{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9996121024049651, "eval_steps": 500, "global_step": 2577, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1412.0, "completions/max_terminated_length": 1412.0, "completions/mean_length": 647.7125244140625, "completions/mean_terminated_length": 647.7125244140625, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.0003878975950349108, "grad_norm": 0.9261648654937744, "kl": 0.0, "learning_rate": 0.0, "loss": 0.0537, "num_tokens": 101713.0, "reward": 0.11124999821186066, "reward_std": 0.11114578694105148, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.13592161238193512, "step": 1 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2048.0, "completions/max_terminated_length": 2048.0, "completions/mean_length": 614.7374877929688, "completions/mean_terminated_length": 614.7374877929688, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "epoch": 0.0007757951900698216, "grad_norm": 1.7771210670471191, "kl": 0.0, "learning_rate": 6.41025641025641e-08, "loss": -0.0179, "num_tokens": 202084.0, "reward": 0.08875000476837158, "reward_std": 0.07942459732294083, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.09409797936677933, "step": 2 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1289.0, "completions/max_terminated_length": 1289.0, "completions/mean_length": 553.9249877929688, "completions/mean_terminated_length": 553.9249877929688, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.0011636927851047323, "grad_norm": 1.122003436088562, "kl": 0.0003743171691894531, "learning_rate": 1.282051282051282e-07, "loss": 0.0181, "num_tokens": 291078.0, "reward": 0.2175000160932541, "reward_std": 0.1725599765777588, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.2872832417488098, "step": 3 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1292.0, "completions/max_terminated_length": 1292.0, "completions/mean_length": 516.5125122070312, "completions/mean_terminated_length": 516.5125122070312, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.0015515903801396431, "grad_norm": 1.370988368988037, "kl": 0.0004405975341796875, "learning_rate": 1.9230769230769234e-07, "loss": 0.0464, "num_tokens": 377295.0, "reward": 0.16375000774860382, "reward_std": 0.17549264430999756, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.23813621699810028, "step": 4 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2464.0, "completions/max_terminated_length": 2464.0, "completions/mean_length": 658.4500122070312, "completions/mean_terminated_length": 658.4500122070312, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "epoch": 0.001939487975174554, "grad_norm": 1.0354154109954834, "kl": 0.00037384033203125, "learning_rate": 2.564102564102564e-07, "loss": 0.0538, "num_tokens": 474275.0, "reward": 0.14750002324581146, "reward_std": 0.08601570874452591, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.08499999344348907, "rewards/unicoder_reward_fn/std": 0.2153007984161377, "step": 5 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1515.0, "completions/max_terminated_length": 1515.0, "completions/mean_length": 685.2250366210938, "completions/mean_terminated_length": 685.2250366210938, "completions/min_length": 254.0, "completions/min_terminated_length": 254.0, "epoch": 0.0023273855702094647, "grad_norm": 1.007523775100708, "kl": 0.0004482269287109375, "learning_rate": 3.205128205128205e-07, "loss": 0.0172, "num_tokens": 586277.0, "reward": 0.1275000125169754, "reward_std": 0.1667095571756363, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.07249999791383743, "rewards/unicoder_reward_fn/std": 0.19420398771762848, "step": 6 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1126.0, "completions/max_terminated_length": 1126.0, "completions/mean_length": 525.9125366210938, "completions/mean_terminated_length": 525.9125366210938, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "epoch": 0.0027152831652443757, "grad_norm": 0.9636631011962891, "kl": 0.0003361701965332031, "learning_rate": 3.846153846153847e-07, "loss": 0.0399, "num_tokens": 671038.0, "reward": 0.2175000160932541, "reward_std": 0.21542425453662872, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.27419620752334595, "step": 7 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1721.0, "completions/max_terminated_length": 1721.0, "completions/mean_length": 606.6000366210938, "completions/mean_terminated_length": 606.6000366210938, "completions/min_length": 324.0, "completions/min_terminated_length": 324.0, "epoch": 0.0031031807602792862, "grad_norm": 1.081107497215271, "kl": 0.0004477500915527344, "learning_rate": 4.4871794871794876e-07, "loss": 0.0523, "num_tokens": 771862.0, "reward": 0.13375000655651093, "reward_std": 0.1293293535709381, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.057499997317790985, "rewards/unicoder_reward_fn/std": 0.16286595165729523, "step": 8 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1419.0, "completions/max_terminated_length": 1419.0, "completions/mean_length": 667.7625122070312, "completions/mean_terminated_length": 667.7625122070312, "completions/min_length": 268.0, "completions/min_terminated_length": 268.0, "epoch": 0.0034910783553141972, "grad_norm": 0.7759409546852112, "kl": 0.0005016326904296875, "learning_rate": 5.128205128205128e-07, "loss": 0.0212, "num_tokens": 870395.0, "reward": 0.11000000685453415, "reward_std": 0.09061162173748016, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.15581552684307098, "step": 9 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 931.0, "completions/max_terminated_length": 931.0, "completions/mean_length": 521.2125244140625, "completions/mean_terminated_length": 521.2125244140625, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.003878975950349108, "grad_norm": 1.088796854019165, "kl": 0.0009222030639648438, "learning_rate": 5.76923076923077e-07, "loss": 0.0058, "num_tokens": 958140.0, "reward": 0.16750000417232513, "reward_std": 0.17442810535430908, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.27374568581581116, "step": 10 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2443.0, "completions/max_terminated_length": 2443.0, "completions/mean_length": 715.2625122070312, "completions/mean_terminated_length": 715.2625122070312, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.004266873545384018, "grad_norm": 0.9076523780822754, "kl": 0.001621246337890625, "learning_rate": 6.41025641025641e-07, "loss": 0.0476, "num_tokens": 1068241.0, "reward": 0.16624999046325684, "reward_std": 0.11536617577075958, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.16441303491592407, "step": 11 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 882.0, "completions/max_terminated_length": 882.0, "completions/mean_length": 547.0625, "completions/mean_terminated_length": 547.0625, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.004654771140418929, "grad_norm": 1.0733799934387207, "kl": 0.0011272430419921875, "learning_rate": 7.051282051282052e-07, "loss": 0.0357, "num_tokens": 1159206.0, "reward": 0.12999999523162842, "reward_std": 0.08741534501314163, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.10542260855436325, "step": 12 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1180.0, "completions/max_terminated_length": 1180.0, "completions/mean_length": 592.6749877929688, "completions/mean_terminated_length": 592.6749877929688, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "epoch": 0.00504266873545384, "grad_norm": 0.7592774033546448, "kl": 0.0010280609130859375, "learning_rate": 7.692307692307694e-07, "loss": 0.0305, "num_tokens": 1247228.0, "reward": 0.21875, "reward_std": 0.18157492578029633, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.2513256072998047, "step": 13 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1179.0, "completions/max_terminated_length": 1179.0, "completions/mean_length": 548.3500366210938, "completions/mean_terminated_length": 548.3500366210938, "completions/min_length": 251.0, "completions/min_terminated_length": 251.0, "epoch": 0.005430566330488751, "grad_norm": 0.7369411587715149, "kl": 0.0018768310546875, "learning_rate": 8.333333333333333e-07, "loss": 0.0587, "num_tokens": 1338736.0, "reward": 0.125, "reward_std": 0.09042035043239594, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.13905194401741028, "step": 14 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1358.0, "completions/max_terminated_length": 1358.0, "completions/mean_length": 546.5499877929688, "completions/mean_terminated_length": 546.5499877929688, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "epoch": 0.0058184639255236615, "grad_norm": 0.7558188438415527, "kl": 0.0006780624389648438, "learning_rate": 8.974358974358975e-07, "loss": 0.0306, "num_tokens": 1426172.0, "reward": 0.18375001847743988, "reward_std": 0.14703132212162018, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.23166976869106293, "step": 15 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1265.0, "completions/max_terminated_length": 1265.0, "completions/mean_length": 625.2250366210938, "completions/mean_terminated_length": 625.2250366210938, "completions/min_length": 350.0, "completions/min_terminated_length": 350.0, "epoch": 0.0062063615205585725, "grad_norm": 0.6734300851821899, "kl": 0.0013523101806640625, "learning_rate": 9.615384615384617e-07, "loss": 0.0293, "num_tokens": 1526998.0, "reward": 0.16500000655651093, "reward_std": 0.08202370256185532, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.17413440346717834, "step": 16 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1308.0, "completions/max_terminated_length": 1308.0, "completions/mean_length": 613.2625122070312, "completions/mean_terminated_length": 613.2625122070312, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.0065942591155934835, "grad_norm": 0.7809385657310486, "kl": 0.0013027191162109375, "learning_rate": 1.0256410256410257e-06, "loss": 0.0853, "num_tokens": 1623715.0, "reward": 0.17750000953674316, "reward_std": 0.13872848451137543, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.18718433380126953, "step": 17 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1330.0, "completions/mean_length": 634.6625366210938, "completions/mean_terminated_length": 590.84814453125, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.0069821567106283944, "grad_norm": 0.8303524851799011, "kl": 0.00140380859375, "learning_rate": 1.0897435897435899e-06, "loss": 0.0078, "num_tokens": 1724760.0, "reward": 0.1562500149011612, "reward_std": 0.11153042316436768, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.08000000566244125, "rewards/unicoder_reward_fn/std": 0.20524761080741882, "step": 18 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1640.0, "completions/max_terminated_length": 1640.0, "completions/mean_length": 630.4125366210938, "completions/mean_terminated_length": 630.4125366210938, "completions/min_length": 242.0, "completions/min_terminated_length": 242.0, "epoch": 0.007370054305663305, "grad_norm": 0.7600616216659546, "kl": 0.001995086669921875, "learning_rate": 1.153846153846154e-06, "loss": 0.0203, "num_tokens": 1826841.0, "reward": 0.13125000894069672, "reward_std": 0.07013656198978424, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.10262380540370941, "step": 19 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1767.0, "completions/max_terminated_length": 1767.0, "completions/mean_length": 532.8875122070312, "completions/mean_terminated_length": 532.8875122070312, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "epoch": 0.007757951900698216, "grad_norm": 0.7621126174926758, "kl": 0.002109527587890625, "learning_rate": 1.217948717948718e-06, "loss": 0.0358, "num_tokens": 1915344.0, "reward": 0.1575000137090683, "reward_std": 0.08701261132955551, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.14372971653938293, "step": 20 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1382.0, "completions/max_terminated_length": 1382.0, "completions/mean_length": 516.1000366210938, "completions/mean_terminated_length": 516.1000366210938, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.008145849495733127, "grad_norm": 0.6807729005813599, "kl": 0.00177001953125, "learning_rate": 1.282051282051282e-06, "loss": 0.0098, "num_tokens": 2003216.0, "reward": 0.20624999701976776, "reward_std": 0.10166865587234497, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.2544664442539215, "step": 21 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1223.0, "completions/max_terminated_length": 1223.0, "completions/mean_length": 511.0625, "completions/mean_terminated_length": 511.0625, "completions/min_length": 241.0, "completions/min_terminated_length": 241.0, "epoch": 0.008533747090768037, "grad_norm": 0.8069402575492859, "kl": 0.002410888671875, "learning_rate": 1.3461538461538462e-06, "loss": 0.0199, "num_tokens": 2093701.0, "reward": 0.23499999940395355, "reward_std": 0.12779128551483154, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.26355651021003723, "step": 22 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1271.0, "completions/max_terminated_length": 1271.0, "completions/mean_length": 591.7000122070312, "completions/mean_terminated_length": 591.7000122070312, "completions/min_length": 224.0, "completions/min_terminated_length": 224.0, "epoch": 0.008921644685802949, "grad_norm": 0.750952422618866, "kl": 0.00327301025390625, "learning_rate": 1.4102564102564104e-06, "loss": 0.0361, "num_tokens": 2195285.0, "reward": 0.125, "reward_std": 0.08125855773687363, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.030000001192092896, "rewards/unicoder_reward_fn/std": 0.12768632173538208, "step": 23 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1226.0, "completions/max_terminated_length": 1226.0, "completions/mean_length": 583.7875366210938, "completions/mean_terminated_length": 583.7875366210938, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.009309542280837859, "grad_norm": 0.40481671690940857, "kl": 0.0019130706787109375, "learning_rate": 1.4743589743589745e-06, "loss": 0.0197, "num_tokens": 2285652.0, "reward": 0.17000000178813934, "reward_std": 0.07045123726129532, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.21370746195316315, "step": 24 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1655.0, "completions/max_terminated_length": 1655.0, "completions/mean_length": 575.7125244140625, "completions/mean_terminated_length": 575.7125244140625, "completions/min_length": 284.0, "completions/min_terminated_length": 284.0, "epoch": 0.009697439875872769, "grad_norm": 73.037109375, "kl": 0.4680023193359375, "learning_rate": 1.5384615384615387e-06, "loss": 0.0144, "num_tokens": 2379549.0, "reward": 0.20125000178813934, "reward_std": 0.16601689159870148, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.10500000417232513, "rewards/unicoder_reward_fn/std": 0.2661968469619751, "step": 25 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1382.0, "completions/max_terminated_length": 1382.0, "completions/mean_length": 533.8500366210938, "completions/mean_terminated_length": 533.8500366210938, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.01008533747090768, "grad_norm": 0.6820522546768188, "kl": 0.004451751708984375, "learning_rate": 1.602564102564103e-06, "loss": 0.0075, "num_tokens": 2468401.0, "reward": 0.24625001847743988, "reward_std": 0.12913106381893158, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.3225589990615845, "step": 26 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2085.0, "completions/max_terminated_length": 2085.0, "completions/mean_length": 576.0375366210938, "completions/mean_terminated_length": 576.0375366210938, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.01047323506594259, "grad_norm": 0.8537800312042236, "kl": 0.003826141357421875, "learning_rate": 1.6666666666666667e-06, "loss": 0.0132, "num_tokens": 2558460.0, "reward": 0.20375001430511475, "reward_std": 0.10090043395757675, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.10500000417232513, "rewards/unicoder_reward_fn/std": 0.27369946241378784, "step": 27 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1658.0, "completions/max_terminated_length": 1658.0, "completions/mean_length": 622.75, "completions/mean_terminated_length": 622.75, "completions/min_length": 251.0, "completions/min_terminated_length": 251.0, "epoch": 0.010861132660977503, "grad_norm": 0.5907183885574341, "kl": 0.003398895263671875, "learning_rate": 1.7307692307692308e-06, "loss": 0.0315, "num_tokens": 2663168.0, "reward": 0.14250001311302185, "reward_std": 0.05506539344787598, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.042500000447034836, "rewards/unicoder_reward_fn/std": 0.10406302660703659, "step": 28 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1868.0, "completions/max_terminated_length": 1868.0, "completions/mean_length": 602.9750366210938, "completions/mean_terminated_length": 602.9750366210938, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.011249030256012413, "grad_norm": 0.7479941844940186, "kl": 0.004856109619140625, "learning_rate": 1.794871794871795e-06, "loss": -0.0, "num_tokens": 2760286.0, "reward": 0.13375000655651093, "reward_std": 0.07486527413129807, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.12363174557685852, "step": 29 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1528.0, "completions/max_terminated_length": 1528.0, "completions/mean_length": 521.1000366210938, "completions/mean_terminated_length": 521.1000366210938, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.011636927851047323, "grad_norm": 0.8341169357299805, "kl": 0.005218505859375, "learning_rate": 1.8589743589743592e-06, "loss": 0.0081, "num_tokens": 2842750.0, "reward": 0.20750001072883606, "reward_std": 0.10487490892410278, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.27178502082824707, "step": 30 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 853.0, "completions/max_terminated_length": 853.0, "completions/mean_length": 433.8125, "completions/mean_terminated_length": 433.8125, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.012024825446082235, "grad_norm": 0.8686825037002563, "kl": 0.0073699951171875, "learning_rate": 1.9230769230769234e-06, "loss": -0.0161, "num_tokens": 2916335.0, "reward": 0.30000001192092896, "reward_std": 0.15951655805110931, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.3724703788757324, "step": 31 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1438.0, "completions/max_terminated_length": 1438.0, "completions/mean_length": 520.7750244140625, "completions/mean_terminated_length": 520.7750244140625, "completions/min_length": 277.0, "completions/min_terminated_length": 277.0, "epoch": 0.012412723041117145, "grad_norm": 0.4418981671333313, "kl": 0.0067138671875, "learning_rate": 1.987179487179487e-06, "loss": 0.0011, "num_tokens": 3011461.0, "reward": 0.12250000238418579, "reward_std": 0.029022613540291786, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.022499999031424522, "rewards/unicoder_reward_fn/std": 0.07790776342153549, "step": 32 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1589.0, "completions/max_terminated_length": 1589.0, "completions/mean_length": 527.2125244140625, "completions/mean_terminated_length": 527.2125244140625, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.012800620636152055, "grad_norm": 0.4720829129219055, "kl": 0.00836944580078125, "learning_rate": 2.0512820512820513e-06, "loss": 0.0011, "num_tokens": 3106454.0, "reward": 0.11874999850988388, "reward_std": 0.03773065656423569, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.020000001415610313, "rewards/unicoder_reward_fn/std": 0.09859776496887207, "step": 33 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 839.0, "completions/max_terminated_length": 839.0, "completions/mean_length": 426.6125183105469, "completions/mean_terminated_length": 426.6125183105469, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.013188518231186967, "grad_norm": 1.0938643217086792, "kl": 0.011016845703125, "learning_rate": 2.1153846153846155e-06, "loss": 0.0006, "num_tokens": 3181735.0, "reward": 0.19750000536441803, "reward_std": 0.17248670756816864, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.19091716408729553, "step": 34 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1004.0, "completions/max_terminated_length": 1004.0, "completions/mean_length": 499.3374938964844, "completions/mean_terminated_length": 499.3374938964844, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.013576415826221877, "grad_norm": 0.5843201875686646, "kl": 0.00885772705078125, "learning_rate": 2.1794871794871797e-06, "loss": 0.0155, "num_tokens": 3273106.0, "reward": 0.13125000894069672, "reward_std": 0.06271100044250488, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.032499998807907104, "rewards/unicoder_reward_fn/std": 0.12505695223808289, "step": 35 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1476.0, "completions/max_terminated_length": 1476.0, "completions/mean_length": 534.7750244140625, "completions/mean_terminated_length": 534.7750244140625, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.013964313421256789, "grad_norm": 0.5502732992172241, "kl": 0.00862884521484375, "learning_rate": 2.243589743589744e-06, "loss": 0.0339, "num_tokens": 3368440.0, "reward": 0.14375001192092896, "reward_std": 0.05580168962478638, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.04500000178813934, "rewards/unicoder_reward_fn/std": 0.15253138542175293, "step": 36 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1676.0, "completions/max_terminated_length": 1676.0, "completions/mean_length": 524.5625, "completions/mean_terminated_length": 524.5625, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.014352211016291699, "grad_norm": 0.8830205798149109, "kl": 0.0152587890625, "learning_rate": 2.307692307692308e-06, "loss": 0.0025, "num_tokens": 3462117.0, "reward": 0.1900000125169754, "reward_std": 0.10545708984136581, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09000000357627869, "rewards/unicoder_reward_fn/std": 0.19066838920116425, "step": 37 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 955.0, "completions/max_terminated_length": 955.0, "completions/mean_length": 493.3000183105469, "completions/mean_terminated_length": 493.3000183105469, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.01474010861132661, "grad_norm": 0.9368318915367126, "kl": 0.009796142578125, "learning_rate": 2.371794871794872e-06, "loss": 0.0202, "num_tokens": 3549693.0, "reward": 0.2750000059604645, "reward_std": 0.16001980006694794, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.2513256072998047, "step": 38 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 855.0, "completions/max_terminated_length": 855.0, "completions/mean_length": 449.3999938964844, "completions/mean_terminated_length": 449.3999938964844, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.015128006206361521, "grad_norm": 1.020815372467041, "kl": 0.0263671875, "learning_rate": 2.435897435897436e-06, "loss": 0.0103, "num_tokens": 3641509.0, "reward": 0.14750000834465027, "reward_std": 0.06859543174505234, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.10670555382966995, "step": 39 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 853.0, "completions/max_terminated_length": 853.0, "completions/mean_length": 438.4375, "completions/mean_terminated_length": 438.4375, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.015515903801396431, "grad_norm": 0.752341628074646, "kl": 0.0147705078125, "learning_rate": 2.5e-06, "loss": 0.0118, "num_tokens": 3721736.0, "reward": 0.20250001549720764, "reward_std": 0.07891489565372467, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.10249999910593033, "rewards/unicoder_reward_fn/std": 0.2505563497543335, "step": 40 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1174.0, "completions/max_terminated_length": 1174.0, "completions/mean_length": 463.51251220703125, "completions/mean_terminated_length": 463.51251220703125, "completions/min_length": 159.0, "completions/min_terminated_length": 159.0, "epoch": 0.01590380139643134, "grad_norm": 0.806076169013977, "kl": 0.0181427001953125, "learning_rate": 2.564102564102564e-06, "loss": 0.0449, "num_tokens": 3810673.0, "reward": 0.3125000298023224, "reward_std": 0.13470660150051117, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.21250000596046448, "rewards/unicoder_reward_fn/std": 0.3796650767326355, "step": 41 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1117.0, "completions/max_terminated_length": 1117.0, "completions/mean_length": 473.0500183105469, "completions/mean_terminated_length": 473.0500183105469, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.016291698991466253, "grad_norm": 0.8992795348167419, "kl": 0.0135345458984375, "learning_rate": 2.6282051282051286e-06, "loss": -0.0142, "num_tokens": 3900293.0, "reward": 0.16750000417232513, "reward_std": 0.0921839103102684, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.13850468397140503, "step": 42 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 996.0, "completions/max_terminated_length": 996.0, "completions/mean_length": 430.1750183105469, "completions/mean_terminated_length": 430.1750183105469, "completions/min_length": 202.0, "completions/min_terminated_length": 202.0, "epoch": 0.016679596586501165, "grad_norm": 0.7114393711090088, "kl": 0.0145263671875, "learning_rate": 2.6923076923076923e-06, "loss": -0.016, "num_tokens": 3989467.0, "reward": 0.13750000298023224, "reward_std": 0.05973212048411369, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.11946611851453781, "step": 43 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 737.0, "completions/max_terminated_length": 737.0, "completions/mean_length": 387.3500061035156, "completions/mean_terminated_length": 387.3500061035156, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.017067494181536073, "grad_norm": 0.9069808721542358, "kl": 0.01361083984375, "learning_rate": 2.756410256410257e-06, "loss": 0.0145, "num_tokens": 4069823.0, "reward": 0.2525000274181366, "reward_std": 0.1630178540945053, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.24493606388568878, "step": 44 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 945.0, "completions/max_terminated_length": 945.0, "completions/mean_length": 424.6625061035156, "completions/mean_terminated_length": 424.6625061035156, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.017455391776570985, "grad_norm": 0.8797289133071899, "kl": 0.0135498046875, "learning_rate": 2.8205128205128207e-06, "loss": 0.0075, "num_tokens": 4145788.0, "reward": 0.21250000596046448, "reward_std": 0.1218840628862381, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.2640363276004791, "step": 45 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 971.0, "completions/max_terminated_length": 971.0, "completions/mean_length": 436.3125, "completions/mean_terminated_length": 436.3125, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.017843289371605897, "grad_norm": 0.8683922290802002, "kl": 0.0163421630859375, "learning_rate": 2.8846153846153845e-06, "loss": 0.008, "num_tokens": 4227965.0, "reward": 0.2262500375509262, "reward_std": 0.10596662014722824, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.2570401430130005, "step": 46 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1003.0, "completions/max_terminated_length": 1003.0, "completions/mean_length": 480.875, "completions/mean_terminated_length": 480.875, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.018231186966640806, "grad_norm": 0.687198281288147, "kl": 0.0150604248046875, "learning_rate": 2.948717948717949e-06, "loss": 0.009, "num_tokens": 4321507.0, "reward": 0.26500001549720764, "reward_std": 0.10024459660053253, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.30095839500427246, "step": 47 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1993.0, "completions/max_terminated_length": 1993.0, "completions/mean_length": 498.38751220703125, "completions/mean_terminated_length": 498.38751220703125, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.018619084561675717, "grad_norm": 0.8283015489578247, "kl": 0.0171966552734375, "learning_rate": 3.012820512820513e-06, "loss": -0.0152, "num_tokens": 4408962.0, "reward": 0.2900000214576721, "reward_std": 0.20099197328090668, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1900000125169754, "rewards/unicoder_reward_fn/std": 0.2897401750087738, "step": 48 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1344.0, "completions/max_terminated_length": 1344.0, "completions/mean_length": 514.2000122070312, "completions/mean_terminated_length": 514.2000122070312, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.01900698215671063, "grad_norm": 0.5997982621192932, "kl": 0.014404296875, "learning_rate": 3.0769230769230774e-06, "loss": 0.0266, "num_tokens": 4499138.0, "reward": 0.18500001728534698, "reward_std": 0.10362284630537033, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.08499999344348907, "rewards/unicoder_reward_fn/std": 0.24188081920146942, "step": 49 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 897.0, "completions/max_terminated_length": 897.0, "completions/mean_length": 412.8374938964844, "completions/mean_terminated_length": 412.8374938964844, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.019394879751745538, "grad_norm": 0.8736696839332581, "kl": 0.015411376953125, "learning_rate": 3.141025641025641e-06, "loss": 0.045, "num_tokens": 4577301.0, "reward": 0.26250001788139343, "reward_std": 0.1321004182100296, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.3003689646720886, "step": 50 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1402.0, "completions/max_terminated_length": 1402.0, "completions/mean_length": 508.51251220703125, "completions/mean_terminated_length": 508.51251220703125, "completions/min_length": 169.0, "completions/min_terminated_length": 169.0, "epoch": 0.01978277734678045, "grad_norm": 0.5378665924072266, "kl": 0.016998291015625, "learning_rate": 3.205128205128206e-06, "loss": -0.0107, "num_tokens": 4670422.0, "reward": 0.16750000417232513, "reward_std": 0.08090989291667938, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.21333695948123932, "step": 51 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1104.0, "completions/max_terminated_length": 1104.0, "completions/mean_length": 468.125, "completions/mean_terminated_length": 468.125, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.02017067494181536, "grad_norm": 0.7122435569763184, "kl": 0.017669677734375, "learning_rate": 3.2692307692307696e-06, "loss": 0.0332, "num_tokens": 4759616.0, "reward": 0.17499999701976776, "reward_std": 0.09049614518880844, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.2071506679058075, "step": 52 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 943.0, "completions/max_terminated_length": 943.0, "completions/mean_length": 424.875, "completions/mean_terminated_length": 424.875, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.020558572536850273, "grad_norm": 0.9005601406097412, "kl": 0.02008056640625, "learning_rate": 3.3333333333333333e-06, "loss": 0.0315, "num_tokens": 4837350.0, "reward": 0.2200000137090683, "reward_std": 0.13544128835201263, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.1671806424856186, "step": 53 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1005.0, "completions/max_terminated_length": 1005.0, "completions/mean_length": 445.32501220703125, "completions/mean_terminated_length": 445.32501220703125, "completions/min_length": 196.0, "completions/min_terminated_length": 196.0, "epoch": 0.02094647013188518, "grad_norm": 0.6599981784820557, "kl": 0.02069091796875, "learning_rate": 3.397435897435898e-06, "loss": 0.0039, "num_tokens": 4917720.0, "reward": 0.19500000774860382, "reward_std": 0.07491317391395569, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.2316151261329651, "step": 54 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 811.0, "completions/max_terminated_length": 811.0, "completions/mean_length": 391.375, "completions/mean_terminated_length": 391.375, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.021334367726920093, "grad_norm": 0.9016393423080444, "kl": 0.0234375, "learning_rate": 3.4615384615384617e-06, "loss": -0.0166, "num_tokens": 4997118.0, "reward": 0.18125002086162567, "reward_std": 0.12434946745634079, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.20792831480503082, "step": 55 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1142.0, "completions/max_terminated_length": 1142.0, "completions/mean_length": 551.2000122070312, "completions/mean_terminated_length": 551.2000122070312, "completions/min_length": 259.0, "completions/min_terminated_length": 259.0, "epoch": 0.021722265321955005, "grad_norm": 0.5501932501792908, "kl": 0.0154876708984375, "learning_rate": 3.5256410256410263e-06, "loss": 0.0287, "num_tokens": 5094318.0, "reward": 0.16250000894069672, "reward_std": 0.06876759976148605, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.15126053988933563, "step": 56 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1092.0, "completions/max_terminated_length": 1092.0, "completions/mean_length": 481.6750183105469, "completions/mean_terminated_length": 481.6750183105469, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.022110162916989914, "grad_norm": 0.6910412907600403, "kl": 0.019287109375, "learning_rate": 3.58974358974359e-06, "loss": -0.0146, "num_tokens": 5183268.0, "reward": 0.1550000011920929, "reward_std": 0.09150350093841553, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.15325580537319183, "step": 57 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1198.0, "completions/max_terminated_length": 1198.0, "completions/mean_length": 381.26251220703125, "completions/mean_terminated_length": 381.26251220703125, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.022498060512024826, "grad_norm": 0.9556728601455688, "kl": 0.02484130859375, "learning_rate": 3.653846153846154e-06, "loss": 0.0303, "num_tokens": 5255777.0, "reward": 0.3799999952316284, "reward_std": 0.20241861045360565, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.2825000286102295, "rewards/unicoder_reward_fn/std": 0.3665698170661926, "step": 58 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1065.0, "completions/max_terminated_length": 1065.0, "completions/mean_length": 550.2625122070312, "completions/mean_terminated_length": 550.2625122070312, "completions/min_length": 234.0, "completions/min_terminated_length": 234.0, "epoch": 0.022885958107059737, "grad_norm": 0.7146815657615662, "kl": 0.015716552734375, "learning_rate": 3.7179487179487184e-06, "loss": 0.017, "num_tokens": 5342398.0, "reward": 0.2224999964237213, "reward_std": 0.16440020501613617, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12250001728534698, "rewards/unicoder_reward_fn/std": 0.2728077173233032, "step": 59 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 883.0, "completions/max_terminated_length": 883.0, "completions/mean_length": 453.125, "completions/mean_terminated_length": 453.125, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.023273855702094646, "grad_norm": 0.8617904782295227, "kl": 0.019317626953125, "learning_rate": 3.782051282051282e-06, "loss": 0.0219, "num_tokens": 5418296.0, "reward": 0.22874999046325684, "reward_std": 0.14883527159690857, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.2307801991701126, "step": 60 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1365.0, "completions/max_terminated_length": 1365.0, "completions/mean_length": 457.8500061035156, "completions/mean_terminated_length": 457.8500061035156, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.023661753297129558, "grad_norm": 0.920381486415863, "kl": 0.026153564453125, "learning_rate": 3.846153846153847e-06, "loss": -0.0079, "num_tokens": 5500972.0, "reward": 0.27000001072883606, "reward_std": 0.23184971511363983, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.31357473134994507, "step": 61 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1372.0, "completions/max_terminated_length": 1372.0, "completions/mean_length": 528.9125366210938, "completions/mean_terminated_length": 528.9125366210938, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "epoch": 0.02404965089216447, "grad_norm": 0.720321774482727, "kl": 0.023712158203125, "learning_rate": 3.910256410256411e-06, "loss": 0.0023, "num_tokens": 5592965.0, "reward": 0.17999999225139618, "reward_std": 0.13727466762065887, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.18445952236652374, "step": 62 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1626.0, "completions/max_terminated_length": 1626.0, "completions/mean_length": 528.375, "completions/mean_terminated_length": 528.375, "completions/min_length": 239.0, "completions/min_terminated_length": 239.0, "epoch": 0.024437548487199378, "grad_norm": 0.6912801861763, "kl": 0.02313232421875, "learning_rate": 3.974358974358974e-06, "loss": 0.0381, "num_tokens": 5682043.0, "reward": 0.22374999523162842, "reward_std": 0.1832731068134308, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.24726349115371704, "step": 63 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1723.0, "completions/max_terminated_length": 1723.0, "completions/mean_length": 491.0500183105469, "completions/mean_terminated_length": 491.0500183105469, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.02482544608223429, "grad_norm": 0.7913193702697754, "kl": 0.025665283203125, "learning_rate": 4.0384615384615385e-06, "loss": 0.011, "num_tokens": 5770791.0, "reward": 0.29500001668930054, "reward_std": 0.14022321999073029, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19499999284744263, "rewards/unicoder_reward_fn/std": 0.2791941463947296, "step": 64 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1562.0, "completions/max_terminated_length": 1562.0, "completions/mean_length": 467.875, "completions/mean_terminated_length": 467.875, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.0252133436772692, "grad_norm": 0.7653983235359192, "kl": 0.03021240234375, "learning_rate": 4.102564102564103e-06, "loss": -0.0033, "num_tokens": 5862301.0, "reward": 0.24500000476837158, "reward_std": 0.13780155777931213, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14499999582767487, "rewards/unicoder_reward_fn/std": 0.26236510276794434, "step": 65 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1280.0, "completions/max_terminated_length": 1280.0, "completions/mean_length": 530.25, "completions/mean_terminated_length": 530.25, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.02560124127230411, "grad_norm": 0.812148928642273, "kl": 0.02197265625, "learning_rate": 4.166666666666667e-06, "loss": -0.0063, "num_tokens": 5946729.0, "reward": 0.2474999874830246, "reward_std": 0.13942794501781464, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.2787744402885437, "step": 66 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1135.0, "completions/max_terminated_length": 1135.0, "completions/mean_length": 463.45001220703125, "completions/mean_terminated_length": 463.45001220703125, "completions/min_length": 259.0, "completions/min_terminated_length": 259.0, "epoch": 0.025989138867339022, "grad_norm": 0.9509928226470947, "kl": 0.032562255859375, "learning_rate": 4.230769230769231e-06, "loss": -0.0239, "num_tokens": 6025485.0, "reward": 0.29500001668930054, "reward_std": 0.15629923343658447, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.19750002026557922, "rewards/unicoder_reward_fn/std": 0.2652798891067505, "step": 67 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1261.0, "completions/max_terminated_length": 1261.0, "completions/mean_length": 518.0875244140625, "completions/mean_terminated_length": 518.0875244140625, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.026377036462373934, "grad_norm": 1.4543869495391846, "kl": 0.05853271484375, "learning_rate": 4.294871794871795e-06, "loss": 0.0545, "num_tokens": 6118716.0, "reward": 0.29749998450279236, "reward_std": 0.19831015169620514, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.2985195815563202, "step": 68 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1014.0, "completions/max_terminated_length": 1014.0, "completions/mean_length": 476.8500061035156, "completions/mean_terminated_length": 476.8500061035156, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.026764934057408846, "grad_norm": 0.7347418665885925, "kl": 0.0299072265625, "learning_rate": 4.358974358974359e-06, "loss": 0.0156, "num_tokens": 6201664.0, "reward": 0.17249999940395355, "reward_std": 0.10632925480604172, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.1660981923341751, "step": 69 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 988.0, "completions/max_terminated_length": 988.0, "completions/mean_length": 489.88751220703125, "completions/mean_terminated_length": 489.88751220703125, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.027152831652443754, "grad_norm": 0.5525808930397034, "kl": 0.0347900390625, "learning_rate": 4.423076923076924e-06, "loss": -0.0129, "num_tokens": 6288775.0, "reward": 0.17000000178813934, "reward_std": 0.11483939737081528, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.22969433665275574, "step": 70 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 844.0, "completions/max_terminated_length": 844.0, "completions/mean_length": 426.2250061035156, "completions/mean_terminated_length": 426.2250061035156, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.027540729247478666, "grad_norm": 0.8707143068313599, "kl": 0.0367431640625, "learning_rate": 4.487179487179488e-06, "loss": 0.0202, "num_tokens": 6363049.0, "reward": 0.21250000596046448, "reward_std": 0.14800523221492767, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.20584805309772491, "step": 71 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 900.0, "completions/max_terminated_length": 900.0, "completions/mean_length": 512.875, "completions/mean_terminated_length": 512.875, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.027928626842513578, "grad_norm": 0.7828732132911682, "kl": 0.03570556640625, "learning_rate": 4.551282051282052e-06, "loss": 0.0055, "num_tokens": 6451663.0, "reward": 0.2449999898672104, "reward_std": 0.16403710842132568, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2810018062591553, "step": 72 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 989.0, "completions/max_terminated_length": 989.0, "completions/mean_length": 486.8000183105469, "completions/mean_terminated_length": 486.8000183105469, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.028316524437548486, "grad_norm": 0.623846709728241, "kl": 0.036376953125, "learning_rate": 4.615384615384616e-06, "loss": 0.0199, "num_tokens": 6545567.0, "reward": 0.21500001847743988, "reward_std": 0.12787054479122162, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.24396513402462006, "step": 73 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1043.0, "completions/max_terminated_length": 1043.0, "completions/mean_length": 503.7875061035156, "completions/mean_terminated_length": 503.7875061035156, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.028704422032583398, "grad_norm": 0.568273663520813, "kl": 0.03680419921875, "learning_rate": 4.6794871794871795e-06, "loss": 0.0147, "num_tokens": 6630110.0, "reward": 0.25, "reward_std": 0.0500543974339962, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.31421995162963867, "step": 74 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 706.0, "completions/max_terminated_length": 706.0, "completions/mean_length": 427.2375183105469, "completions/mean_terminated_length": 427.2375183105469, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "epoch": 0.02909231962761831, "grad_norm": 0.4973991811275482, "kl": 0.0379638671875, "learning_rate": 4.743589743589744e-06, "loss": -0.0031, "num_tokens": 6711025.0, "reward": 0.17249999940395355, "reward_std": 0.06750421971082687, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.20187413692474365, "step": 75 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 714.0, "completions/max_terminated_length": 714.0, "completions/mean_length": 373.7124938964844, "completions/mean_terminated_length": 373.7124938964844, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.02948021722265322, "grad_norm": 0.9061908721923828, "kl": 0.0528564453125, "learning_rate": 4.807692307692308e-06, "loss": 0.0068, "num_tokens": 6785858.0, "reward": 0.2900000214576721, "reward_std": 0.17934346199035645, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1900000125169754, "rewards/unicoder_reward_fn/std": 0.2660541534423828, "step": 76 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 727.0, "completions/max_terminated_length": 727.0, "completions/mean_length": 353.7250061035156, "completions/mean_terminated_length": 353.7250061035156, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.02986811481768813, "grad_norm": 0.9673826694488525, "kl": 0.05596923828125, "learning_rate": 4.871794871794872e-06, "loss": 0.0148, "num_tokens": 6861828.0, "reward": 0.35750001668930054, "reward_std": 0.1838989555835724, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2574999928474426, "rewards/unicoder_reward_fn/std": 0.3430226743221283, "step": 77 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 746.0, "completions/max_terminated_length": 746.0, "completions/mean_length": 372.8374938964844, "completions/mean_terminated_length": 372.8374938964844, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.030256012412723042, "grad_norm": 0.8349303603172302, "kl": 0.04486083984375, "learning_rate": 4.935897435897436e-06, "loss": -0.0007, "num_tokens": 6940175.0, "reward": 0.30250000953674316, "reward_std": 0.18500085175037384, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.32684725522994995, "step": 78 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1150.0, "completions/max_terminated_length": 1150.0, "completions/mean_length": 442.95001220703125, "completions/mean_terminated_length": 442.95001220703125, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.03064391000775795, "grad_norm": 0.77904212474823, "kl": 0.05303955078125, "learning_rate": 5e-06, "loss": 0.0393, "num_tokens": 7032619.0, "reward": 0.26375001668930054, "reward_std": 0.15441323816776276, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.16499999165534973, "rewards/unicoder_reward_fn/std": 0.31786471605300903, "step": 79 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 971.0, "completions/max_terminated_length": 971.0, "completions/mean_length": 392.9875183105469, "completions/mean_terminated_length": 392.9875183105469, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.031031807602792862, "grad_norm": 0.6622040271759033, "kl": 0.05255126953125, "learning_rate": 4.999998223471442e-06, "loss": 0.0029, "num_tokens": 7114306.0, "reward": 0.19500000774860382, "reward_std": 0.10067303478717804, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0949999988079071, "rewards/unicoder_reward_fn/std": 0.1799437552690506, "step": 80 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 891.0, "completions/max_terminated_length": 891.0, "completions/mean_length": 372.7375183105469, "completions/mean_terminated_length": 372.7375183105469, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.031419705197827774, "grad_norm": 0.9596009254455566, "kl": 0.0465087890625, "learning_rate": 4.999992893888572e-06, "loss": 0.0038, "num_tokens": 7180237.0, "reward": 0.3400000035762787, "reward_std": 0.1611824333667755, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24000000953674316, "rewards/unicoder_reward_fn/std": 0.3025628924369812, "step": 81 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1574.0, "completions/max_terminated_length": 1574.0, "completions/mean_length": 411.63751220703125, "completions/mean_terminated_length": 411.63751220703125, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.03180760279286268, "grad_norm": 0.7552454471588135, "kl": 0.0496826171875, "learning_rate": 4.999984011259806e-06, "loss": 0.0144, "num_tokens": 7262960.0, "reward": 0.21500001847743988, "reward_std": 0.0806669369339943, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.282439649105072, "step": 82 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 847.0, "completions/max_terminated_length": 847.0, "completions/mean_length": 426.20001220703125, "completions/mean_terminated_length": 426.20001220703125, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.0321955003878976, "grad_norm": 0.7512875199317932, "kl": 0.0465087890625, "learning_rate": 4.999971575599174e-06, "loss": -0.012, "num_tokens": 7341096.0, "reward": 0.24750001728534698, "reward_std": 0.09416606277227402, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.23328964412212372, "step": 83 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 732.0, "completions/max_terminated_length": 732.0, "completions/mean_length": 386.5874938964844, "completions/mean_terminated_length": 386.5874938964844, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.032583397982932506, "grad_norm": 0.886727511882782, "kl": 0.0516357421875, "learning_rate": 4.9999555869263085e-06, "loss": 0.005, "num_tokens": 7414799.0, "reward": 0.23749999701976776, "reward_std": 0.15079715847969055, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.2456585168838501, "step": 84 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 798.0, "completions/max_terminated_length": 798.0, "completions/mean_length": 463.70001220703125, "completions/mean_terminated_length": 463.70001220703125, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.032971295577967415, "grad_norm": 0.6678573489189148, "kl": 0.05560302734375, "learning_rate": 4.9999360452664626e-06, "loss": -0.0107, "num_tokens": 7504895.0, "reward": 0.17000000178813934, "reward_std": 0.11974883079528809, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.18853197991847992, "step": 85 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1119.0, "completions/max_terminated_length": 1119.0, "completions/mean_length": 440.6499938964844, "completions/mean_terminated_length": 440.6499938964844, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.03335919317300233, "grad_norm": 0.7660518884658813, "kl": 0.05499267578125, "learning_rate": 4.999912950650492e-06, "loss": 0.0299, "num_tokens": 7585155.0, "reward": 0.2775000035762787, "reward_std": 0.16469421982765198, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.31015512347221375, "step": 86 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 852.0, "completions/max_terminated_length": 852.0, "completions/mean_length": 447.63751220703125, "completions/mean_terminated_length": 447.63751220703125, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.03374709076803724, "grad_norm": 0.6622053384780884, "kl": 0.07464599609375, "learning_rate": 4.999886303114866e-06, "loss": 0.0284, "num_tokens": 7661806.0, "reward": 0.26500001549720764, "reward_std": 0.09573165327310562, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16499999165534973, "rewards/unicoder_reward_fn/std": 0.29126524925231934, "step": 87 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 848.0, "completions/max_terminated_length": 848.0, "completions/mean_length": 448.8000183105469, "completions/mean_terminated_length": 448.8000183105469, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "epoch": 0.03413498836307215, "grad_norm": 0.8376213312149048, "kl": 0.0498046875, "learning_rate": 4.999856102701668e-06, "loss": -0.0285, "num_tokens": 7741598.0, "reward": 0.2175000160932541, "reward_std": 0.11794539541006088, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.18471668660640717, "step": 88 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 854.0, "completions/max_terminated_length": 854.0, "completions/mean_length": 484.5, "completions/mean_terminated_length": 484.5, "completions/min_length": 251.0, "completions/min_terminated_length": 251.0, "epoch": 0.03452288595810706, "grad_norm": 0.4069496989250183, "kl": 0.050048828125, "learning_rate": 4.999822349458587e-06, "loss": 0.0, "num_tokens": 7827214.0, "reward": 0.19500000774860382, "reward_std": 0.03265853971242905, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0949999988079071, "rewards/unicoder_reward_fn/std": 0.27737465500831604, "step": 89 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2080.0, "completions/max_terminated_length": 2080.0, "completions/mean_length": 538.1375122070312, "completions/mean_terminated_length": 538.1375122070312, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.03491078355314197, "grad_norm": 292.2913513183594, "kl": 8.17431640625, "learning_rate": 4.999785043438922e-06, "loss": 0.0668, "num_tokens": 7918913.0, "reward": 0.2224999964237213, "reward_std": 0.11574418842792511, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12250000238418579, "rewards/unicoder_reward_fn/std": 0.25156474113464355, "step": 90 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1453.0, "completions/max_terminated_length": 1453.0, "completions/mean_length": 520.5625, "completions/mean_terminated_length": 520.5625, "completions/min_length": 234.0, "completions/min_terminated_length": 234.0, "epoch": 0.03529868114817688, "grad_norm": 0.5854523181915283, "kl": 0.05145263671875, "learning_rate": 4.999744184701587e-06, "loss": -0.0006, "num_tokens": 8005822.0, "reward": 0.2200000137090683, "reward_std": 0.15373609960079193, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.28921544551849365, "step": 91 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1201.0, "completions/max_terminated_length": 1201.0, "completions/mean_length": 549.9625244140625, "completions/mean_terminated_length": 549.9625244140625, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "epoch": 0.035686578743211794, "grad_norm": 0.5861818790435791, "kl": 0.05718994140625, "learning_rate": 4.999699773311102e-06, "loss": 0.0289, "num_tokens": 8109627.0, "reward": 0.18500001728534698, "reward_std": 0.12647615373134613, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.20567886531352997, "step": 92 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1461.0, "completions/max_terminated_length": 1461.0, "completions/mean_length": 563.2750244140625, "completions/mean_terminated_length": 563.2750244140625, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "epoch": 0.0360744763382467, "grad_norm": 0.7537212371826172, "kl": 0.05841064453125, "learning_rate": 4.9996518093376e-06, "loss": 0.006, "num_tokens": 8198377.0, "reward": 0.4162500500679016, "reward_std": 0.26385578513145447, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.3175000250339508, "rewards/unicoder_reward_fn/std": 0.4096231758594513, "step": 93 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1384.0, "completions/max_terminated_length": 1384.0, "completions/mean_length": 592.2125244140625, "completions/mean_terminated_length": 592.2125244140625, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "epoch": 0.03646237393328161, "grad_norm": 0.5742750763893127, "kl": 0.06494140625, "learning_rate": 4.999600292856821e-06, "loss": 0.0173, "num_tokens": 8288450.0, "reward": 0.26250001788139343, "reward_std": 0.18922506272792816, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.16499999165534973, "rewards/unicoder_reward_fn/std": 0.29813769459724426, "step": 94 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1234.0, "completions/max_terminated_length": 1234.0, "completions/mean_length": 674.2999877929688, "completions/mean_terminated_length": 674.2999877929688, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.036850271528316526, "grad_norm": 0.7128031253814697, "kl": 0.0731201171875, "learning_rate": 4.999545223950117e-06, "loss": 0.0243, "num_tokens": 8387658.0, "reward": 0.21375000476837158, "reward_std": 0.16922013461589813, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.2300110012292862, "step": 95 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2346.0, "completions/max_terminated_length": 2346.0, "completions/mean_length": 796.2875366210938, "completions/mean_terminated_length": 796.2875366210938, "completions/min_length": 354.0, "completions/min_terminated_length": 354.0, "epoch": 0.037238169123351435, "grad_norm": 0.513206958770752, "kl": 0.0533447265625, "learning_rate": 4.999486602704451e-06, "loss": 0.0051, "num_tokens": 8500033.0, "reward": 0.17000000178813934, "reward_std": 0.08945737034082413, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.17208696901798248, "step": 96 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2801.0, "completions/max_terminated_length": 2801.0, "completions/mean_length": 794.5625, "completions/mean_terminated_length": 794.5625, "completions/min_length": 295.0, "completions/min_terminated_length": 295.0, "epoch": 0.03762606671838634, "grad_norm": 4.892813682556152, "kl": 0.18896484375, "learning_rate": 4.9994244292123905e-06, "loss": 0.026, "num_tokens": 8610678.0, "reward": 0.257500022649765, "reward_std": 0.11637254059314728, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.2437964677810669, "step": 97 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2210.0, "completions/max_terminated_length": 2210.0, "completions/mean_length": 640.9375, "completions/mean_terminated_length": 640.9375, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "epoch": 0.03801396431342126, "grad_norm": 0.6314482688903809, "kl": 0.0789794921875, "learning_rate": 4.999358703572119e-06, "loss": 0.0286, "num_tokens": 8706185.0, "reward": 0.3449999988079071, "reward_std": 0.15989021956920624, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24500000476837158, "rewards/unicoder_reward_fn/std": 0.3697073757648468, "step": 98 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1683.0, "completions/max_terminated_length": 1683.0, "completions/mean_length": 720.4000244140625, "completions/mean_terminated_length": 720.4000244140625, "completions/min_length": 375.0, "completions/min_terminated_length": 375.0, "epoch": 0.03840186190845617, "grad_norm": 0.509837806224823, "kl": 0.07275390625, "learning_rate": 4.999289425887425e-06, "loss": 0.0247, "num_tokens": 8811129.0, "reward": 0.22874999046325684, "reward_std": 0.160537987947464, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.2626062035560608, "step": 99 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1713.0, "completions/mean_length": 673.875, "completions/mean_terminated_length": 630.5570068359375, "completions/min_length": 350.0, "completions/min_terminated_length": 350.0, "epoch": 0.038789759503491075, "grad_norm": 0.7430384159088135, "kl": 0.0828857421875, "learning_rate": 4.999216596267708e-06, "loss": 0.0531, "num_tokens": 8906863.0, "reward": 0.27625003457069397, "reward_std": 0.15935559570789337, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.17749999463558197, "rewards/unicoder_reward_fn/std": 0.30020034313201904, "step": 100 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1907.0, "completions/max_terminated_length": 1907.0, "completions/mean_length": 733.3250122070312, "completions/mean_terminated_length": 733.3250122070312, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "epoch": 0.03917765709852599, "grad_norm": 0.6189298033714294, "kl": 0.07421875, "learning_rate": 4.999140214827976e-06, "loss": -0.0079, "num_tokens": 9015801.0, "reward": 0.21500001847743988, "reward_std": 0.15202419459819794, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.21745072305202484, "step": 101 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2016.0, "completions/max_terminated_length": 2016.0, "completions/mean_length": 767.7750244140625, "completions/mean_terminated_length": 767.7750244140625, "completions/min_length": 348.0, "completions/min_terminated_length": 348.0, "epoch": 0.0395655546935609, "grad_norm": 4.246227741241455, "kl": 0.533203125, "learning_rate": 4.999060281688845e-06, "loss": 0.0144, "num_tokens": 9134023.0, "reward": 0.14874999225139618, "reward_std": 0.08554533869028091, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.13687285780906677, "step": 102 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2455.0, "completions/max_terminated_length": 2455.0, "completions/mean_length": 674.9625244140625, "completions/mean_terminated_length": 674.9625244140625, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "epoch": 0.03995345228859581, "grad_norm": 0.6863650679588318, "kl": 0.0821533203125, "learning_rate": 4.99897679697654e-06, "loss": -0.0171, "num_tokens": 9234076.0, "reward": 0.23250000178813934, "reward_std": 0.11810418218374252, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.2561151087284088, "step": 103 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1791.0, "completions/max_terminated_length": 1791.0, "completions/mean_length": 747.4500122070312, "completions/mean_terminated_length": 747.4500122070312, "completions/min_length": 356.0, "completions/min_terminated_length": 356.0, "epoch": 0.04034134988363072, "grad_norm": 0.5646711587905884, "kl": 0.0687255859375, "learning_rate": 4.9988897608228966e-06, "loss": 0.0419, "num_tokens": 9343840.0, "reward": 0.24249999225139618, "reward_std": 0.1663830727338791, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2584765553474426, "step": 104 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2728.0, "completions/max_terminated_length": 2728.0, "completions/mean_length": 766.8875122070312, "completions/mean_terminated_length": 766.8875122070312, "completions/min_length": 337.0, "completions/min_terminated_length": 337.0, "epoch": 0.04072924747866563, "grad_norm": 0.5746595859527588, "kl": 0.0772705078125, "learning_rate": 4.998799173365355e-06, "loss": -0.0098, "num_tokens": 9457679.0, "reward": 0.2837500274181366, "reward_std": 0.15318447351455688, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.1850000023841858, "rewards/unicoder_reward_fn/std": 0.35260334610939026, "step": 105 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1588.0, "completions/max_terminated_length": 1588.0, "completions/mean_length": 737.6749877929688, "completions/mean_terminated_length": 737.6749877929688, "completions/min_length": 331.0, "completions/min_terminated_length": 331.0, "epoch": 0.041117145073700546, "grad_norm": 0.6349690556526184, "kl": 0.075927734375, "learning_rate": 4.998705034746965e-06, "loss": 0.0063, "num_tokens": 9567173.0, "reward": 0.1837500035762787, "reward_std": 0.12005230039358139, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.15186604857444763, "step": 106 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3178.0, "completions/max_terminated_length": 3178.0, "completions/mean_length": 787.1625366210938, "completions/mean_terminated_length": 787.1625366210938, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.041505042668735455, "grad_norm": 1.401928424835205, "kl": 0.12939453125, "learning_rate": 4.998607345116386e-06, "loss": 0.0788, "num_tokens": 9683634.0, "reward": 0.2912500202655792, "reward_std": 0.1435823291540146, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.19500000774860382, "rewards/unicoder_reward_fn/std": 0.31012454628944397, "step": 107 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1523.0, "completions/max_terminated_length": 1523.0, "completions/mean_length": 754.2374877929688, "completions/mean_terminated_length": 754.2374877929688, "completions/min_length": 283.0, "completions/min_terminated_length": 283.0, "epoch": 0.04189294026377036, "grad_norm": 0.5199242830276489, "kl": 0.076416015625, "learning_rate": 4.9985061046278824e-06, "loss": -0.0036, "num_tokens": 9790997.0, "reward": 0.20999999344348907, "reward_std": 0.12655074894428253, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.2153007984161377, "step": 108 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1210.0, "completions/max_terminated_length": 1210.0, "completions/mean_length": 650.7374877929688, "completions/mean_terminated_length": 650.7374877929688, "completions/min_length": 345.0, "completions/min_terminated_length": 345.0, "epoch": 0.04228083785880528, "grad_norm": 0.6833178400993347, "kl": 0.0777587890625, "learning_rate": 4.9984013134413265e-06, "loss": 0.0123, "num_tokens": 9888408.0, "reward": 0.2800000309944153, "reward_std": 0.17940165102481842, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.2867429554462433, "step": 109 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1413.0, "completions/max_terminated_length": 1413.0, "completions/mean_length": 688.3624877929688, "completions/mean_terminated_length": 688.3624877929688, "completions/min_length": 308.0, "completions/min_terminated_length": 308.0, "epoch": 0.04266873545384019, "grad_norm": 0.5465185642242432, "kl": 0.0888671875, "learning_rate": 4.998292971722197e-06, "loss": 0.0094, "num_tokens": 9988997.0, "reward": 0.14500001072883606, "reward_std": 0.08145972341299057, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.12825508415699005, "step": 110 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1716.0, "completions/max_terminated_length": 1716.0, "completions/mean_length": 596.8875122070312, "completions/mean_terminated_length": 596.8875122070312, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.043056633048875095, "grad_norm": 0.5824081897735596, "kl": 0.08740234375, "learning_rate": 4.998181079641583e-06, "loss": 0.0175, "num_tokens": 10084180.0, "reward": 0.21125002205371857, "reward_std": 0.13424620032310486, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.21546241641044617, "step": 111 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1608.0, "completions/max_terminated_length": 1608.0, "completions/mean_length": 648.4000244140625, "completions/mean_terminated_length": 648.4000244140625, "completions/min_length": 305.0, "completions/min_terminated_length": 305.0, "epoch": 0.04344453064391001, "grad_norm": 0.6946550607681274, "kl": 0.0731201171875, "learning_rate": 4.998065637376174e-06, "loss": 0.0241, "num_tokens": 10180540.0, "reward": 0.2800000309944153, "reward_std": 0.1808169186115265, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.33252963423728943, "step": 112 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1107.0, "completions/max_terminated_length": 1107.0, "completions/mean_length": 523.3375244140625, "completions/mean_terminated_length": 523.3375244140625, "completions/min_length": 284.0, "completions/min_terminated_length": 284.0, "epoch": 0.04383242823894492, "grad_norm": 0.6432906985282898, "kl": 0.0977783203125, "learning_rate": 4.9979466451082726e-06, "loss": 0.0043, "num_tokens": 10271207.0, "reward": 0.36250001192092896, "reward_std": 0.1313905566930771, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.26749998331069946, "rewards/unicoder_reward_fn/std": 0.34705817699432373, "step": 113 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1072.0, "completions/max_terminated_length": 1072.0, "completions/mean_length": 501.0874938964844, "completions/mean_terminated_length": 501.0874938964844, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "epoch": 0.04422032583397983, "grad_norm": 0.7505579590797424, "kl": 0.0943603515625, "learning_rate": 4.997824103025781e-06, "loss": 0.017, "num_tokens": 10354774.0, "reward": 0.26750001311302185, "reward_std": 0.18299490213394165, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.2812269628047943, "step": 114 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1053.0, "completions/max_terminated_length": 1053.0, "completions/mean_length": 452.0874938964844, "completions/mean_terminated_length": 452.0874938964844, "completions/min_length": 241.0, "completions/min_terminated_length": 241.0, "epoch": 0.04460822342901474, "grad_norm": 0.7407625913619995, "kl": 0.094482421875, "learning_rate": 4.997698011322211e-06, "loss": -0.0015, "num_tokens": 10441557.0, "reward": 0.26125001907348633, "reward_std": 0.12256067246198654, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.2521301507949829, "step": 115 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1231.0, "completions/max_terminated_length": 1231.0, "completions/mean_length": 469.2749938964844, "completions/mean_terminated_length": 469.2749938964844, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.04499612102404965, "grad_norm": 0.637856662273407, "kl": 0.093017578125, "learning_rate": 4.997568370196679e-06, "loss": 0.0093, "num_tokens": 10529859.0, "reward": 0.32625001668930054, "reward_std": 0.1353728026151657, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.22750000655651093, "rewards/unicoder_reward_fn/std": 0.3256833255290985, "step": 116 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 804.0, "completions/max_terminated_length": 804.0, "completions/mean_length": 467.2875061035156, "completions/mean_terminated_length": 467.2875061035156, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.04538401861908456, "grad_norm": 0.7106099724769592, "kl": 0.088134765625, "learning_rate": 4.9974351798539045e-06, "loss": -0.002, "num_tokens": 10615586.0, "reward": 0.24375000596046448, "reward_std": 0.1897091120481491, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.14499999582767487, "rewards/unicoder_reward_fn/std": 0.2791941463947296, "step": 117 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 916.0, "completions/max_terminated_length": 916.0, "completions/mean_length": 477.6499938964844, "completions/mean_terminated_length": 477.6499938964844, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.045771916214119475, "grad_norm": 0.7194799780845642, "kl": 0.0858154296875, "learning_rate": 4.997298440504215e-06, "loss": 0.019, "num_tokens": 10706670.0, "reward": 0.3150000274181366, "reward_std": 0.1582345962524414, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2150000035762787, "rewards/unicoder_reward_fn/std": 0.3130495250225067, "step": 118 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1145.0, "completions/max_terminated_length": 1145.0, "completions/mean_length": 443.375, "completions/mean_terminated_length": 443.375, "completions/min_length": 241.0, "completions/min_terminated_length": 241.0, "epoch": 0.04615981380915438, "grad_norm": 0.7510746717453003, "kl": 0.0914306640625, "learning_rate": 4.99715815236354e-06, "loss": 0.0413, "num_tokens": 10800404.0, "reward": 0.24249999225139618, "reward_std": 0.12139517068862915, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.27594488859176636, "step": 119 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 775.0, "completions/max_terminated_length": 775.0, "completions/mean_length": 377.4125061035156, "completions/mean_terminated_length": 377.4125061035156, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "epoch": 0.04654771140418929, "grad_norm": 0.9184646606445312, "kl": 0.1016845703125, "learning_rate": 4.997014315653413e-06, "loss": 0.0185, "num_tokens": 10872077.0, "reward": 0.45250001549720764, "reward_std": 0.18905103206634521, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3525000214576721, "rewards/unicoder_reward_fn/std": 0.34309646487236023, "step": 120 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1619.0, "completions/max_terminated_length": 1619.0, "completions/mean_length": 431.0375061035156, "completions/mean_terminated_length": 431.0375061035156, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.04693560899922421, "grad_norm": 0.6000313758850098, "kl": 0.088134765625, "learning_rate": 4.996866930600973e-06, "loss": 0.0178, "num_tokens": 10964544.0, "reward": 0.19750002026557922, "reward_std": 0.09892836958169937, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.20374655723571777, "step": 121 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 695.0, "completions/max_terminated_length": 695.0, "completions/mean_length": 356.6499938964844, "completions/mean_terminated_length": 356.6499938964844, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.047323506594259115, "grad_norm": 0.7749669551849365, "kl": 0.1007080078125, "learning_rate": 4.99671599743896e-06, "loss": -0.0075, "num_tokens": 11041452.0, "reward": 0.32875001430511475, "reward_std": 0.15132203698158264, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.23000001907348633, "rewards/unicoder_reward_fn/std": 0.3820000886917114, "step": 122 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 824.0, "completions/max_terminated_length": 824.0, "completions/mean_length": 408.76251220703125, "completions/mean_terminated_length": 408.76251220703125, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.047711404189294024, "grad_norm": 0.5975648760795593, "kl": 0.086181640625, "learning_rate": 4.996561516405718e-06, "loss": 0.0071, "num_tokens": 11119993.0, "reward": 0.26750001311302185, "reward_std": 0.07181212306022644, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.30761435627937317, "step": 123 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 741.0, "completions/max_terminated_length": 741.0, "completions/mean_length": 337.5, "completions/mean_terminated_length": 337.5, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.04809930178432894, "grad_norm": 0.7817389369010925, "kl": 0.1060791015625, "learning_rate": 4.996403487745195e-06, "loss": 0.018, "num_tokens": 11196793.0, "reward": 0.30000001192092896, "reward_std": 0.21806326508522034, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.32452499866485596, "step": 124 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 957.0, "completions/max_terminated_length": 957.0, "completions/mean_length": 332.8625183105469, "completions/mean_terminated_length": 332.8625183105469, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.04848719937936385, "grad_norm": 0.7606878280639648, "kl": 0.095458984375, "learning_rate": 4.996241911706938e-06, "loss": -0.0071, "num_tokens": 11269446.0, "reward": 0.42500001192092896, "reward_std": 0.09999766200780869, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.32499998807907104, "rewards/unicoder_reward_fn/std": 0.41169610619544983, "step": 125 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 575.0, "completions/max_terminated_length": 575.0, "completions/mean_length": 314.45001220703125, "completions/mean_terminated_length": 314.45001220703125, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.048875096974398756, "grad_norm": 0.83009934425354, "kl": 0.126220703125, "learning_rate": 4.996076788546099e-06, "loss": 0.0023, "num_tokens": 11343242.0, "reward": 0.28125, "reward_std": 0.07942771166563034, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.3216944932937622, "step": 126 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 697.0, "completions/max_terminated_length": 697.0, "completions/mean_length": 381.3500061035156, "completions/mean_terminated_length": 381.3500061035156, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.04926299456943367, "grad_norm": 0.717721164226532, "kl": 0.091796875, "learning_rate": 4.995908118523429e-06, "loss": -0.0007, "num_tokens": 11426198.0, "reward": 0.19249999523162842, "reward_std": 0.10295168310403824, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.18811187148094177, "step": 127 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 799.0, "completions/max_terminated_length": 799.0, "completions/mean_length": 379.2250061035156, "completions/mean_terminated_length": 379.2250061035156, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.04965089216446858, "grad_norm": 0.9794763922691345, "kl": 0.09716796875, "learning_rate": 4.995735901905282e-06, "loss": 0.0304, "num_tokens": 11506856.0, "reward": 0.24250002205371857, "reward_std": 0.14830456674098969, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.19666524231433868, "step": 128 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 645.0, "completions/max_terminated_length": 645.0, "completions/mean_length": 321.01251220703125, "completions/mean_terminated_length": 321.01251220703125, "completions/min_length": 152.0, "completions/min_terminated_length": 152.0, "epoch": 0.05003878975950349, "grad_norm": 0.7199133038520813, "kl": 0.1014404296875, "learning_rate": 4.995560138963612e-06, "loss": 0.0155, "num_tokens": 11578865.0, "reward": 0.18250000476837158, "reward_std": 0.10112505406141281, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.18471668660640717, "step": 129 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 796.0, "completions/max_terminated_length": 796.0, "completions/mean_length": 326.0500183105469, "completions/mean_terminated_length": 326.0500183105469, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.0504266873545384, "grad_norm": 0.8115301132202148, "kl": 0.10107421875, "learning_rate": 4.9953808299759694e-06, "loss": -0.0108, "num_tokens": 11646813.0, "reward": 0.29749998450279236, "reward_std": 0.12670481204986572, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19750000536441803, "rewards/unicoder_reward_fn/std": 0.29251205921173096, "step": 130 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 566.0, "completions/max_terminated_length": 566.0, "completions/mean_length": 306.0874938964844, "completions/mean_terminated_length": 306.0874938964844, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.05081458494957331, "grad_norm": 0.9443567395210266, "kl": 0.1033935546875, "learning_rate": 4.995197975225512e-06, "loss": -0.0317, "num_tokens": 11709180.0, "reward": 0.4100000560283661, "reward_std": 0.1868554651737213, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3100000023841858, "rewards/unicoder_reward_fn/std": 0.3843785226345062, "step": 131 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 836.0, "completions/max_terminated_length": 836.0, "completions/mean_length": 332.8125, "completions/mean_terminated_length": 332.8125, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.05120248254460822, "grad_norm": 0.8449292778968811, "kl": 0.111328125, "learning_rate": 4.99501157500099e-06, "loss": 0.011, "num_tokens": 11783493.0, "reward": 0.3124999701976776, "reward_std": 0.12641015648841858, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.21250002086162567, "rewards/unicoder_reward_fn/std": 0.3357873857021332, "step": 132 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1471.0, "completions/max_terminated_length": 1471.0, "completions/mean_length": 455.2749938964844, "completions/mean_terminated_length": 455.2749938964844, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "epoch": 0.051590380139643136, "grad_norm": 0.6684557795524597, "kl": 0.1031494140625, "learning_rate": 4.994821629596755e-06, "loss": 0.0022, "num_tokens": 11875571.0, "reward": 0.19500000774860382, "reward_std": 0.09156139940023422, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0949999988079071, "rewards/unicoder_reward_fn/std": 0.18548600375652313, "step": 133 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 804.0, "completions/max_terminated_length": 804.0, "completions/mean_length": 420.4375, "completions/mean_terminated_length": 420.4375, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.051978277734678044, "grad_norm": 0.5835564732551575, "kl": 0.0892333984375, "learning_rate": 4.994628139312757e-06, "loss": -0.0081, "num_tokens": 11964590.0, "reward": 0.2200000137090683, "reward_std": 0.12413938343524933, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.25576889514923096, "step": 134 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1007.0, "completions/max_terminated_length": 1007.0, "completions/mean_length": 341.8000183105469, "completions/mean_terminated_length": 341.8000183105469, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.05236617532971296, "grad_norm": 0.9242949485778809, "kl": 0.099853515625, "learning_rate": 4.9944311044545435e-06, "loss": 0.0048, "num_tokens": 12040294.0, "reward": 0.26875001192092896, "reward_std": 0.15461403131484985, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.2466997355222702, "step": 135 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 930.0, "completions/max_terminated_length": 930.0, "completions/mean_length": 342.45001220703125, "completions/mean_terminated_length": 342.45001220703125, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.05275407292474787, "grad_norm": 0.9130709767341614, "kl": 0.1195068359375, "learning_rate": 4.994230525333259e-06, "loss": -0.0267, "num_tokens": 12120346.0, "reward": 0.19249999523162842, "reward_std": 0.09376432746648788, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09250000864267349, "rewards/unicoder_reward_fn/std": 0.15893235802650452, "step": 136 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 643.0, "completions/max_terminated_length": 643.0, "completions/mean_length": 324.4125061035156, "completions/mean_terminated_length": 324.4125061035156, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.053141970519782776, "grad_norm": 0.9632866978645325, "kl": 0.1007080078125, "learning_rate": 4.994026402265646e-06, "loss": -0.014, "num_tokens": 12194115.0, "reward": 0.25, "reward_std": 0.14740632474422455, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.27466896176338196, "step": 137 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 710.0, "completions/max_terminated_length": 710.0, "completions/mean_length": 323.01251220703125, "completions/mean_terminated_length": 323.01251220703125, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.05352986811481769, "grad_norm": 1.055999517440796, "kl": 0.1243896484375, "learning_rate": 4.993818735574042e-06, "loss": 0.0006, "num_tokens": 12263476.0, "reward": 0.2874999940395355, "reward_std": 0.1940165013074875, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.255261093378067, "step": 138 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 654.0, "completions/max_terminated_length": 654.0, "completions/mean_length": 317.3625183105469, "completions/mean_terminated_length": 317.3625183105469, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.0539177657098526, "grad_norm": 0.8405289649963379, "kl": 0.1185302734375, "learning_rate": 4.9936075255863814e-06, "loss": -0.0021, "num_tokens": 12331209.0, "reward": 0.29250001907348633, "reward_std": 0.13439567387104034, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.28004294633865356, "step": 139 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 974.0, "completions/max_terminated_length": 974.0, "completions/mean_length": 374.4624938964844, "completions/mean_terminated_length": 374.4624938964844, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.05430566330488751, "grad_norm": 0.7320149540901184, "kl": 0.107177734375, "learning_rate": 4.9933927726361945e-06, "loss": 0.0145, "num_tokens": 12408502.0, "reward": 0.29500001668930054, "reward_std": 0.12206808477640152, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19499999284744263, "rewards/unicoder_reward_fn/std": 0.31012454628944397, "step": 140 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1167.0, "completions/max_terminated_length": 1167.0, "completions/mean_length": 398.3374938964844, "completions/mean_terminated_length": 398.3374938964844, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.054693560899922423, "grad_norm": 0.8347128629684448, "kl": 0.1129150390625, "learning_rate": 4.993174477062604e-06, "loss": -0.0237, "num_tokens": 12489497.0, "reward": 0.14750000834465027, "reward_std": 0.0807124599814415, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.11580756306648254, "step": 141 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 980.0, "completions/max_terminated_length": 980.0, "completions/mean_length": 439.51251220703125, "completions/mean_terminated_length": 439.51251220703125, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.05508145849495733, "grad_norm": 0.6793642640113831, "kl": 0.104736328125, "learning_rate": 4.992952639210329e-06, "loss": 0.004, "num_tokens": 12569634.0, "reward": 0.1287499964237213, "reward_std": 0.05157065391540527, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0325000025331974, "rewards/unicoder_reward_fn/std": 0.07424727827310562, "step": 142 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 713.0, "completions/mean_length": 394.8125, "completions/mean_terminated_length": 347.9620361328125, "completions/min_length": 152.0, "completions/min_terminated_length": 152.0, "epoch": 0.05546935608999224, "grad_norm": 0.9312738180160522, "kl": 0.102783203125, "learning_rate": 4.9927272594296816e-06, "loss": 0.0538, "num_tokens": 12648835.0, "reward": 0.26625001430511475, "reward_std": 0.149911567568779, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.23045086860656738, "step": 143 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 574.0, "completions/max_terminated_length": 574.0, "completions/mean_length": 314.2875061035156, "completions/mean_terminated_length": 314.2875061035156, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.055857253685027156, "grad_norm": 0.7925715446472168, "kl": 0.1461181640625, "learning_rate": 4.992498338076567e-06, "loss": 0.0261, "num_tokens": 12723250.0, "reward": 0.2874999940395355, "reward_std": 0.0967419371008873, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.24923935532569885, "step": 144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 729.0, "completions/max_terminated_length": 729.0, "completions/mean_length": 310.0375061035156, "completions/mean_terminated_length": 310.0375061035156, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.056245151280062064, "grad_norm": 0.9252837300300598, "kl": 0.133056640625, "learning_rate": 4.992265875512485e-06, "loss": 0.0751, "num_tokens": 12803157.0, "reward": 0.30750003457069397, "reward_std": 0.21213355660438538, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.2958681881427765, "step": 145 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2189.0, "completions/mean_length": 495.0625, "completions/mean_terminated_length": 449.48101806640625, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.05663304887509697, "grad_norm": 0.46331092715263367, "kl": 0.10693359375, "learning_rate": 4.9920298721045245e-06, "loss": 0.1225, "num_tokens": 12887930.0, "reward": 0.17374999821186066, "reward_std": 0.09508509188890457, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.17824745178222656, "step": 146 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1888.0, "completions/max_terminated_length": 1888.0, "completions/mean_length": 382.7375183105469, "completions/mean_terminated_length": 382.7375183105469, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.05702094647013189, "grad_norm": 0.622003436088562, "kl": 0.1007080078125, "learning_rate": 4.991790328225366e-06, "loss": -0.0149, "num_tokens": 12966477.0, "reward": 0.33500000834465027, "reward_std": 0.06824255734682083, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23500001430511475, "rewards/unicoder_reward_fn/std": 0.3482688069343567, "step": 147 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1387.0, "completions/max_terminated_length": 1387.0, "completions/mean_length": 355.2124938964844, "completions/mean_terminated_length": 355.2124938964844, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.057408844065166796, "grad_norm": 1.001321792602539, "kl": 0.1212158203125, "learning_rate": 4.9915472442532834e-06, "loss": 0.0247, "num_tokens": 13044910.0, "reward": 0.43250003457069397, "reward_std": 0.18012695014476776, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.33249998092651367, "rewards/unicoder_reward_fn/std": 0.41606202721595764, "step": 148 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1543.0, "completions/max_terminated_length": 1543.0, "completions/mean_length": 384.9250183105469, "completions/mean_terminated_length": 384.9250183105469, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.057796741660201705, "grad_norm": 1.3920271396636963, "kl": 0.197509765625, "learning_rate": 4.991300620572139e-06, "loss": 0.0017, "num_tokens": 13128328.0, "reward": 0.16625000536441803, "reward_std": 0.04199902340769768, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.15242762863636017, "step": 149 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1139.0, "completions/max_terminated_length": 1139.0, "completions/mean_length": 422.1000061035156, "completions/mean_terminated_length": 422.1000061035156, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.05818463925523662, "grad_norm": 0.6105234622955322, "kl": 0.1162109375, "learning_rate": 4.991050457571384e-06, "loss": 0.0241, "num_tokens": 13209136.0, "reward": 0.23250003159046173, "reward_std": 0.1010664626955986, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.2564484775066376, "step": 150 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 849.0, "completions/max_terminated_length": 849.0, "completions/mean_length": 355.2749938964844, "completions/mean_terminated_length": 355.2749938964844, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.05857253685027153, "grad_norm": 0.5300409197807312, "kl": 0.1317138671875, "learning_rate": 4.990796755646063e-06, "loss": 0.0251, "num_tokens": 13283134.0, "reward": 0.18250000476837158, "reward_std": 0.08043359965085983, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.19794835150241852, "step": 151 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 639.0, "completions/max_terminated_length": 639.0, "completions/mean_length": 345.9750061035156, "completions/mean_terminated_length": 345.9750061035156, "completions/min_length": 152.0, "completions/min_terminated_length": 152.0, "epoch": 0.05896043444530644, "grad_norm": 0.5914447903633118, "kl": 0.149169921875, "learning_rate": 4.990539515196804e-06, "loss": 0.0126, "num_tokens": 13359852.0, "reward": 0.18000000715255737, "reward_std": 0.09504849463701248, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.1671806424856186, "step": 152 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 738.0, "completions/max_terminated_length": 738.0, "completions/mean_length": 324.5375061035156, "completions/mean_terminated_length": 324.5375061035156, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.05934833204034135, "grad_norm": 0.6847733855247498, "kl": 0.14599609375, "learning_rate": 4.990278736629824e-06, "loss": -0.0132, "num_tokens": 13424007.0, "reward": 0.26750001311302185, "reward_std": 0.12172949314117432, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.2782290279865265, "step": 153 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3067.0, "completions/max_terminated_length": 3067.0, "completions/mean_length": 393.6750183105469, "completions/mean_terminated_length": 393.6750183105469, "completions/min_length": 75.0, "completions/min_terminated_length": 75.0, "epoch": 0.05973622963537626, "grad_norm": 0.7113120555877686, "kl": 0.1475830078125, "learning_rate": 4.99001442035693e-06, "loss": 0.0327, "num_tokens": 13506965.0, "reward": 0.2837500274181366, "reward_std": 0.1636272817850113, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.18500001728534698, "rewards/unicoder_reward_fn/std": 0.3130495250225067, "step": 154 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1330.0, "completions/max_terminated_length": 1330.0, "completions/mean_length": 410.625, "completions/mean_terminated_length": 410.625, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.06012412723041117, "grad_norm": 0.6558129787445068, "kl": 0.1263427734375, "learning_rate": 4.989746566795513e-06, "loss": 0.0402, "num_tokens": 13592023.0, "reward": 0.2474999874830246, "reward_std": 0.1615317165851593, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.25406816601753235, "step": 155 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 809.0, "completions/max_terminated_length": 809.0, "completions/mean_length": 353.8125, "completions/mean_terminated_length": 353.8125, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.060512024825446084, "grad_norm": 0.7079071402549744, "kl": 0.124267578125, "learning_rate": 4.989475176368548e-06, "loss": 0.0248, "num_tokens": 13664552.0, "reward": 0.35500001907348633, "reward_std": 0.16502873599529266, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2550000250339508, "rewards/unicoder_reward_fn/std": 0.2916127145290375, "step": 156 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 786.0, "completions/max_terminated_length": 786.0, "completions/mean_length": 261.76251220703125, "completions/mean_terminated_length": 261.76251220703125, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.06089992242048099, "grad_norm": 0.9276941418647766, "kl": 0.170166015625, "learning_rate": 4.9892002495046e-06, "loss": -0.0128, "num_tokens": 13725757.0, "reward": 0.26750001311302185, "reward_std": 0.14398124814033508, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.26897603273391724, "step": 157 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 722.0, "completions/max_terminated_length": 722.0, "completions/mean_length": 334.5249938964844, "completions/mean_terminated_length": 334.5249938964844, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "epoch": 0.0612878200155159, "grad_norm": 0.7318286299705505, "kl": 0.147216796875, "learning_rate": 4.988921786637816e-06, "loss": 0.0173, "num_tokens": 13797159.0, "reward": 0.19875000417232513, "reward_std": 0.13438506424427032, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.21108520030975342, "step": 158 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2465.0, "completions/max_terminated_length": 2465.0, "completions/mean_length": 303.4624938964844, "completions/mean_terminated_length": 303.4624938964844, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.061675717610550816, "grad_norm": 0.7510900497436523, "kl": 0.1435546875, "learning_rate": 4.988639788207926e-06, "loss": 0.0268, "num_tokens": 13869804.0, "reward": 0.25, "reward_std": 0.17024655640125275, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.2709570527076721, "step": 159 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 683.0, "completions/max_terminated_length": 683.0, "completions/mean_length": 322.0249938964844, "completions/mean_terminated_length": 322.0249938964844, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.062063615205585725, "grad_norm": 0.9066020846366882, "kl": 0.1904296875, "learning_rate": 4.988354254660243e-06, "loss": 0.0205, "num_tokens": 13936310.0, "reward": 0.2800000309944153, "reward_std": 0.1786544770002365, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.2517784833908081, "step": 160 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 963.0, "completions/max_terminated_length": 963.0, "completions/mean_length": 348.45001220703125, "completions/mean_terminated_length": 348.45001220703125, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.06245151280062063, "grad_norm": 0.7214123606681824, "kl": 0.17529296875, "learning_rate": 4.988065186445666e-06, "loss": 0.0051, "num_tokens": 14015842.0, "reward": 0.18250000476837158, "reward_std": 0.09264849871397018, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.14821724593639374, "step": 161 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4062.0, "completions/max_terminated_length": 4062.0, "completions/mean_length": 451.26251220703125, "completions/mean_terminated_length": 451.26251220703125, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.06283941039565555, "grad_norm": 0.7397768497467041, "kl": 0.155517578125, "learning_rate": 4.987772584020671e-06, "loss": 0.0186, "num_tokens": 14093271.0, "reward": 0.17249999940395355, "reward_std": 0.08064081519842148, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.1566864252090454, "step": 162 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2817.0, "completions/mean_length": 532.9625244140625, "completions/mean_terminated_length": 441.6025695800781, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.06322730799069046, "grad_norm": 0.8928542733192444, "kl": 0.1383056640625, "learning_rate": 4.987476447847319e-06, "loss": -0.0037, "num_tokens": 14188068.0, "reward": 0.29500001668930054, "reward_std": 0.13800378143787384, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19500000774860382, "rewards/unicoder_reward_fn/std": 0.29506489634513855, "step": 163 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3836.0, "completions/mean_length": 664.3500366210938, "completions/mean_terminated_length": 620.9114379882812, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.06361520558572537, "grad_norm": 0.6007050275802612, "kl": 0.1131591796875, "learning_rate": 4.9871767783932475e-06, "loss": 0.0058, "num_tokens": 14289352.0, "reward": 0.2449999898672104, "reward_std": 0.10146994888782501, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.28100183606147766, "step": 164 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3859.0, "completions/mean_length": 806.9375, "completions/mean_terminated_length": 722.6026000976562, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.06400310318076027, "grad_norm": 0.7033397555351257, "kl": 0.1396484375, "learning_rate": 4.9868735761316775e-06, "loss": 0.0435, "num_tokens": 14401291.0, "reward": 0.30000004172325134, "reward_std": 0.1440671682357788, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.28902938961982727, "step": 165 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2455.0, "completions/mean_length": 523.0, "completions/mean_terminated_length": 477.77215576171875, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.0643910007757952, "grad_norm": 0.615270733833313, "kl": 0.126220703125, "learning_rate": 4.986566841541405e-06, "loss": -0.0637, "num_tokens": 14493811.0, "reward": 0.29875001311302185, "reward_std": 0.19343259930610657, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.28816312551498413, "step": 166 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1449.0, "completions/mean_length": 443.3500061035156, "completions/mean_terminated_length": 397.1139221191406, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.0647788983708301, "grad_norm": 0.6911518573760986, "kl": 0.1348876953125, "learning_rate": 4.986256575106806e-06, "loss": 0.0791, "num_tokens": 14583935.0, "reward": 0.25, "reward_std": 0.15772591531276703, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.25950294733047485, "step": 167 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1289.0, "completions/max_terminated_length": 1289.0, "completions/mean_length": 352.5, "completions/mean_terminated_length": 352.5, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.06516679596586501, "grad_norm": 0.9088248610496521, "kl": 0.15234375, "learning_rate": 4.985942777317834e-06, "loss": 0.0211, "num_tokens": 14651511.0, "reward": 0.320000022649765, "reward_std": 0.1770070344209671, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.2224999964237213, "rewards/unicoder_reward_fn/std": 0.2899039387702942, "step": 168 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1570.0, "completions/max_terminated_length": 1570.0, "completions/mean_length": 475.45001220703125, "completions/mean_terminated_length": 475.45001220703125, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.06555469356089992, "grad_norm": 0.7465201020240784, "kl": 0.1181640625, "learning_rate": 4.985625448670019e-06, "loss": 0.0524, "num_tokens": 14741163.0, "reward": 0.23375001549720764, "reward_std": 0.13580851256847382, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.23117750883102417, "step": 169 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1877.0, "completions/max_terminated_length": 1877.0, "completions/mean_length": 369.0874938964844, "completions/mean_terminated_length": 369.0874938964844, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.06594259115593483, "grad_norm": 0.7619454264640808, "kl": 0.1298828125, "learning_rate": 4.985304589664466e-06, "loss": 0.0271, "num_tokens": 14818146.0, "reward": 0.20624999701976776, "reward_std": 0.12928104400634766, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.22713126242160797, "step": 170 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 865.0, "completions/max_terminated_length": 865.0, "completions/mean_length": 387.9750061035156, "completions/mean_terminated_length": 387.9750061035156, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.06633048875096974, "grad_norm": 0.6737417578697205, "kl": 0.1416015625, "learning_rate": 4.984980200807855e-06, "loss": 0.0018, "num_tokens": 14899984.0, "reward": 0.32874998450279236, "reward_std": 0.08800918608903885, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.22999998927116394, "rewards/unicoder_reward_fn/std": 0.3231177031993866, "step": 171 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1885.0, "completions/max_terminated_length": 1885.0, "completions/mean_length": 404.1625061035156, "completions/mean_terminated_length": 404.1625061035156, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.06671838634600466, "grad_norm": 0.781979501247406, "kl": 0.12646484375, "learning_rate": 4.98465228261244e-06, "loss": 0.0265, "num_tokens": 14973477.0, "reward": 0.2549999952316284, "reward_std": 0.2064262479543686, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.28408190608024597, "step": 172 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 943.0, "completions/max_terminated_length": 943.0, "completions/mean_length": 402.25, "completions/mean_terminated_length": 402.25, "completions/min_length": 241.0, "completions/min_terminated_length": 241.0, "epoch": 0.06710628394103957, "grad_norm": 0.6136270761489868, "kl": 0.140380859375, "learning_rate": 4.984320835596048e-06, "loss": 0.0114, "num_tokens": 15060841.0, "reward": 0.2837500274181366, "reward_std": 0.11466412991285324, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.1850000023841858, "rewards/unicoder_reward_fn/std": 0.27147045731544495, "step": 173 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1108.0, "completions/max_terminated_length": 1108.0, "completions/mean_length": 459.0375061035156, "completions/mean_terminated_length": 459.0375061035156, "completions/min_length": 217.0, "completions/min_terminated_length": 217.0, "epoch": 0.06749418153607448, "grad_norm": 0.6078804731369019, "kl": 0.1202392578125, "learning_rate": 4.9839858602820815e-06, "loss": -0.0007, "num_tokens": 15145604.0, "reward": 0.20750001072883606, "reward_std": 0.09437689185142517, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.1852811872959137, "step": 174 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1065.0, "completions/max_terminated_length": 1065.0, "completions/mean_length": 359.7124938964844, "completions/mean_terminated_length": 359.7124938964844, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.06788207913110939, "grad_norm": 0.6904677152633667, "kl": 0.140380859375, "learning_rate": 4.98364735719951e-06, "loss": -0.0304, "num_tokens": 15218053.0, "reward": 0.27000004053115845, "reward_std": 0.09850352257490158, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.2900894582271576, "step": 175 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 799.0, "completions/max_terminated_length": 799.0, "completions/mean_length": 375.0249938964844, "completions/mean_terminated_length": 375.0249938964844, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.0682699767261443, "grad_norm": 0.8315485119819641, "kl": 0.1357421875, "learning_rate": 4.983305326882875e-06, "loss": 0.0323, "num_tokens": 15299727.0, "reward": 0.3400000035762787, "reward_std": 0.15266548097133636, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.24250002205371857, "rewards/unicoder_reward_fn/std": 0.33329007029533386, "step": 176 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1254.0, "completions/max_terminated_length": 1254.0, "completions/mean_length": 505.6875, "completions/mean_terminated_length": 505.6875, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.0686578743211792, "grad_norm": 0.6327449083328247, "kl": 0.132568359375, "learning_rate": 4.982959769872294e-06, "loss": 0.0832, "num_tokens": 15398062.0, "reward": 0.20750001072883606, "reward_std": 0.09307727217674255, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.16504698991775513, "step": 177 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 895.0, "completions/max_terminated_length": 895.0, "completions/mean_length": 412.0375061035156, "completions/mean_terminated_length": 412.0375061035156, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.06904577191621412, "grad_norm": 0.7468381524085999, "kl": 0.1282958984375, "learning_rate": 4.9826106867134435e-06, "loss": 0.0333, "num_tokens": 15484313.0, "reward": 0.19750002026557922, "reward_std": 0.09191574156284332, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.14923858642578125, "step": 178 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 793.0, "completions/max_terminated_length": 793.0, "completions/mean_length": 344.82501220703125, "completions/mean_terminated_length": 344.82501220703125, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.06943366951124903, "grad_norm": 0.9415863752365112, "kl": 0.134765625, "learning_rate": 4.982258077957576e-06, "loss": -0.0258, "num_tokens": 15552755.0, "reward": 0.3725000321865082, "reward_std": 0.19005921483039856, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.27250000834465027, "rewards/unicoder_reward_fn/std": 0.2989327013492584, "step": 179 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1433.0, "completions/max_terminated_length": 1433.0, "completions/mean_length": 356.6875, "completions/mean_terminated_length": 356.6875, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.06982156710628394, "grad_norm": 0.7579959630966187, "kl": 0.154296875, "learning_rate": 4.981901944161507e-06, "loss": 0.0128, "num_tokens": 15631146.0, "reward": 0.2549999952316284, "reward_std": 0.07076601684093475, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.24848909676074982, "step": 180 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1917.0, "completions/max_terminated_length": 1917.0, "completions/mean_length": 331.1625061035156, "completions/mean_terminated_length": 331.1625061035156, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.07020946470131885, "grad_norm": 0.695427656173706, "kl": 0.1669921875, "learning_rate": 4.981542285887623e-06, "loss": -0.0016, "num_tokens": 15708815.0, "reward": 0.35500001907348633, "reward_std": 0.14948557317256927, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2549999952316284, "rewards/unicoder_reward_fn/std": 0.3500090539455414, "step": 181 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 837.0, "completions/max_terminated_length": 837.0, "completions/mean_length": 342.9624938964844, "completions/mean_terminated_length": 342.9624938964844, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.07059736229635376, "grad_norm": 0.8505650758743286, "kl": 0.1474609375, "learning_rate": 4.981179103703872e-06, "loss": -0.0002, "num_tokens": 15780164.0, "reward": 0.30250000953674316, "reward_std": 0.14486202597618103, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.32993096113204956, "step": 182 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1608.0, "completions/max_terminated_length": 1608.0, "completions/mean_length": 415.1125183105469, "completions/mean_terminated_length": 415.1125183105469, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.07098525989138868, "grad_norm": 0.5601588487625122, "kl": 0.154541015625, "learning_rate": 4.980812398183768e-06, "loss": -0.0279, "num_tokens": 15863317.0, "reward": 0.32875004410743713, "reward_std": 0.09785236418247223, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.22999998927116394, "rewards/unicoder_reward_fn/std": 0.33086004853248596, "step": 183 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 732.0, "completions/max_terminated_length": 732.0, "completions/mean_length": 348.0874938964844, "completions/mean_terminated_length": 348.0874938964844, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.07137315748642359, "grad_norm": 0.9052456021308899, "kl": 0.15234375, "learning_rate": 4.98044216990639e-06, "loss": 0.0037, "num_tokens": 15942532.0, "reward": 0.26500001549720764, "reward_std": 0.12740038335323334, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16499999165534973, "rewards/unicoder_reward_fn/std": 0.22675475478172302, "step": 184 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 736.0, "completions/max_terminated_length": 736.0, "completions/mean_length": 378.9375, "completions/mean_terminated_length": 378.9375, "completions/min_length": 159.0, "completions/min_terminated_length": 159.0, "epoch": 0.0717610550814585, "grad_norm": 0.7681582570075989, "kl": 0.154541015625, "learning_rate": 4.980068419456379e-06, "loss": 0.0378, "num_tokens": 16022047.0, "reward": 0.38750001788139343, "reward_std": 0.1610429286956787, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.3439807891845703, "step": 185 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 596.0, "completions/max_terminated_length": 596.0, "completions/mean_length": 351.2250061035156, "completions/mean_terminated_length": 351.2250061035156, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.0721489526764934, "grad_norm": 0.6600393652915955, "kl": 0.15625, "learning_rate": 4.979691147423936e-06, "loss": 0.0085, "num_tokens": 16092449.0, "reward": 0.4037500321865082, "reward_std": 0.09322072565555573, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.3050000071525574, "rewards/unicoder_reward_fn/std": 0.38447731733322144, "step": 186 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 696.0, "completions/max_terminated_length": 696.0, "completions/mean_length": 354.9250183105469, "completions/mean_terminated_length": 354.9250183105469, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.07253685027152831, "grad_norm": 0.7983233332633972, "kl": 0.179443359375, "learning_rate": 4.9793103544048275e-06, "loss": -0.0002, "num_tokens": 16172035.0, "reward": 0.2800000309944153, "reward_std": 0.14599613845348358, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.32855647802352905, "step": 187 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1333.0, "completions/max_terminated_length": 1333.0, "completions/mean_length": 476.88751220703125, "completions/mean_terminated_length": 476.88751220703125, "completions/min_length": 242.0, "completions/min_terminated_length": 242.0, "epoch": 0.07292474786656322, "grad_norm": 0.6302588582038879, "kl": 0.12890625, "learning_rate": 4.978926041000376e-06, "loss": -0.0096, "num_tokens": 16259330.0, "reward": 0.2875000238418579, "reward_std": 0.09889662265777588, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.253269761800766, "step": 188 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 897.0, "completions/max_terminated_length": 897.0, "completions/mean_length": 459.9375, "completions/mean_terminated_length": 459.9375, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "epoch": 0.07331264546159814, "grad_norm": 0.5369707345962524, "kl": 0.135986328125, "learning_rate": 4.978538207817466e-06, "loss": 0.0253, "num_tokens": 16350245.0, "reward": 0.16500000655651093, "reward_std": 0.0764847844839096, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0650000050663948, "rewards/unicoder_reward_fn/std": 0.15186603367328644, "step": 189 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 819.0, "completions/max_terminated_length": 819.0, "completions/mean_length": 335.6499938964844, "completions/mean_terminated_length": 335.6499938964844, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "epoch": 0.07370054305663305, "grad_norm": 0.7881373763084412, "kl": 0.162841796875, "learning_rate": 4.978146855468538e-06, "loss": 0.0167, "num_tokens": 16415153.0, "reward": 0.38750001788139343, "reward_std": 0.1864456832408905, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.3803313076496124, "step": 190 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1081.0, "completions/max_terminated_length": 1081.0, "completions/mean_length": 352.5375061035156, "completions/mean_terminated_length": 352.5375061035156, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.07408844065166796, "grad_norm": 0.840052604675293, "kl": 0.156005859375, "learning_rate": 4.977751984571592e-06, "loss": 0.042, "num_tokens": 16489124.0, "reward": 0.30250000953674316, "reward_std": 0.19631895422935486, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.28372523188591003, "step": 191 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 774.0, "completions/max_terminated_length": 774.0, "completions/mean_length": 373.7875061035156, "completions/mean_terminated_length": 373.7875061035156, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.07447633824670287, "grad_norm": 0.6509885787963867, "kl": 0.16015625, "learning_rate": 4.977353595750181e-06, "loss": -0.011, "num_tokens": 16565043.0, "reward": 0.3425000309944153, "reward_std": 0.10418402403593063, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24249999225139618, "rewards/unicoder_reward_fn/std": 0.35818636417388916, "step": 192 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2734.0, "completions/max_terminated_length": 2734.0, "completions/mean_length": 537.0499877929688, "completions/mean_terminated_length": 537.0499877929688, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.07486423584173778, "grad_norm": 0.7917982935905457, "kl": 0.137451171875, "learning_rate": 4.9769516896334175e-06, "loss": 0.0041, "num_tokens": 16659975.0, "reward": 0.3487499952316284, "reward_std": 0.15029816329479218, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.25, "rewards/unicoder_reward_fn/std": 0.3785381615161896, "step": 193 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3343.0, "completions/max_terminated_length": 3343.0, "completions/mean_length": 602.1000366210938, "completions/mean_terminated_length": 602.1000366210938, "completions/min_length": 275.0, "completions/min_terminated_length": 275.0, "epoch": 0.07525213343677269, "grad_norm": 0.42735305428504944, "kl": 0.11474609375, "learning_rate": 4.976546266855965e-06, "loss": 0.0224, "num_tokens": 16766383.0, "reward": 0.1550000011920929, "reward_std": 0.06739810854196548, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.14222463965415955, "step": 194 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 969.0, "completions/max_terminated_length": 969.0, "completions/mean_length": 471.3374938964844, "completions/mean_terminated_length": 471.3374938964844, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "epoch": 0.07564003103180761, "grad_norm": 0.48070913553237915, "kl": 0.1334228515625, "learning_rate": 4.976137328058042e-06, "loss": 0.0288, "num_tokens": 16857338.0, "reward": 0.13750000298023224, "reward_std": 0.043164756149053574, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.09053288400173187, "step": 195 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2480.0, "completions/max_terminated_length": 2480.0, "completions/mean_length": 557.7750244140625, "completions/mean_terminated_length": 557.7750244140625, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.07602792862684252, "grad_norm": 0.610073983669281, "kl": 0.127197265625, "learning_rate": 4.975724873885418e-06, "loss": 0.0197, "num_tokens": 16954280.0, "reward": 0.2200000137090683, "reward_std": 0.08647775650024414, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.20769988000392914, "step": 196 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 867.0, "completions/max_terminated_length": 867.0, "completions/mean_length": 437.8125, "completions/mean_terminated_length": 437.8125, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.07641582622187743, "grad_norm": 0.8149270415306091, "kl": 0.1170654296875, "learning_rate": 4.975308904989415e-06, "loss": -0.0086, "num_tokens": 17046721.0, "reward": 0.2874999940395355, "reward_std": 0.1948740929365158, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.18999998271465302, "rewards/unicoder_reward_fn/std": 0.2790580987930298, "step": 197 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1061.0, "completions/max_terminated_length": 1061.0, "completions/mean_length": 487.1625061035156, "completions/mean_terminated_length": 487.1625061035156, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.07680372381691233, "grad_norm": 0.4917312264442444, "kl": 0.120361328125, "learning_rate": 4.974889422026905e-06, "loss": 0.0317, "num_tokens": 17121510.0, "reward": 0.25, "reward_std": 0.11377506703138351, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.3390231728553772, "step": 198 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 874.0, "completions/max_terminated_length": 874.0, "completions/mean_length": 460.1000061035156, "completions/mean_terminated_length": 460.1000061035156, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.07719162141194724, "grad_norm": 0.6986340284347534, "kl": 0.1314697265625, "learning_rate": 4.974466425660307e-06, "loss": -0.0322, "num_tokens": 17201638.0, "reward": 0.29250001907348633, "reward_std": 0.14314042031764984, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.23045086860656738, "step": 199 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1144.0, "completions/max_terminated_length": 1144.0, "completions/mean_length": 504.88751220703125, "completions/mean_terminated_length": 504.88751220703125, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "epoch": 0.07757951900698215, "grad_norm": 0.6592660546302795, "kl": 0.133056640625, "learning_rate": 4.974039916557592e-06, "loss": 0.0218, "num_tokens": 17289197.0, "reward": 0.24375000596046448, "reward_std": 0.10192903131246567, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.19086743891239166, "step": 200 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1376.0, "completions/max_terminated_length": 1376.0, "completions/mean_length": 537.0625, "completions/mean_terminated_length": 537.0625, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.07796741660201707, "grad_norm": 0.5784204602241516, "kl": 0.134521484375, "learning_rate": 4.9736098953922755e-06, "loss": 0.0424, "num_tokens": 17379010.0, "reward": 0.3750000298023224, "reward_std": 0.13842399418354034, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.3836863040924072, "step": 201 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1336.0, "completions/max_terminated_length": 1336.0, "completions/mean_length": 484.6499938964844, "completions/mean_terminated_length": 484.6499938964844, "completions/min_length": 200.0, "completions/min_terminated_length": 200.0, "epoch": 0.07835531419705198, "grad_norm": 0.6285555958747864, "kl": 0.122314453125, "learning_rate": 4.973176362843418e-06, "loss": -0.0209, "num_tokens": 17469142.0, "reward": 0.2824999988079071, "reward_std": 0.12071912735700607, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.2777737081050873, "step": 202 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 957.0, "completions/mean_length": 578.0750122070312, "completions/mean_terminated_length": 533.5443115234375, "completions/min_length": 271.0, "completions/min_terminated_length": 271.0, "epoch": 0.07874321179208689, "grad_norm": 0.6344579458236694, "kl": 0.12646484375, "learning_rate": 4.97273931959563e-06, "loss": 0.0591, "num_tokens": 17559236.0, "reward": 0.24124999344348907, "reward_std": 0.12572196125984192, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.21861185133457184, "step": 203 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1181.0, "completions/max_terminated_length": 1181.0, "completions/mean_length": 554.4500122070312, "completions/mean_terminated_length": 554.4500122070312, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "epoch": 0.0791311093871218, "grad_norm": 0.6375933885574341, "kl": 0.1185302734375, "learning_rate": 4.97229876633906e-06, "loss": 0.0397, "num_tokens": 17652880.0, "reward": 0.3725000321865082, "reward_std": 0.15498560667037964, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2724999785423279, "rewards/unicoder_reward_fn/std": 0.3489496111869812, "step": 204 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2593.0, "completions/max_terminated_length": 2593.0, "completions/mean_length": 542.7125244140625, "completions/mean_terminated_length": 542.7125244140625, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.0795190069821567, "grad_norm": 0.5053399205207825, "kl": 0.1246337890625, "learning_rate": 4.971854703769402e-06, "loss": 0.0178, "num_tokens": 17736137.0, "reward": 0.4000000059604645, "reward_std": 0.1957085132598877, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.30000001192092896, "rewards/unicoder_reward_fn/std": 0.42217040061950684, "step": 205 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2775.0, "completions/max_terminated_length": 2775.0, "completions/mean_length": 554.6625366210938, "completions/mean_terminated_length": 554.6625366210938, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.07990690457719161, "grad_norm": 0.6508405208587646, "kl": 0.1380615234375, "learning_rate": 4.971407132587892e-06, "loss": -0.0014, "num_tokens": 17834630.0, "reward": 0.2175000160932541, "reward_std": 0.12133164703845978, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.18471670150756836, "step": 206 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2807.0, "completions/max_terminated_length": 2807.0, "completions/mean_length": 603.9249877929688, "completions/mean_terminated_length": 603.9249877929688, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.08029480217222654, "grad_norm": 0.5423001646995544, "kl": 0.1234130859375, "learning_rate": 4.9709560535013055e-06, "loss": 0.0343, "num_tokens": 17935056.0, "reward": 0.36500003933906555, "reward_std": 0.15734592080116272, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.26499998569488525, "rewards/unicoder_reward_fn/std": 0.3815690577030182, "step": 207 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 928.0, "completions/max_terminated_length": 928.0, "completions/mean_length": 518.9125366210938, "completions/mean_terminated_length": 518.9125366210938, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "epoch": 0.08068269976726145, "grad_norm": 0.636978805065155, "kl": 0.12451171875, "learning_rate": 4.970501467221958e-06, "loss": 0.0145, "num_tokens": 18025265.0, "reward": 0.19750000536441803, "reward_std": 0.11083877086639404, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09750001132488251, "rewards/unicoder_reward_fn/std": 0.20621666312217712, "step": 208 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3643.0, "completions/max_terminated_length": 3643.0, "completions/mean_length": 606.9000244140625, "completions/mean_terminated_length": 606.9000244140625, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.08107059736229635, "grad_norm": 0.6641480922698975, "kl": 0.11767578125, "learning_rate": 4.9700433744677035e-06, "loss": 0.0788, "num_tokens": 18120865.0, "reward": 0.2512499988079071, "reward_std": 0.16978582739830017, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.21166908740997314, "step": 209 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1036.0, "completions/max_terminated_length": 1036.0, "completions/mean_length": 490.3625183105469, "completions/mean_terminated_length": 490.3625183105469, "completions/min_length": 234.0, "completions/min_terminated_length": 234.0, "epoch": 0.08145849495733126, "grad_norm": 0.5294095873832703, "kl": 0.125, "learning_rate": 4.9695817759619324e-06, "loss": 0.0058, "num_tokens": 18199350.0, "reward": 0.30000004172325134, "reward_std": 0.13617883622646332, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.3035653233528137, "step": 210 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1182.0, "completions/max_terminated_length": 1182.0, "completions/mean_length": 565.25, "completions/mean_terminated_length": 565.25, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.08184639255236617, "grad_norm": 0.45027196407318115, "kl": 0.115966796875, "learning_rate": 4.969116672433573e-06, "loss": 0.0457, "num_tokens": 18288842.0, "reward": 0.2224999964237213, "reward_std": 0.05977179482579231, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12250000238418579, "rewards/unicoder_reward_fn/std": 0.19998417794704437, "step": 211 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1282.0, "completions/max_terminated_length": 1282.0, "completions/mean_length": 524.5125122070312, "completions/mean_terminated_length": 524.5125122070312, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.08223429014740109, "grad_norm": 0.5149121880531311, "kl": 0.130859375, "learning_rate": 4.9686480646170856e-06, "loss": 0.0268, "num_tokens": 18382011.0, "reward": 0.30000001192092896, "reward_std": 0.09162890166044235, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.32917237281799316, "step": 212 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1366.0, "completions/max_terminated_length": 1366.0, "completions/mean_length": 602.9249877929688, "completions/mean_terminated_length": 602.9249877929688, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.082622187742436, "grad_norm": 0.6666144132614136, "kl": 0.1104736328125, "learning_rate": 4.968175953252468e-06, "loss": 0.0088, "num_tokens": 18477981.0, "reward": 0.3175000250339508, "reward_std": 0.22333459556102753, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2175000011920929, "rewards/unicoder_reward_fn/std": 0.2971489131450653, "step": 213 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2306.0, "completions/max_terminated_length": 2306.0, "completions/mean_length": 604.6500244140625, "completions/mean_terminated_length": 604.6500244140625, "completions/min_length": 257.0, "completions/min_terminated_length": 257.0, "epoch": 0.08301008533747091, "grad_norm": 0.4824112355709076, "kl": 0.1156005859375, "learning_rate": 4.967700339085246e-06, "loss": 0.0147, "num_tokens": 18571945.0, "reward": 0.16624999046325684, "reward_std": 0.10903497040271759, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.18264926970005035, "step": 214 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1897.0, "completions/max_terminated_length": 1897.0, "completions/mean_length": 732.0, "completions/mean_terminated_length": 732.0, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.08339798293250582, "grad_norm": 0.439778596162796, "kl": 0.10546875, "learning_rate": 4.9672212228664814e-06, "loss": 0.0084, "num_tokens": 18682465.0, "reward": 0.23624999821186066, "reward_std": 0.05723550543189049, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.2926850914955139, "step": 215 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1175.0, "completions/max_terminated_length": 1175.0, "completions/mean_length": 507.6000061035156, "completions/mean_terminated_length": 507.6000061035156, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.08378588052754073, "grad_norm": 0.5554835796356201, "kl": 0.135498046875, "learning_rate": 4.9667386053527625e-06, "loss": 0.0149, "num_tokens": 18768625.0, "reward": 0.23000001907348633, "reward_std": 0.09127090126276016, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.20151327550411224, "step": 216 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1132.0, "completions/max_terminated_length": 1132.0, "completions/mean_length": 589.4500122070312, "completions/mean_terminated_length": 589.4500122070312, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.08417377812257563, "grad_norm": 0.5687991976737976, "kl": 0.130615234375, "learning_rate": 4.966252487306209e-06, "loss": 0.0089, "num_tokens": 18856117.0, "reward": 0.26750001311302185, "reward_std": 0.17255084216594696, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16750000417232513, "rewards/unicoder_reward_fn/std": 0.28541553020477295, "step": 217 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1066.0, "completions/max_terminated_length": 1066.0, "completions/mean_length": 555.0125122070312, "completions/mean_terminated_length": 555.0125122070312, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.08456167571761056, "grad_norm": 0.6186714172363281, "kl": 0.1209716796875, "learning_rate": 4.965762869494468e-06, "loss": -0.0243, "num_tokens": 18947798.0, "reward": 0.2874999940395355, "reward_std": 0.14413680136203766, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.30079010128974915, "step": 218 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1830.0, "completions/mean_length": 674.0375366210938, "completions/mean_terminated_length": 630.7215576171875, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.08494957331264547, "grad_norm": 0.6153665781021118, "kl": 0.1085205078125, "learning_rate": 4.965269752690713e-06, "loss": 0.0449, "num_tokens": 19050537.0, "reward": 0.1862500160932541, "reward_std": 0.08684924244880676, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.08750000596046448, "rewards/unicoder_reward_fn/std": 0.16177420318126678, "step": 219 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1772.0, "completions/max_terminated_length": 1772.0, "completions/mean_length": 615.6875, "completions/mean_terminated_length": 615.6875, "completions/min_length": 312.0, "completions/min_terminated_length": 312.0, "epoch": 0.08533747090768037, "grad_norm": 0.5182559490203857, "kl": 0.132080078125, "learning_rate": 4.964773137673642e-06, "loss": 0.0165, "num_tokens": 19148048.0, "reward": 0.17249999940395355, "reward_std": 0.07645450532436371, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.16911910474300385, "step": 220 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2463.0, "completions/max_terminated_length": 2463.0, "completions/mean_length": 650.4750366210938, "completions/mean_terminated_length": 650.4750366210938, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "epoch": 0.08572536850271528, "grad_norm": 0.6005114316940308, "kl": 0.1190185546875, "learning_rate": 4.9642730252274766e-06, "loss": 0.0357, "num_tokens": 19250126.0, "reward": 0.2800000309944153, "reward_std": 0.11694544553756714, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.18000002205371857, "rewards/unicoder_reward_fn/std": 0.30122116208076477, "step": 221 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1331.0, "completions/max_terminated_length": 1331.0, "completions/mean_length": 572.4000244140625, "completions/mean_terminated_length": 572.4000244140625, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.08611326609775019, "grad_norm": 0.6565302014350891, "kl": 0.1290283203125, "learning_rate": 4.963769416141966e-06, "loss": 0.0058, "num_tokens": 19341382.0, "reward": 0.29749998450279236, "reward_std": 0.18267658352851868, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19750000536441803, "rewards/unicoder_reward_fn/std": 0.23491178452968597, "step": 222 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2295.0, "completions/max_terminated_length": 2295.0, "completions/mean_length": 578.75, "completions/mean_terminated_length": 578.75, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.0865011636927851, "grad_norm": 0.7366396188735962, "kl": 0.14111328125, "learning_rate": 4.9632623112123755e-06, "loss": -0.0072, "num_tokens": 19433458.0, "reward": 0.20875000953674316, "reward_std": 0.15025866031646729, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.20351344347000122, "step": 223 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2120.0, "completions/max_terminated_length": 2120.0, "completions/mean_length": 818.1000366210938, "completions/mean_terminated_length": 818.1000366210938, "completions/min_length": 359.0, "completions/min_terminated_length": 359.0, "epoch": 0.08688906128782002, "grad_norm": 0.40322956442832947, "kl": 0.1043701171875, "learning_rate": 4.962751711239493e-06, "loss": 0.0183, "num_tokens": 19557586.0, "reward": 0.18125000596046448, "reward_std": 0.09429683536291122, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.20299966633319855, "step": 224 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1207.0, "completions/max_terminated_length": 1207.0, "completions/mean_length": 601.75, "completions/mean_terminated_length": 601.75, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.08727695888285493, "grad_norm": 0.5838728547096252, "kl": 0.1268310546875, "learning_rate": 4.962237617029625e-06, "loss": 0.0024, "num_tokens": 19649094.0, "reward": 0.27250000834465027, "reward_std": 0.15554679930210114, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17250001430511475, "rewards/unicoder_reward_fn/std": 0.2732713222503662, "step": 225 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1249.0, "completions/max_terminated_length": 1249.0, "completions/mean_length": 574.2875366210938, "completions/mean_terminated_length": 574.2875366210938, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.08766485647788984, "grad_norm": 0.6256162524223328, "kl": 0.1270751953125, "learning_rate": 4.961720029394598e-06, "loss": -0.0159, "num_tokens": 19738749.0, "reward": 0.23499999940395355, "reward_std": 0.08083128184080124, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.23117752373218536, "step": 226 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1618.0, "completions/max_terminated_length": 1618.0, "completions/mean_length": 584.2625122070312, "completions/mean_terminated_length": 584.2625122070312, "completions/min_length": 312.0, "completions/min_terminated_length": 312.0, "epoch": 0.08805275407292475, "grad_norm": 0.606743574142456, "kl": 0.1192626953125, "learning_rate": 4.961198949151753e-06, "loss": 0.0377, "num_tokens": 19829594.0, "reward": 0.23499999940395355, "reward_std": 0.15588393807411194, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13499999046325684, "rewards/unicoder_reward_fn/std": 0.2580844759941101, "step": 227 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1668.0, "completions/max_terminated_length": 1668.0, "completions/mean_length": 613.5750122070312, "completions/mean_terminated_length": 613.5750122070312, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "epoch": 0.08844065166795965, "grad_norm": 0.6333354711532593, "kl": 0.1273193359375, "learning_rate": 4.960674377123947e-06, "loss": 0.0421, "num_tokens": 19929168.0, "reward": 0.24375000596046448, "reward_std": 0.11257711797952652, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.31012454628944397, "step": 228 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1926.0, "completions/max_terminated_length": 1926.0, "completions/mean_length": 686.0625, "completions/mean_terminated_length": 686.0625, "completions/min_length": 321.0, "completions/min_terminated_length": 321.0, "epoch": 0.08882854926299456, "grad_norm": 0.6180984377861023, "kl": 0.11083984375, "learning_rate": 4.96014631413955e-06, "loss": 0.0316, "num_tokens": 20038589.0, "reward": 0.29375001788139343, "reward_std": 0.18031208217144012, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.19500000774860382, "rewards/unicoder_reward_fn/std": 0.2967759370803833, "step": 229 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1540.0, "completions/max_terminated_length": 1540.0, "completions/mean_length": 559.5125122070312, "completions/mean_terminated_length": 559.5125122070312, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "epoch": 0.08921644685802949, "grad_norm": 0.8247970342636108, "kl": 0.142578125, "learning_rate": 4.959614761032446e-06, "loss": 0.0716, "num_tokens": 20124926.0, "reward": 0.26750001311302185, "reward_std": 0.16438858211040497, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.29927125573158264, "step": 230 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1583.0, "completions/max_terminated_length": 1583.0, "completions/mean_length": 680.0625, "completions/mean_terminated_length": 680.0625, "completions/min_length": 316.0, "completions/min_terminated_length": 316.0, "epoch": 0.0896043444530644, "grad_norm": 0.5177686810493469, "kl": 0.108154296875, "learning_rate": 4.959079718642031e-06, "loss": -0.0071, "num_tokens": 20227603.0, "reward": 0.2525000274181366, "reward_std": 0.1647680103778839, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.2814856767654419, "step": 231 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1202.0, "completions/max_terminated_length": 1202.0, "completions/mean_length": 580.4125366210938, "completions/mean_terminated_length": 580.4125366210938, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "epoch": 0.0899922420480993, "grad_norm": 0.5660355687141418, "kl": 0.11865234375, "learning_rate": 4.958541187813207e-06, "loss": 0.0332, "num_tokens": 20324244.0, "reward": 0.3162499964237213, "reward_std": 0.1307021677494049, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.2175000011920929, "rewards/unicoder_reward_fn/std": 0.33252963423728943, "step": 232 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 988.0, "completions/max_terminated_length": 988.0, "completions/mean_length": 578.8500366210938, "completions/mean_terminated_length": 578.8500366210938, "completions/min_length": 303.0, "completions/min_terminated_length": 303.0, "epoch": 0.09038013964313421, "grad_norm": 0.5680636167526245, "kl": 0.1241455078125, "learning_rate": 4.957999169396392e-06, "loss": 0.0293, "num_tokens": 20424392.0, "reward": 0.26500001549720764, "reward_std": 0.13515999913215637, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.27883121371269226, "step": 233 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1044.0, "completions/max_terminated_length": 1044.0, "completions/mean_length": 569.875, "completions/mean_terminated_length": 569.875, "completions/min_length": 313.0, "completions/min_terminated_length": 313.0, "epoch": 0.09076803723816912, "grad_norm": 0.5908787846565247, "kl": 0.1212158203125, "learning_rate": 4.957453664247502e-06, "loss": 0.0044, "num_tokens": 20514342.0, "reward": 0.24124999344348907, "reward_std": 0.11625418812036514, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.2971488833427429, "step": 234 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1060.0, "completions/max_terminated_length": 1060.0, "completions/mean_length": 573.1875, "completions/mean_terminated_length": 573.1875, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "epoch": 0.09115593483320403, "grad_norm": 0.6202223300933838, "kl": 0.12109375, "learning_rate": 4.956904673227968e-06, "loss": -0.0046, "num_tokens": 20610205.0, "reward": 0.3125, "reward_std": 0.17571832239627838, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.21250002086162567, "rewards/unicoder_reward_fn/std": 0.2687877118587494, "step": 235 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 817.0, "completions/max_terminated_length": 817.0, "completions/mean_length": 471.5375061035156, "completions/mean_terminated_length": 471.5375061035156, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "epoch": 0.09154383242823895, "grad_norm": 0.6726969480514526, "kl": 0.131591796875, "learning_rate": 4.95635219720472e-06, "loss": -0.022, "num_tokens": 20690488.0, "reward": 0.23749999701976776, "reward_std": 0.17461137473583221, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.26737117767333984, "step": 236 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1043.0, "completions/max_terminated_length": 1043.0, "completions/mean_length": 515.3375244140625, "completions/mean_terminated_length": 515.3375244140625, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "epoch": 0.09193173002327386, "grad_norm": 0.6393038034439087, "kl": 0.1236572265625, "learning_rate": 4.955796237050195e-06, "loss": -0.0123, "num_tokens": 20774523.0, "reward": 0.3525000214576721, "reward_std": 0.11577508598566055, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2524999976158142, "rewards/unicoder_reward_fn/std": 0.3256833255290985, "step": 237 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1382.0, "completions/max_terminated_length": 1382.0, "completions/mean_length": 524.4500122070312, "completions/mean_terminated_length": 524.4500122070312, "completions/min_length": 224.0, "completions/min_terminated_length": 224.0, "epoch": 0.09231962761830877, "grad_norm": 0.4821893274784088, "kl": 0.13720703125, "learning_rate": 4.9552367936423275e-06, "loss": -0.0082, "num_tokens": 20863775.0, "reward": 0.33249998092651367, "reward_std": 0.09878727048635483, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23250000178813934, "rewards/unicoder_reward_fn/std": 0.32985422015190125, "step": 238 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1943.0, "completions/max_terminated_length": 1943.0, "completions/mean_length": 622.4375, "completions/mean_terminated_length": 622.4375, "completions/min_length": 356.0, "completions/min_terminated_length": 356.0, "epoch": 0.09270752521334368, "grad_norm": 0.5668232440948486, "kl": 0.130615234375, "learning_rate": 4.954673867864556e-06, "loss": 0.0339, "num_tokens": 20955994.0, "reward": 0.24500000476837158, "reward_std": 0.08750516176223755, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.15903188288211823, "step": 239 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1468.0, "completions/max_terminated_length": 1468.0, "completions/mean_length": 646.9625244140625, "completions/mean_terminated_length": 646.9625244140625, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.09309542280837858, "grad_norm": 0.5247519612312317, "kl": 0.1229248046875, "learning_rate": 4.954107460605817e-06, "loss": 0.0247, "num_tokens": 21057271.0, "reward": 0.17749999463558197, "reward_std": 0.08460516482591629, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.1292382776737213, "step": 240 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 927.0, "completions/max_terminated_length": 927.0, "completions/mean_length": 522.9375, "completions/mean_terminated_length": 522.9375, "completions/min_length": 301.0, "completions/min_terminated_length": 301.0, "epoch": 0.0934833204034135, "grad_norm": 0.45199891924858093, "kl": 0.134033203125, "learning_rate": 4.953537572760545e-06, "loss": 0.008, "num_tokens": 21143226.0, "reward": 0.20750001072883606, "reward_std": 0.08968909829854965, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.23801659047603607, "step": 241 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1097.0, "completions/max_terminated_length": 1097.0, "completions/mean_length": 520.4625244140625, "completions/mean_terminated_length": 520.4625244140625, "completions/min_length": 275.0, "completions/min_terminated_length": 275.0, "epoch": 0.09387121799844841, "grad_norm": 0.628884494304657, "kl": 0.152099609375, "learning_rate": 4.952964205228671e-06, "loss": 0.0197, "num_tokens": 21227599.0, "reward": 0.21500001847743988, "reward_std": 0.12813766300678253, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.21763981878757477, "step": 242 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1219.0, "completions/max_terminated_length": 1219.0, "completions/mean_length": 587.2750244140625, "completions/mean_terminated_length": 587.2750244140625, "completions/min_length": 326.0, "completions/min_terminated_length": 326.0, "epoch": 0.09425911559348332, "grad_norm": 0.5939902067184448, "kl": 0.150390625, "learning_rate": 4.952387358915621e-06, "loss": 0.0415, "num_tokens": 21329741.0, "reward": 0.2775000035762787, "reward_std": 0.13290873169898987, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.28638955950737, "step": 243 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1477.0, "completions/max_terminated_length": 1477.0, "completions/mean_length": 549.1124877929688, "completions/mean_terminated_length": 549.1124877929688, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.09464701318851823, "grad_norm": 0.609360933303833, "kl": 0.1396484375, "learning_rate": 4.951807034732313e-06, "loss": -0.0153, "num_tokens": 21419838.0, "reward": 0.24000000953674316, "reward_std": 0.10452865809202194, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14000001549720764, "rewards/unicoder_reward_fn/std": 0.2957932949066162, "step": 244 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1280.0, "completions/max_terminated_length": 1280.0, "completions/mean_length": 530.2875366210938, "completions/mean_terminated_length": 530.2875366210938, "completions/min_length": 258.0, "completions/min_terminated_length": 258.0, "epoch": 0.09503491078355314, "grad_norm": 0.7368508577346802, "kl": 0.14404296875, "learning_rate": 4.951223233595158e-06, "loss": 0.0051, "num_tokens": 21516949.0, "reward": 0.2212500125169754, "reward_std": 0.17252914607524872, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.12250000238418579, "rewards/unicoder_reward_fn/std": 0.21930557489395142, "step": 245 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1761.0, "completions/max_terminated_length": 1761.0, "completions/mean_length": 577.3624877929688, "completions/mean_terminated_length": 577.3624877929688, "completions/min_length": 278.0, "completions/min_terminated_length": 278.0, "epoch": 0.09542280837858805, "grad_norm": 0.616127073764801, "kl": 0.1259765625, "learning_rate": 4.9506359564260585e-06, "loss": -0.0002, "num_tokens": 21612258.0, "reward": 0.28999999165534973, "reward_std": 0.17127439379692078, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1899999976158142, "rewards/unicoder_reward_fn/std": 0.2949361801147461, "step": 246 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1168.0, "completions/max_terminated_length": 1168.0, "completions/mean_length": 502.1000061035156, "completions/mean_terminated_length": 502.1000061035156, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.09581070597362297, "grad_norm": 0.6955956220626831, "kl": 0.147705078125, "learning_rate": 4.950045204152405e-06, "loss": 0.0203, "num_tokens": 21698434.0, "reward": 0.2862499952316284, "reward_std": 0.12135125696659088, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.33275794982910156, "step": 247 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1767.0, "completions/max_terminated_length": 1767.0, "completions/mean_length": 661.9750366210938, "completions/mean_terminated_length": 661.9750366210938, "completions/min_length": 331.0, "completions/min_terminated_length": 331.0, "epoch": 0.09619860356865788, "grad_norm": 0.6523640155792236, "kl": 0.128662109375, "learning_rate": 4.949450977707076e-06, "loss": 0.0069, "num_tokens": 21810928.0, "reward": 0.19750000536441803, "reward_std": 0.1311897486448288, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.18824639916419983, "step": 248 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1276.0, "completions/max_terminated_length": 1276.0, "completions/mean_length": 543.4750366210938, "completions/mean_terminated_length": 543.4750366210938, "completions/min_length": 280.0, "completions/min_terminated_length": 280.0, "epoch": 0.09658650116369279, "grad_norm": 0.6260305643081665, "kl": 0.11767578125, "learning_rate": 4.948853278028436e-06, "loss": 0.0143, "num_tokens": 21898758.0, "reward": 0.2862499952316284, "reward_std": 0.1741112768650055, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.1875000149011612, "rewards/unicoder_reward_fn/std": 0.28700771927833557, "step": 249 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1097.0, "completions/max_terminated_length": 1097.0, "completions/mean_length": 526.4625244140625, "completions/mean_terminated_length": 526.4625244140625, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "epoch": 0.0969743987587277, "grad_norm": 0.7393983602523804, "kl": 0.131103515625, "learning_rate": 4.948252106060334e-06, "loss": 0.0144, "num_tokens": 21987779.0, "reward": 0.3400000035762787, "reward_std": 0.19722728431224823, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24000000953674316, "rewards/unicoder_reward_fn/std": 0.3343612849712372, "step": 250 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1206.0, "completions/max_terminated_length": 1206.0, "completions/mean_length": 552.875, "completions/mean_terminated_length": 552.875, "completions/min_length": 310.0, "completions/min_terminated_length": 310.0, "epoch": 0.0973622963537626, "grad_norm": 0.48460811376571655, "kl": 0.137939453125, "learning_rate": 4.947647462752104e-06, "loss": 0.003, "num_tokens": 22089153.0, "reward": 0.15125000476837158, "reward_std": 0.05366116389632225, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.05250000208616257, "rewards/unicoder_reward_fn/std": 0.10905230790376663, "step": 251 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 907.0, "completions/max_terminated_length": 907.0, "completions/mean_length": 535.5, "completions/mean_terminated_length": 535.5, "completions/min_length": 305.0, "completions/min_terminated_length": 305.0, "epoch": 0.09775019394879751, "grad_norm": 0.5846840739250183, "kl": 0.154296875, "learning_rate": 4.9470393490585585e-06, "loss": 0.004, "num_tokens": 22175985.0, "reward": 0.22500000894069672, "reward_std": 0.09504849463701248, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.18657474219799042, "step": 252 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1158.0, "completions/max_terminated_length": 1158.0, "completions/mean_length": 532.6124877929688, "completions/mean_terminated_length": 532.6124877929688, "completions/min_length": 242.0, "completions/min_terminated_length": 242.0, "epoch": 0.09813809154383243, "grad_norm": 0.6002660989761353, "kl": 0.1304931640625, "learning_rate": 4.9464277659399936e-06, "loss": 0.0064, "num_tokens": 22267858.0, "reward": 0.3575000464916229, "reward_std": 0.14995543658733368, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.257500022649765, "rewards/unicoder_reward_fn/std": 0.32011669874191284, "step": 253 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1399.0, "completions/max_terminated_length": 1399.0, "completions/mean_length": 508.5375061035156, "completions/mean_terminated_length": 508.5375061035156, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.09852598913886734, "grad_norm": 0.6817594766616821, "kl": 0.148681640625, "learning_rate": 4.945812714362182e-06, "loss": -0.0017, "num_tokens": 22357637.0, "reward": 0.2825000286102295, "reward_std": 0.09715896844863892, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.3294702470302582, "step": 254 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1430.0, "completions/max_terminated_length": 1430.0, "completions/mean_length": 538.0125122070312, "completions/mean_terminated_length": 538.0125122070312, "completions/min_length": 295.0, "completions/min_terminated_length": 295.0, "epoch": 0.09891388673390225, "grad_norm": 0.5336037278175354, "kl": 0.1358642578125, "learning_rate": 4.945194195296374e-06, "loss": 0.0047, "num_tokens": 22452430.0, "reward": 0.23000001907348633, "reward_std": 0.10527608543634415, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.20647738873958588, "step": 255 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1914.0, "completions/max_terminated_length": 1914.0, "completions/mean_length": 537.5125122070312, "completions/mean_terminated_length": 537.5125122070312, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.09930178432893716, "grad_norm": 0.5909339785575867, "kl": 0.1474609375, "learning_rate": 4.944572209719295e-06, "loss": 0.0371, "num_tokens": 22541343.0, "reward": 0.257500022649765, "reward_std": 0.12644517421722412, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.3030540645122528, "step": 256 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1070.0, "completions/max_terminated_length": 1070.0, "completions/mean_length": 520.6124877929688, "completions/mean_terminated_length": 520.6124877929688, "completions/min_length": 283.0, "completions/min_terminated_length": 283.0, "epoch": 0.09968968192397207, "grad_norm": 0.6363409161567688, "kl": 0.1435546875, "learning_rate": 4.943946758613147e-06, "loss": 0.0224, "num_tokens": 22631832.0, "reward": 0.27125000953674316, "reward_std": 0.17842164635658264, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.17249999940395355, "rewards/unicoder_reward_fn/std": 0.314612478017807, "step": 257 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 776.0, "completions/max_terminated_length": 776.0, "completions/mean_length": 480.0500183105469, "completions/mean_terminated_length": 480.0500183105469, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.10007757951900698, "grad_norm": 0.7723165154457092, "kl": 0.154296875, "learning_rate": 4.9433178429656e-06, "loss": 0.0046, "num_tokens": 22715508.0, "reward": 0.25, "reward_std": 0.1342344582080841, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.25557082891464233, "step": 258 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1074.0, "completions/max_terminated_length": 1074.0, "completions/mean_length": 473.875, "completions/mean_terminated_length": 473.875, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.1004654771140419, "grad_norm": 0.7763696312904358, "kl": 0.169921875, "learning_rate": 4.942685463769799e-06, "loss": 0.0184, "num_tokens": 22805354.0, "reward": 0.24249999225139618, "reward_std": 0.14219455420970917, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14250001311302185, "rewards/unicoder_reward_fn/std": 0.19407358765602112, "step": 259 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 808.0, "completions/max_terminated_length": 808.0, "completions/mean_length": 466.6625061035156, "completions/mean_terminated_length": 466.6625061035156, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "epoch": 0.1008533747090768, "grad_norm": 0.47525516152381897, "kl": 0.15478515625, "learning_rate": 4.942049622024357e-06, "loss": 0.0073, "num_tokens": 22892055.0, "reward": 0.27250000834465027, "reward_std": 0.06100421026349068, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17249999940395355, "rewards/unicoder_reward_fn/std": 0.22444023191928864, "step": 260 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1856.0, "completions/max_terminated_length": 1856.0, "completions/mean_length": 589.4375, "completions/mean_terminated_length": 589.4375, "completions/min_length": 261.0, "completions/min_terminated_length": 261.0, "epoch": 0.10124127230411172, "grad_norm": 0.6531454920768738, "kl": 0.141845703125, "learning_rate": 4.941410318733356e-06, "loss": 0.0051, "num_tokens": 22996674.0, "reward": 0.17499999701976776, "reward_std": 0.12795203924179077, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.16342845559120178, "step": 261 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 791.0, "completions/max_terminated_length": 791.0, "completions/mean_length": 513.0125122070312, "completions/mean_terminated_length": 513.0125122070312, "completions/min_length": 290.0, "completions/min_terminated_length": 290.0, "epoch": 0.10162916989914662, "grad_norm": 0.6214700937271118, "kl": 0.150634765625, "learning_rate": 4.940767554906341e-06, "loss": 0.0084, "num_tokens": 23080891.0, "reward": 0.23749999701976776, "reward_std": 0.15492728352546692, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.3207487463951111, "step": 262 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1170.0, "completions/max_terminated_length": 1170.0, "completions/mean_length": 583.4500122070312, "completions/mean_terminated_length": 583.4500122070312, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "epoch": 0.10201706749418153, "grad_norm": 0.544352650642395, "kl": 0.1380615234375, "learning_rate": 4.940121331558326e-06, "loss": 0.036, "num_tokens": 23184327.0, "reward": 0.24000000953674316, "reward_std": 0.1437753438949585, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.14250001311302185, "rewards/unicoder_reward_fn/std": 0.2867429554462433, "step": 263 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 873.0, "completions/max_terminated_length": 873.0, "completions/mean_length": 482.01251220703125, "completions/mean_terminated_length": 482.01251220703125, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.10240496508921644, "grad_norm": 0.5174285173416138, "kl": 0.170166015625, "learning_rate": 4.9394716497097855e-06, "loss": -0.004, "num_tokens": 23275968.0, "reward": 0.20000000298023224, "reward_std": 0.09538667649030685, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.24649439752101898, "step": 264 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 942.0, "completions/max_terminated_length": 942.0, "completions/mean_length": 435.0874938964844, "completions/mean_terminated_length": 435.0874938964844, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.10279286268425136, "grad_norm": 0.5823934674263, "kl": 0.177734375, "learning_rate": 4.938818510386654e-06, "loss": -0.0129, "num_tokens": 23361767.0, "reward": 0.24250002205371857, "reward_std": 0.1425909847021103, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.27959057688713074, "step": 265 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 849.0, "completions/max_terminated_length": 849.0, "completions/mean_length": 446.5, "completions/mean_terminated_length": 446.5, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.10318076027928627, "grad_norm": 0.675086259841919, "kl": 0.167236328125, "learning_rate": 4.938161914620329e-06, "loss": 0.008, "num_tokens": 23445175.0, "reward": 0.23249998688697815, "reward_std": 0.10630675405263901, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.26036733388900757, "step": 266 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1299.0, "completions/max_terminated_length": 1299.0, "completions/mean_length": 552.9750366210938, "completions/mean_terminated_length": 552.9750366210938, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.10356865787432118, "grad_norm": 0.6754739880561829, "kl": 0.14306640625, "learning_rate": 4.937501863447665e-06, "loss": 0.0442, "num_tokens": 23549389.0, "reward": 0.26625001430511475, "reward_std": 0.16440080106258392, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.16750000417232513, "rewards/unicoder_reward_fn/std": 0.2632681727409363, "step": 267 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1221.0, "completions/max_terminated_length": 1221.0, "completions/mean_length": 413.3000183105469, "completions/mean_terminated_length": 413.3000183105469, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.10395655546935609, "grad_norm": 0.8321073055267334, "kl": 0.164306640625, "learning_rate": 4.936838357910972e-06, "loss": -0.0073, "num_tokens": 23621173.0, "reward": 0.39500001072883606, "reward_std": 0.237833172082901, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.29500001668930054, "rewards/unicoder_reward_fn/std": 0.35432231426239014, "step": 268 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 844.0, "completions/max_terminated_length": 844.0, "completions/mean_length": 481.0, "completions/mean_terminated_length": 481.0, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "epoch": 0.104344453064391, "grad_norm": 0.492504358291626, "kl": 0.17041015625, "learning_rate": 4.936171399058017e-06, "loss": 0.0028, "num_tokens": 23710269.0, "reward": 0.26750001311302185, "reward_std": 0.07633309811353683, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.29757457971572876, "step": 269 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1068.0, "completions/max_terminated_length": 1068.0, "completions/mean_length": 491.75, "completions/mean_terminated_length": 491.75, "completions/min_length": 288.0, "completions/min_terminated_length": 288.0, "epoch": 0.10473235065942592, "grad_norm": 0.6644993424415588, "kl": 0.18408203125, "learning_rate": 4.935500987942018e-06, "loss": 0.0146, "num_tokens": 23799841.0, "reward": 0.22500000894069672, "reward_std": 0.12938550114631653, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.1945458948612213, "step": 270 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 763.0, "completions/max_terminated_length": 763.0, "completions/mean_length": 386.7250061035156, "completions/mean_terminated_length": 386.7250061035156, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.10512024825446083, "grad_norm": 0.8506864309310913, "kl": 0.205078125, "learning_rate": 4.934827125621646e-06, "loss": 0.0251, "num_tokens": 23887819.0, "reward": 0.23500001430511475, "reward_std": 0.18116824328899384, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.22897681593894958, "step": 271 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1213.0, "completions/max_terminated_length": 1213.0, "completions/mean_length": 488.875, "completions/mean_terminated_length": 488.875, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.10550814584949574, "grad_norm": 0.7202557921409607, "kl": 0.16455078125, "learning_rate": 4.934149813161022e-06, "loss": 0.0436, "num_tokens": 23972505.0, "reward": 0.24500000476837158, "reward_std": 0.14395014941692352, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2157706320285797, "step": 272 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 822.0, "completions/max_terminated_length": 822.0, "completions/mean_length": 465.0249938964844, "completions/mean_terminated_length": 465.0249938964844, "completions/min_length": 288.0, "completions/min_terminated_length": 288.0, "epoch": 0.10589604344453064, "grad_norm": 0.7181267142295837, "kl": 0.1669921875, "learning_rate": 4.933469051629715e-06, "loss": 0.017, "num_tokens": 24061491.0, "reward": 0.2137499898672104, "reward_std": 0.1311948150396347, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.19815605878829956, "step": 273 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1018.0, "completions/max_terminated_length": 1018.0, "completions/mean_length": 445.0249938964844, "completions/mean_terminated_length": 445.0249938964844, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.10628394103956555, "grad_norm": 0.4985988736152649, "kl": 0.172119140625, "learning_rate": 4.932784842102739e-06, "loss": 0.0134, "num_tokens": 24149181.0, "reward": 0.27250000834465027, "reward_std": 0.09661700576543808, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17250001430511475, "rewards/unicoder_reward_fn/std": 0.2841710150241852, "step": 274 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 848.0, "completions/max_terminated_length": 848.0, "completions/mean_length": 441.0625, "completions/mean_terminated_length": 441.0625, "completions/min_length": 210.0, "completions/min_terminated_length": 210.0, "epoch": 0.10667183863460046, "grad_norm": 0.6757827997207642, "kl": 0.179443359375, "learning_rate": 4.9320971856605544e-06, "loss": 0.0034, "num_tokens": 24236994.0, "reward": 0.3737500011920929, "reward_std": 0.14049172401428223, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.34622740745544434, "step": 275 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 915.0, "completions/max_terminated_length": 915.0, "completions/mean_length": 431.7375183105469, "completions/mean_terminated_length": 431.7375183105469, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.10705973622963538, "grad_norm": 0.6428701281547546, "kl": 0.176025390625, "learning_rate": 4.931406083389067e-06, "loss": -0.0116, "num_tokens": 24312757.0, "reward": 0.2874999940395355, "reward_std": 0.12280664592981339, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.27804699540138245, "step": 276 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1020.0, "completions/max_terminated_length": 1020.0, "completions/mean_length": 386.8500061035156, "completions/mean_terminated_length": 386.8500061035156, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.10744763382467029, "grad_norm": 0.8871065974235535, "kl": 0.2021484375, "learning_rate": 4.9307115363796186e-06, "loss": -0.0035, "num_tokens": 24394897.0, "reward": 0.3225000500679016, "reward_std": 0.178131565451622, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2224999964237213, "rewards/unicoder_reward_fn/std": 0.26623252034187317, "step": 277 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 634.0, "completions/max_terminated_length": 634.0, "completions/mean_length": 348.8125, "completions/mean_terminated_length": 348.8125, "completions/min_length": 152.0, "completions/min_terminated_length": 152.0, "epoch": 0.1078355314197052, "grad_norm": 0.699694812297821, "kl": 0.234130859375, "learning_rate": 4.930013545728996e-06, "loss": 0.0001, "num_tokens": 24469178.0, "reward": 0.2912500202655792, "reward_std": 0.15275797247886658, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.3535623550415039, "step": 278 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1980.0, "completions/max_terminated_length": 1980.0, "completions/mean_length": 434.0500183105469, "completions/mean_terminated_length": 434.0500183105469, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.10822342901474011, "grad_norm": 0.6421154141426086, "kl": 0.184326171875, "learning_rate": 4.9293121125394205e-06, "loss": 0.0103, "num_tokens": 24545342.0, "reward": 0.3225000202655792, "reward_std": 0.13630501925945282, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2224999964237213, "rewards/unicoder_reward_fn/std": 0.29850897192955017, "step": 279 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 847.0, "completions/max_terminated_length": 847.0, "completions/mean_length": 419.1499938964844, "completions/mean_terminated_length": 419.1499938964844, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.10861132660977502, "grad_norm": 0.6725829243659973, "kl": 0.178466796875, "learning_rate": 4.928607237918551e-06, "loss": -0.0031, "num_tokens": 24625914.0, "reward": 0.23000001907348633, "reward_std": 0.13525645434856415, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.23621487617492676, "step": 280 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 693.0, "completions/max_terminated_length": 693.0, "completions/mean_length": 383.2124938964844, "completions/mean_terminated_length": 383.2124938964844, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.10899922420480992, "grad_norm": 0.690222442150116, "kl": 0.22509765625, "learning_rate": 4.927898922979481e-06, "loss": 0.0041, "num_tokens": 24703779.0, "reward": 0.3187500238418579, "reward_std": 0.09508258104324341, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.2199999988079071, "rewards/unicoder_reward_fn/std": 0.2748532295227051, "step": 281 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 738.0, "completions/max_terminated_length": 738.0, "completions/mean_length": 365.13751220703125, "completions/mean_terminated_length": 365.13751220703125, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.10938712179984485, "grad_norm": 0.8612980246543884, "kl": 0.223388671875, "learning_rate": 4.927187168840736e-06, "loss": 0.0358, "num_tokens": 24784390.0, "reward": 0.2549999952316284, "reward_std": 0.13024230301380157, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.2827979624271393, "step": 282 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 839.0, "completions/max_terminated_length": 839.0, "completions/mean_length": 412.7124938964844, "completions/mean_terminated_length": 412.7124938964844, "completions/min_length": 200.0, "completions/min_terminated_length": 200.0, "epoch": 0.10977501939487976, "grad_norm": 0.7258760929107666, "kl": 0.198486328125, "learning_rate": 4.9264719766262735e-06, "loss": -0.0095, "num_tokens": 24864911.0, "reward": 0.2550000250339508, "reward_std": 0.12335040420293808, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.21810461580753326, "step": 283 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 743.0, "completions/max_terminated_length": 743.0, "completions/mean_length": 445.125, "completions/mean_terminated_length": 445.125, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.11016291698991466, "grad_norm": 0.5923596620559692, "kl": 0.181640625, "learning_rate": 4.925753347465479e-06, "loss": 0.0395, "num_tokens": 24948321.0, "reward": 0.21500001847743988, "reward_std": 0.095413438975811, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.19815605878829956, "step": 284 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1731.0, "completions/max_terminated_length": 1731.0, "completions/mean_length": 436.9875183105469, "completions/mean_terminated_length": 436.9875183105469, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.11055081458494957, "grad_norm": 0.7096765041351318, "kl": 0.165283203125, "learning_rate": 4.925031282493166e-06, "loss": -0.027, "num_tokens": 25035088.0, "reward": 0.2800000011920929, "reward_std": 0.16279065608978271, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.18000000715255737, "rewards/unicoder_reward_fn/std": 0.2654826045036316, "step": 285 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 660.0, "completions/max_terminated_length": 660.0, "completions/mean_length": 344.38751220703125, "completions/mean_terminated_length": 344.38751220703125, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.11093871217998448, "grad_norm": 0.7686842679977417, "kl": 0.18701171875, "learning_rate": 4.924305782849575e-06, "loss": 0.0399, "num_tokens": 25105463.0, "reward": 0.30750003457069397, "reward_std": 0.1592330038547516, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.2958681881427765, "step": 286 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 993.0, "completions/max_terminated_length": 993.0, "completions/mean_length": 396.6750183105469, "completions/mean_terminated_length": 396.6750183105469, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.11132660977501939, "grad_norm": 0.7269200086593628, "kl": 0.191162109375, "learning_rate": 4.923576849680369e-06, "loss": 0.044, "num_tokens": 25186277.0, "reward": 0.19750000536441803, "reward_std": 0.11964404582977295, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.20621666312217712, "step": 287 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 802.0, "completions/max_terminated_length": 802.0, "completions/mean_length": 396.3625183105469, "completions/mean_terminated_length": 396.3625183105469, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.11171450737005431, "grad_norm": 0.8135611414909363, "kl": 0.201416015625, "learning_rate": 4.92284448413663e-06, "loss": 0.0094, "num_tokens": 25261594.0, "reward": 0.29749998450279236, "reward_std": 0.17270749807357788, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19750002026557922, "rewards/unicoder_reward_fn/std": 0.2594907283782959, "step": 288 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 884.0, "completions/max_terminated_length": 884.0, "completions/mean_length": 408.76251220703125, "completions/mean_terminated_length": 408.76251220703125, "completions/min_length": 231.0, "completions/min_terminated_length": 231.0, "epoch": 0.11210240496508922, "grad_norm": 0.8864731192588806, "kl": 0.19775390625, "learning_rate": 4.922108687374866e-06, "loss": 0.0244, "num_tokens": 25348239.0, "reward": 0.24000000953674316, "reward_std": 0.14268063008785248, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13999998569488525, "rewards/unicoder_reward_fn/std": 0.19974668323993683, "step": 289 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 606.0, "completions/max_terminated_length": 606.0, "completions/mean_length": 360.6625061035156, "completions/mean_terminated_length": 360.6625061035156, "completions/min_length": 193.0, "completions/min_terminated_length": 193.0, "epoch": 0.11249030256012413, "grad_norm": 0.6646974086761475, "kl": 0.179443359375, "learning_rate": 4.921369460557001e-06, "loss": -0.0022, "num_tokens": 25426100.0, "reward": 0.33750003576278687, "reward_std": 0.11884617060422897, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23749999701976776, "rewards/unicoder_reward_fn/std": 0.31517526507377625, "step": 290 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1186.0, "completions/max_terminated_length": 1186.0, "completions/mean_length": 425.6125183105469, "completions/mean_terminated_length": 425.6125183105469, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "epoch": 0.11287820015515904, "grad_norm": 0.7539274096488953, "kl": 0.184814453125, "learning_rate": 4.920626804850372e-06, "loss": -0.0269, "num_tokens": 25507581.0, "reward": 0.2200000137090683, "reward_std": 0.17053528130054474, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.2263636738061905, "step": 291 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 868.0, "completions/max_terminated_length": 868.0, "completions/mean_length": 430.8000183105469, "completions/mean_terminated_length": 430.8000183105469, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.11326609775019394, "grad_norm": 0.7934711575508118, "kl": 0.1884765625, "learning_rate": 4.9198807214277355e-06, "loss": 0.0095, "num_tokens": 25593669.0, "reward": 0.23000001907348633, "reward_std": 0.14744237065315247, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.21839460730552673, "step": 292 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1891.0, "completions/max_terminated_length": 1891.0, "completions/mean_length": 463.13751220703125, "completions/mean_terminated_length": 463.13751220703125, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.11365399534522885, "grad_norm": 0.7781848907470703, "kl": 0.160400390625, "learning_rate": 4.91913121146726e-06, "loss": -0.0076, "num_tokens": 25670016.0, "reward": 0.29499998688697815, "reward_std": 0.15619264543056488, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19500000774860382, "rewards/unicoder_reward_fn/std": 0.32135000824928284, "step": 293 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 853.0, "completions/max_terminated_length": 853.0, "completions/mean_length": 485.4125061035156, "completions/mean_terminated_length": 485.4125061035156, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.11404189294026378, "grad_norm": 0.577975869178772, "kl": 0.178466796875, "learning_rate": 4.9183782761525224e-06, "loss": 0.0009, "num_tokens": 25758625.0, "reward": 0.17749999463558197, "reward_std": 0.06292319297790527, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.14753243327140808, "step": 294 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 894.0, "completions/max_terminated_length": 894.0, "completions/mean_length": 477.7124938964844, "completions/mean_terminated_length": 477.7124938964844, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.11442979053529868, "grad_norm": 0.5966387987136841, "kl": 0.19189453125, "learning_rate": 4.917621916672511e-06, "loss": 0.0194, "num_tokens": 25857530.0, "reward": 0.15250001847743988, "reward_std": 0.06025980785489082, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.05250000208616257, "rewards/unicoder_reward_fn/std": 0.1136004626750946, "step": 295 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 828.0, "completions/max_terminated_length": 828.0, "completions/mean_length": 424.63751220703125, "completions/mean_terminated_length": 424.63751220703125, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.11481768813033359, "grad_norm": 0.7122818827629089, "kl": 0.1875, "learning_rate": 4.916862134221621e-06, "loss": 0.0211, "num_tokens": 25940077.0, "reward": 0.27000004053115845, "reward_std": 0.13234387338161469, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.2527819871902466, "step": 296 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 961.0, "completions/max_terminated_length": 961.0, "completions/mean_length": 473.3000183105469, "completions/mean_terminated_length": 473.3000183105469, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.1152055857253685, "grad_norm": 0.7102165222167969, "kl": 0.236083984375, "learning_rate": 4.916098929999652e-06, "loss": -0.0078, "num_tokens": 26031437.0, "reward": 0.36000001430511475, "reward_std": 0.11608979851007462, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2600000202655792, "rewards/unicoder_reward_fn/std": 0.3188587427139282, "step": 297 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 898.0, "completions/max_terminated_length": 898.0, "completions/mean_length": 461.6875, "completions/mean_terminated_length": 461.6875, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "epoch": 0.11559348332040341, "grad_norm": 0.6243550777435303, "kl": 0.200439453125, "learning_rate": 4.915332305211808e-06, "loss": 0.0183, "num_tokens": 26121708.0, "reward": 0.18999998271465302, "reward_std": 0.07856175303459167, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09000001102685928, "rewards/unicoder_reward_fn/std": 0.1710725724697113, "step": 298 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 740.0, "completions/max_terminated_length": 740.0, "completions/mean_length": 407.8500061035156, "completions/mean_terminated_length": 407.8500061035156, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.11598138091543832, "grad_norm": 0.7579537034034729, "kl": 0.175537109375, "learning_rate": 4.914562261068693e-06, "loss": -0.0098, "num_tokens": 26191408.0, "reward": 0.1900000125169754, "reward_std": 0.08606418967247009, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09000000357627869, "rewards/unicoder_reward_fn/std": 0.1454804390668869, "step": 299 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 878.0, "completions/max_terminated_length": 878.0, "completions/mean_length": 424.5874938964844, "completions/mean_terminated_length": 424.5874938964844, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "epoch": 0.11636927851047324, "grad_norm": 0.6652755737304688, "kl": 0.176513671875, "learning_rate": 4.913788798786315e-06, "loss": 0.0004, "num_tokens": 26268855.0, "reward": 0.1850000023841858, "reward_std": 0.1191520020365715, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.18493926525115967, "step": 300 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1244.0, "completions/max_terminated_length": 1244.0, "completions/mean_length": 486.7124938964844, "completions/mean_terminated_length": 486.7124938964844, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.11675717610550815, "grad_norm": 0.7912869453430176, "kl": 0.17724609375, "learning_rate": 4.913011919586073e-06, "loss": 0.0371, "num_tokens": 26359008.0, "reward": 0.2600000202655792, "reward_std": 0.12756338715553284, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1599999964237213, "rewards/unicoder_reward_fn/std": 0.22591587901115417, "step": 301 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 771.0, "completions/max_terminated_length": 771.0, "completions/mean_length": 411.125, "completions/mean_terminated_length": 411.125, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.11714507370054306, "grad_norm": 0.9114518761634827, "kl": 0.188232421875, "learning_rate": 4.912231624694767e-06, "loss": 0.007, "num_tokens": 26439290.0, "reward": 0.45000001788139343, "reward_std": 0.15913045406341553, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3500000238418579, "rewards/unicoder_reward_fn/std": 0.37585344910621643, "step": 302 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1153.0, "completions/max_terminated_length": 1153.0, "completions/mean_length": 532.375, "completions/mean_terminated_length": 532.375, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.11753297129557796, "grad_norm": 0.6090479493141174, "kl": 0.173828125, "learning_rate": 4.911447915344589e-06, "loss": 0.0192, "num_tokens": 26529544.0, "reward": 0.15125000476837158, "reward_std": 0.0823635458946228, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.12218993902206421, "step": 303 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1385.0, "completions/max_terminated_length": 1385.0, "completions/mean_length": 521.3500366210938, "completions/mean_terminated_length": 521.3500366210938, "completions/min_length": 242.0, "completions/min_terminated_length": 242.0, "epoch": 0.11792086889061287, "grad_norm": 0.6708824634552002, "kl": 0.1572265625, "learning_rate": 4.910660792773122e-06, "loss": 0.03, "num_tokens": 26615028.0, "reward": 0.23250000178813934, "reward_std": 0.10728316754102707, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.19858677685260773, "step": 304 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 924.0, "completions/max_terminated_length": 924.0, "completions/mean_length": 492.0500183105469, "completions/mean_terminated_length": 492.0500183105469, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.1183087664856478, "grad_norm": 0.6978269815444946, "kl": 0.169677734375, "learning_rate": 4.909870258223342e-06, "loss": 0.0212, "num_tokens": 26697736.0, "reward": 0.26250001788139343, "reward_std": 0.12292223423719406, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16250000894069672, "rewards/unicoder_reward_fn/std": 0.21369268000125885, "step": 305 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1128.0, "completions/max_terminated_length": 1128.0, "completions/mean_length": 508.1000061035156, "completions/mean_terminated_length": 508.1000061035156, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.1186966640806827, "grad_norm": 0.8212612867355347, "kl": 0.188232421875, "learning_rate": 4.909076312943609e-06, "loss": -0.0007, "num_tokens": 26787216.0, "reward": 0.32500001788139343, "reward_std": 0.16279709339141846, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.22499999403953552, "rewards/unicoder_reward_fn/std": 0.32040321826934814, "step": 306 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 949.0, "completions/max_terminated_length": 949.0, "completions/mean_length": 432.8999938964844, "completions/mean_terminated_length": 432.8999938964844, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "epoch": 0.11908456167571761, "grad_norm": 0.8716068863868713, "kl": 0.2099609375, "learning_rate": 4.908278958187673e-06, "loss": 0.018, "num_tokens": 26861200.0, "reward": 0.375, "reward_std": 0.204799085855484, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.33733880519866943, "step": 307 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 910.0, "completions/max_terminated_length": 910.0, "completions/mean_length": 459.2375183105469, "completions/mean_terminated_length": 459.2375183105469, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.11947245927075252, "grad_norm": 0.6713731288909912, "kl": 0.172119140625, "learning_rate": 4.907478195214664e-06, "loss": 0.0291, "num_tokens": 26936715.0, "reward": 0.2212500125169754, "reward_std": 0.0804794579744339, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.12250000238418579, "rewards/unicoder_reward_fn/std": 0.263364315032959, "step": 308 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1075.0, "completions/max_terminated_length": 1075.0, "completions/mean_length": 552.5750122070312, "completions/mean_terminated_length": 552.5750122070312, "completions/min_length": 276.0, "completions/min_terminated_length": 276.0, "epoch": 0.11986035686578743, "grad_norm": 0.5698988437652588, "kl": 0.167236328125, "learning_rate": 4.906674025289096e-06, "loss": 0.0036, "num_tokens": 27028233.0, "reward": 0.1925000101327896, "reward_std": 0.10089719295501709, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.2156972885131836, "step": 309 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1016.0, "completions/max_terminated_length": 1016.0, "completions/mean_length": 542.4874877929688, "completions/mean_terminated_length": 542.4874877929688, "completions/min_length": 302.0, "completions/min_terminated_length": 302.0, "epoch": 0.12024825446082234, "grad_norm": 0.5580446720123291, "kl": 0.165771484375, "learning_rate": 4.905866449680864e-06, "loss": 0.0138, "num_tokens": 27117168.0, "reward": 0.1850000023841858, "reward_std": 0.08186791092157364, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.17938010394573212, "step": 310 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 895.0, "completions/max_terminated_length": 895.0, "completions/mean_length": 525.2000122070312, "completions/mean_terminated_length": 525.2000122070312, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.12063615205585726, "grad_norm": 0.5431578755378723, "kl": 0.16650390625, "learning_rate": 4.905055469665241e-06, "loss": 0.0106, "num_tokens": 27205760.0, "reward": 0.26750001311302185, "reward_std": 0.10238859802484512, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16750000417232513, "rewards/unicoder_reward_fn/std": 0.3125132918357849, "step": 311 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1095.0, "completions/max_terminated_length": 1095.0, "completions/mean_length": 514.5625, "completions/mean_terminated_length": 514.5625, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.12102404965089217, "grad_norm": 0.7057257890701294, "kl": 0.20068359375, "learning_rate": 4.904241086522874e-06, "loss": -0.0224, "num_tokens": 27291789.0, "reward": 0.3575000464916229, "reward_std": 0.12466736137866974, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2574999928474426, "rewards/unicoder_reward_fn/std": 0.30388832092285156, "step": 312 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1001.0, "completions/max_terminated_length": 1001.0, "completions/mean_length": 529.8500366210938, "completions/mean_terminated_length": 529.8500366210938, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.12141194724592708, "grad_norm": 0.743370532989502, "kl": 0.182373046875, "learning_rate": 4.903423301539786e-06, "loss": 0.0007, "num_tokens": 27389657.0, "reward": 0.26250001788139343, "reward_std": 0.17457836866378784, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.16499999165534973, "rewards/unicoder_reward_fn/std": 0.32573190331459045, "step": 313 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 864.0, "completions/max_terminated_length": 864.0, "completions/mean_length": 473.4624938964844, "completions/mean_terminated_length": 473.4624938964844, "completions/min_length": 251.0, "completions/min_terminated_length": 251.0, "epoch": 0.12179984484096198, "grad_norm": 0.610395073890686, "kl": 0.1923828125, "learning_rate": 4.902602116007369e-06, "loss": 0.0045, "num_tokens": 27474646.0, "reward": 0.3225000202655792, "reward_std": 0.15795986354351044, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2224999964237213, "rewards/unicoder_reward_fn/std": 0.3306973874568939, "step": 314 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1044.0, "completions/max_terminated_length": 1044.0, "completions/mean_length": 525.7000122070312, "completions/mean_terminated_length": 525.7000122070312, "completions/min_length": 291.0, "completions/min_terminated_length": 291.0, "epoch": 0.1221877424359969, "grad_norm": 0.6670093536376953, "kl": 0.17578125, "learning_rate": 4.9017775312223895e-06, "loss": 0.0041, "num_tokens": 27565902.0, "reward": 0.20000000298023224, "reward_std": 0.14966022968292236, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.25655949115753174, "step": 315 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1046.0, "completions/max_terminated_length": 1046.0, "completions/mean_length": 543.2999877929688, "completions/mean_terminated_length": 543.2999877929688, "completions/min_length": 291.0, "completions/min_terminated_length": 291.0, "epoch": 0.1225756400310318, "grad_norm": 0.6759949922561646, "kl": 0.1875, "learning_rate": 4.900949548486977e-06, "loss": 0.0369, "num_tokens": 27669030.0, "reward": 0.21124999225139618, "reward_std": 0.08920171111822128, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.1648743450641632, "step": 316 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1011.0, "completions/max_terminated_length": 1011.0, "completions/mean_length": 459.1499938964844, "completions/mean_terminated_length": 459.1499938964844, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.12296353762606672, "grad_norm": 0.7631900906562805, "kl": 0.211669921875, "learning_rate": 4.900118169108632e-06, "loss": 0.0024, "num_tokens": 27761978.0, "reward": 0.2874999940395355, "reward_std": 0.11252539604902267, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.21190816164016724, "step": 317 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 836.0, "completions/max_terminated_length": 836.0, "completions/mean_length": 474.5, "completions/mean_terminated_length": 474.5, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "epoch": 0.12335143522110163, "grad_norm": 0.828331470489502, "kl": 0.197021484375, "learning_rate": 4.899283394400214e-06, "loss": 0.0108, "num_tokens": 27841930.0, "reward": 0.2750000059604645, "reward_std": 0.15740719437599182, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.23029975593090057, "step": 318 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1088.0, "completions/max_terminated_length": 1088.0, "completions/mean_length": 548.4750366210938, "completions/mean_terminated_length": 548.4750366210938, "completions/min_length": 251.0, "completions/min_terminated_length": 251.0, "epoch": 0.12373933281613654, "grad_norm": 0.6373922824859619, "kl": 0.180908203125, "learning_rate": 4.898445225679948e-06, "loss": 0.0196, "num_tokens": 27936536.0, "reward": 0.445000022649765, "reward_std": 0.1279619187116623, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3449999988079071, "rewards/unicoder_reward_fn/std": 0.36279505491256714, "step": 319 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1083.0, "completions/max_terminated_length": 1083.0, "completions/mean_length": 633.1749877929688, "completions/mean_terminated_length": 633.1749877929688, "completions/min_length": 308.0, "completions/min_terminated_length": 308.0, "epoch": 0.12412723041117145, "grad_norm": 0.5543761253356934, "kl": 0.167724609375, "learning_rate": 4.897603664271414e-06, "loss": -0.0155, "num_tokens": 28041446.0, "reward": 0.24249999225139618, "reward_std": 0.07661745697259903, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.24275581538677216, "step": 320 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1116.0, "completions/max_terminated_length": 1116.0, "completions/mean_length": 492.07501220703125, "completions/mean_terminated_length": 492.07501220703125, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.12451512800620636, "grad_norm": 0.7415323853492737, "kl": 0.184326171875, "learning_rate": 4.896758711503554e-06, "loss": -0.0026, "num_tokens": 28129308.0, "reward": 0.3137500286102295, "reward_std": 0.12077617645263672, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.2150000035762787, "rewards/unicoder_reward_fn/std": 0.2619788348674774, "step": 321 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1055.0, "completions/max_terminated_length": 1055.0, "completions/mean_length": 538.3500366210938, "completions/mean_terminated_length": 538.3500366210938, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "epoch": 0.12490302560124127, "grad_norm": 0.7193112969398499, "kl": 0.172607421875, "learning_rate": 4.895910368710665e-06, "loss": 0.0069, "num_tokens": 28216256.0, "reward": 0.29249998927116394, "reward_std": 0.13390961289405823, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1925000250339508, "rewards/unicoder_reward_fn/std": 0.2260138988494873, "step": 322 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1143.0, "completions/max_terminated_length": 1143.0, "completions/mean_length": 549.3875122070312, "completions/mean_terminated_length": 549.3875122070312, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.1252909231962762, "grad_norm": 0.5058150291442871, "kl": 0.171142578125, "learning_rate": 4.895058637232394e-06, "loss": 0.0301, "num_tokens": 28306087.0, "reward": 0.2800000309944153, "reward_std": 0.1495622992515564, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.3270117938518524, "step": 323 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 980.0, "completions/max_terminated_length": 980.0, "completions/mean_length": 549.0375366210938, "completions/mean_terminated_length": 549.0375366210938, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.1256788207913111, "grad_norm": 0.6184161305427551, "kl": 0.167236328125, "learning_rate": 4.894203518413743e-06, "loss": 0.0053, "num_tokens": 28398482.0, "reward": 0.20500002801418304, "reward_std": 0.13806647062301636, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.18819595873355865, "step": 324 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 932.0, "completions/max_terminated_length": 932.0, "completions/mean_length": 476.26251220703125, "completions/mean_terminated_length": 476.26251220703125, "completions/min_length": 238.0, "completions/min_terminated_length": 238.0, "epoch": 0.126066718386346, "grad_norm": 0.7511195540428162, "kl": 0.182861328125, "learning_rate": 4.89334501360506e-06, "loss": 0.0214, "num_tokens": 28485231.0, "reward": 0.29249998927116394, "reward_std": 0.15438444912433624, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1925000101327896, "rewards/unicoder_reward_fn/std": 0.2514640688896179, "step": 325 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 810.0, "completions/max_terminated_length": 810.0, "completions/mean_length": 512.2000122070312, "completions/mean_terminated_length": 512.2000122070312, "completions/min_length": 257.0, "completions/min_terminated_length": 257.0, "epoch": 0.1264546159813809, "grad_norm": 0.8195441961288452, "kl": 0.19580078125, "learning_rate": 4.892483124162041e-06, "loss": 0.007, "num_tokens": 28576751.0, "reward": 0.3775000274181366, "reward_std": 0.16343779861927032, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2774999737739563, "rewards/unicoder_reward_fn/std": 0.26718172430992126, "step": 326 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 959.0, "completions/max_terminated_length": 959.0, "completions/mean_length": 499.5249938964844, "completions/mean_terminated_length": 499.5249938964844, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.12684251357641582, "grad_norm": 0.6965732574462891, "kl": 0.180419921875, "learning_rate": 4.891617851445729e-06, "loss": 0.0138, "num_tokens": 28660097.0, "reward": 0.23624999821186066, "reward_std": 0.13310347497463226, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.22183071076869965, "step": 327 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1117.0, "completions/max_terminated_length": 1117.0, "completions/mean_length": 528.0250244140625, "completions/mean_terminated_length": 528.0250244140625, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "epoch": 0.12723041117145073, "grad_norm": 0.7396894693374634, "kl": 0.1943359375, "learning_rate": 4.890749196822506e-06, "loss": 0.0013, "num_tokens": 28750995.0, "reward": 0.24625001847743988, "reward_std": 0.10409671068191528, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.15506227314472198, "step": 328 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 920.0, "completions/max_terminated_length": 920.0, "completions/mean_length": 541.75, "completions/mean_terminated_length": 541.75, "completions/min_length": 330.0, "completions/min_terminated_length": 330.0, "epoch": 0.12761830876648564, "grad_norm": 0.5413792133331299, "kl": 0.186767578125, "learning_rate": 4.889877161664096e-06, "loss": 0.0068, "num_tokens": 28840471.0, "reward": 0.3125, "reward_std": 0.1025177389383316, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.21250000596046448, "rewards/unicoder_reward_fn/std": 0.2870076894760132, "step": 329 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 842.0, "completions/max_terminated_length": 842.0, "completions/mean_length": 540.3875122070312, "completions/mean_terminated_length": 540.3875122070312, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "epoch": 0.12800620636152055, "grad_norm": 0.7199426889419556, "kl": 0.163818359375, "learning_rate": 4.889001747347563e-06, "loss": -0.012, "num_tokens": 28933582.0, "reward": 0.3775000274181366, "reward_std": 0.21813993155956268, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2775000035762787, "rewards/unicoder_reward_fn/std": 0.31259429454803467, "step": 330 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1399.0, "completions/max_terminated_length": 1399.0, "completions/mean_length": 554.5125122070312, "completions/mean_terminated_length": 554.5125122070312, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "epoch": 0.12839410395655548, "grad_norm": 0.5968661308288574, "kl": 0.189208984375, "learning_rate": 4.888122955255304e-06, "loss": 0.0156, "num_tokens": 29030823.0, "reward": 0.3449999988079071, "reward_std": 0.11065230518579483, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24500000476837158, "rewards/unicoder_reward_fn/std": 0.2584765553474426, "step": 331 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1145.0, "completions/max_terminated_length": 1145.0, "completions/mean_length": 621.1375122070312, "completions/mean_terminated_length": 621.1375122070312, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "epoch": 0.1287820015515904, "grad_norm": 0.5512359738349915, "kl": 0.1865234375, "learning_rate": 4.8872407867750545e-06, "loss": 0.0227, "num_tokens": 29123306.0, "reward": 0.24500000476837158, "reward_std": 0.09602485597133636, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.24020035564899445, "step": 332 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 849.0, "completions/max_terminated_length": 849.0, "completions/mean_length": 457.7875061035156, "completions/mean_terminated_length": 457.7875061035156, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.1291698991466253, "grad_norm": 0.7473779916763306, "kl": 0.196533203125, "learning_rate": 4.886355243299876e-06, "loss": 0.0036, "num_tokens": 29205545.0, "reward": 0.5175000429153442, "reward_std": 0.17388784885406494, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.41750001907348633, "rewards/unicoder_reward_fn/std": 0.3781282901763916, "step": 333 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1192.0, "completions/max_terminated_length": 1192.0, "completions/mean_length": 561.5125122070312, "completions/mean_terminated_length": 561.5125122070312, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.1295577967416602, "grad_norm": 0.6978445649147034, "kl": 0.180419921875, "learning_rate": 4.885466326228163e-06, "loss": 0.0039, "num_tokens": 29301594.0, "reward": 0.2950000464916229, "reward_std": 0.14928975701332092, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19500000774860382, "rewards/unicoder_reward_fn/std": 0.26428794860839844, "step": 334 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1309.0, "completions/max_terminated_length": 1309.0, "completions/mean_length": 499.0249938964844, "completions/mean_terminated_length": 499.0249938964844, "completions/min_length": 220.0, "completions/min_terminated_length": 220.0, "epoch": 0.12994569433669512, "grad_norm": 0.7044440507888794, "kl": 0.191162109375, "learning_rate": 4.884574036963639e-06, "loss": 0.0189, "num_tokens": 29381548.0, "reward": 0.42750000953674316, "reward_std": 0.1785697489976883, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.32750001549720764, "rewards/unicoder_reward_fn/std": 0.37144091725349426, "step": 335 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2813.0, "completions/max_terminated_length": 2813.0, "completions/mean_length": 555.2374877929688, "completions/mean_terminated_length": 555.2374877929688, "completions/min_length": 239.0, "completions/min_terminated_length": 239.0, "epoch": 0.13033359193173003, "grad_norm": 0.7157901525497437, "kl": 0.2021484375, "learning_rate": 4.883678376915348e-06, "loss": 0.0044, "num_tokens": 29471911.0, "reward": 0.26250001788139343, "reward_std": 0.14991536736488342, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16250000894069672, "rewards/unicoder_reward_fn/std": 0.29867851734161377, "step": 336 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1066.0, "completions/max_terminated_length": 1066.0, "completions/mean_length": 488.38751220703125, "completions/mean_terminated_length": 488.38751220703125, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.13072148952676493, "grad_norm": 0.7547192573547363, "kl": 0.192138671875, "learning_rate": 4.8827793474976605e-06, "loss": 0.0094, "num_tokens": 29555502.0, "reward": 0.5050000548362732, "reward_std": 0.20460812747478485, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4049999713897705, "rewards/unicoder_reward_fn/std": 0.3961523771286011, "step": 337 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 938.0, "completions/max_terminated_length": 938.0, "completions/mean_length": 566.7625122070312, "completions/mean_terminated_length": 566.7625122070312, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "epoch": 0.13110938712179984, "grad_norm": 0.5531723499298096, "kl": 0.158935546875, "learning_rate": 4.881876950130266e-06, "loss": 0.0147, "num_tokens": 29653067.0, "reward": 0.24500000476837158, "reward_std": 0.09469256550073624, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14499999582767487, "rewards/unicoder_reward_fn/std": 0.24644306302070618, "step": 338 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 885.0, "completions/mean_length": 560.6375122070312, "completions/mean_terminated_length": 515.8861083984375, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "epoch": 0.13149728471683475, "grad_norm": 0.5851719975471497, "kl": 0.18017578125, "learning_rate": 4.880971186238175e-06, "loss": 0.0742, "num_tokens": 29745758.0, "reward": 0.18375001847743988, "reward_std": 0.08441993594169617, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.14151082932949066, "step": 339 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1382.0, "completions/max_terminated_length": 1382.0, "completions/mean_length": 515.2875366210938, "completions/mean_terminated_length": 515.2875366210938, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.13188518231186966, "grad_norm": 0.7348682880401611, "kl": 0.16650390625, "learning_rate": 4.880062057251711e-06, "loss": -0.0027, "num_tokens": 29838653.0, "reward": 0.23750002682209015, "reward_std": 0.14302581548690796, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.2027500867843628, "step": 340 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1262.0, "completions/max_terminated_length": 1262.0, "completions/mean_length": 510.3125, "completions/mean_terminated_length": 510.3125, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "epoch": 0.13227307990690457, "grad_norm": 0.6058952212333679, "kl": 0.18896484375, "learning_rate": 4.879149564606512e-06, "loss": 0.0075, "num_tokens": 29927310.0, "reward": 0.33750003576278687, "reward_std": 0.11682651191949844, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23749999701976776, "rewards/unicoder_reward_fn/std": 0.321537047624588, "step": 341 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 937.0, "completions/max_terminated_length": 937.0, "completions/mean_length": 494.6125183105469, "completions/mean_terminated_length": 494.6125183105469, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.13266097750193948, "grad_norm": 0.7950134873390198, "kl": 0.185546875, "learning_rate": 4.878233709743531e-06, "loss": 0.0434, "num_tokens": 30024847.0, "reward": 0.38499999046325684, "reward_std": 0.20879927277565002, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2850000262260437, "rewards/unicoder_reward_fn/std": 0.3540363907814026, "step": 342 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1126.0, "completions/max_terminated_length": 1126.0, "completions/mean_length": 542.7625122070312, "completions/mean_terminated_length": 542.7625122070312, "completions/min_length": 264.0, "completions/min_terminated_length": 264.0, "epoch": 0.1330488750969744, "grad_norm": 0.6898848414421082, "kl": 0.1787109375, "learning_rate": 4.877314494109024e-06, "loss": -0.0288, "num_tokens": 30113228.0, "reward": 0.30750003457069397, "reward_std": 0.1747378557920456, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.2632681727409363, "step": 343 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1139.0, "completions/max_terminated_length": 1139.0, "completions/mean_length": 464.2749938964844, "completions/mean_terminated_length": 464.2749938964844, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.13343677269200932, "grad_norm": 0.7542030811309814, "kl": 0.192626953125, "learning_rate": 4.876391919154561e-06, "loss": -0.022, "num_tokens": 30200506.0, "reward": 0.38625001907348633, "reward_std": 0.17524342238903046, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.38953882455825806, "step": 344 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1004.0, "completions/max_terminated_length": 1004.0, "completions/mean_length": 522.8125, "completions/mean_terminated_length": 522.8125, "completions/min_length": 233.0, "completions/min_terminated_length": 233.0, "epoch": 0.13382467028704423, "grad_norm": 0.800554096698761, "kl": 0.175048828125, "learning_rate": 4.875465986337013e-06, "loss": -0.0113, "num_tokens": 30284243.0, "reward": 0.30625003576278687, "reward_std": 0.1450517624616623, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.23480400443077087, "step": 345 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1168.0, "completions/max_terminated_length": 1168.0, "completions/mean_length": 575.1875, "completions/mean_terminated_length": 575.1875, "completions/min_length": 303.0, "completions/min_terminated_length": 303.0, "epoch": 0.13421256788207914, "grad_norm": 0.5960347056388855, "kl": 0.17333984375, "learning_rate": 4.874536697118554e-06, "loss": 0.0211, "num_tokens": 30380786.0, "reward": 0.2950000464916229, "reward_std": 0.14862696826457977, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19500000774860382, "rewards/unicoder_reward_fn/std": 0.3051872253417969, "step": 346 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1025.0, "completions/max_terminated_length": 1025.0, "completions/mean_length": 537.5, "completions/mean_terminated_length": 537.5, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "epoch": 0.13460046547711405, "grad_norm": 0.7391052842140198, "kl": 0.183837890625, "learning_rate": 4.8736040529666595e-06, "loss": 0.015, "num_tokens": 30478778.0, "reward": 0.2750000059604645, "reward_std": 0.16175387799739838, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.2431335598230362, "step": 347 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1323.0, "completions/max_terminated_length": 1323.0, "completions/mean_length": 554.7999877929688, "completions/mean_terminated_length": 554.7999877929688, "completions/min_length": 266.0, "completions/min_terminated_length": 266.0, "epoch": 0.13498836307214895, "grad_norm": 0.6457734107971191, "kl": 0.1591796875, "learning_rate": 4.8726680553541e-06, "loss": 0.0051, "num_tokens": 30570186.0, "reward": 0.3225000500679016, "reward_std": 0.12204251438379288, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2225000113248825, "rewards/unicoder_reward_fn/std": 0.20621667802333832, "step": 348 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 962.0, "completions/max_terminated_length": 962.0, "completions/mean_length": 602.3250122070312, "completions/mean_terminated_length": 602.3250122070312, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.13537626066718386, "grad_norm": 0.6298027038574219, "kl": 0.1728515625, "learning_rate": 4.8717287057589455e-06, "loss": -0.0273, "num_tokens": 30657420.0, "reward": 0.2224999964237213, "reward_std": 0.0700276717543602, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12250000238418579, "rewards/unicoder_reward_fn/std": 0.15424376726150513, "step": 349 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1083.0, "completions/max_terminated_length": 1083.0, "completions/mean_length": 585.7625122070312, "completions/mean_terminated_length": 585.7625122070312, "completions/min_length": 301.0, "completions/min_terminated_length": 301.0, "epoch": 0.13576415826221877, "grad_norm": 0.6849412322044373, "kl": 0.16650390625, "learning_rate": 4.870786005664556e-06, "loss": 0.0373, "num_tokens": 30753729.0, "reward": 0.24500000476837158, "reward_std": 0.1126268282532692, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14499999582767487, "rewards/unicoder_reward_fn/std": 0.17129439115524292, "step": 350 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 886.0, "completions/max_terminated_length": 886.0, "completions/mean_length": 528.6625366210938, "completions/mean_terminated_length": 528.6625366210938, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "epoch": 0.13615205585725368, "grad_norm": 0.7189430594444275, "kl": 0.171630859375, "learning_rate": 4.869839956559584e-06, "loss": 0.0196, "num_tokens": 30842038.0, "reward": 0.33500000834465027, "reward_std": 0.14260141551494598, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23499998450279236, "rewards/unicoder_reward_fn/std": 0.23977836966514587, "step": 351 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1349.0, "completions/max_terminated_length": 1349.0, "completions/mean_length": 598.3250122070312, "completions/mean_terminated_length": 598.3250122070312, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.1365399534522886, "grad_norm": 0.6426841616630554, "kl": 0.1787109375, "learning_rate": 4.86889055993797e-06, "loss": -0.0364, "num_tokens": 30928832.0, "reward": 0.38750001788139343, "reward_std": 0.15756569802761078, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.30163058638572693, "step": 352 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1091.0, "completions/max_terminated_length": 1091.0, "completions/mean_length": 540.9375, "completions/mean_terminated_length": 540.9375, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "epoch": 0.1369278510473235, "grad_norm": 0.7209288477897644, "kl": 0.16748046875, "learning_rate": 4.867937817298941e-06, "loss": 0.0115, "num_tokens": 31018051.0, "reward": 0.3962499797344208, "reward_std": 0.19572162628173828, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.29750001430511475, "rewards/unicoder_reward_fn/std": 0.32762089371681213, "step": 353 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 882.0, "completions/max_terminated_length": 882.0, "completions/mean_length": 489.70001220703125, "completions/mean_terminated_length": 489.70001220703125, "completions/min_length": 243.0, "completions/min_terminated_length": 243.0, "epoch": 0.1373157486423584, "grad_norm": 0.7465131282806396, "kl": 0.179931640625, "learning_rate": 4.866981730147008e-06, "loss": -0.0473, "num_tokens": 31105923.0, "reward": 0.3750000298023224, "reward_std": 0.1574910581111908, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.29915493726730347, "step": 354 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1325.0, "completions/max_terminated_length": 1325.0, "completions/mean_length": 613.2625122070312, "completions/mean_terminated_length": 613.2625122070312, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.13770364623739334, "grad_norm": 0.748731255531311, "kl": 0.18212890625, "learning_rate": 4.866022299991962e-06, "loss": 0.0034, "num_tokens": 31206296.0, "reward": 0.3125, "reward_std": 0.12432298809289932, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.21250000596046448, "rewards/unicoder_reward_fn/std": 0.23673677444458008, "step": 355 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1246.0, "completions/max_terminated_length": 1246.0, "completions/mean_length": 558.0750122070312, "completions/mean_terminated_length": 558.0750122070312, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.13809154383242825, "grad_norm": 0.7337028980255127, "kl": 0.197509765625, "learning_rate": 4.865059528348876e-06, "loss": -0.0129, "num_tokens": 31303654.0, "reward": 0.2850000262260437, "reward_std": 0.1251099854707718, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.18500001728534698, "rewards/unicoder_reward_fn/std": 0.27883121371269226, "step": 356 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1393.0, "completions/max_terminated_length": 1393.0, "completions/mean_length": 579.6124877929688, "completions/mean_terminated_length": 579.6124877929688, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "epoch": 0.13847944142746316, "grad_norm": 0.6421045064926147, "kl": 0.157958984375, "learning_rate": 4.864093416738096e-06, "loss": 0.0215, "num_tokens": 31397511.0, "reward": 0.3500000536441803, "reward_std": 0.17458824813365936, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.25, "rewards/unicoder_reward_fn/std": 0.2907868027687073, "step": 357 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1618.0, "completions/mean_length": 664.1875, "completions/mean_terminated_length": 620.746826171875, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "epoch": 0.13886733902249807, "grad_norm": 0.5165175199508667, "kl": 0.140869140625, "learning_rate": 4.863123966685245e-06, "loss": 0.096, "num_tokens": 31496814.0, "reward": 0.4512500464916229, "reward_std": 0.15494577586650848, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.3525000214576721, "rewards/unicoder_reward_fn/std": 0.36316993832588196, "step": 358 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1258.0, "completions/mean_length": 650.3624877929688, "completions/mean_terminated_length": 606.746826171875, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "epoch": 0.13925523661753297, "grad_norm": 0.5616593956947327, "kl": 0.15478515625, "learning_rate": 4.862151179721217e-06, "loss": 0.0292, "num_tokens": 31604467.0, "reward": 0.2487500160932541, "reward_std": 0.12111931294202805, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.21228115260601044, "step": 359 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1011.0, "completions/max_terminated_length": 1011.0, "completions/mean_length": 506.1750183105469, "completions/mean_terminated_length": 506.1750183105469, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.13964313421256788, "grad_norm": 0.7833155393600464, "kl": 0.168701171875, "learning_rate": 4.861175057382175e-06, "loss": -0.0686, "num_tokens": 31686785.0, "reward": 0.29249998927116394, "reward_std": 0.11283562332391739, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.23263761401176453, "step": 360 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2120.0, "completions/max_terminated_length": 2120.0, "completions/mean_length": 722.3250122070312, "completions/mean_terminated_length": 722.3250122070312, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "epoch": 0.1400310318076028, "grad_norm": 0.6429398655891418, "kl": 0.15185546875, "learning_rate": 4.86019560120955e-06, "loss": 0.0198, "num_tokens": 31785291.0, "reward": 0.3225000202655792, "reward_std": 0.15830913186073303, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2225000113248825, "rewards/unicoder_reward_fn/std": 0.2565471827983856, "step": 361 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1282.0, "completions/max_terminated_length": 1282.0, "completions/mean_length": 654.2999877929688, "completions/mean_terminated_length": 654.2999877929688, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.1404189294026377, "grad_norm": 0.5455026626586914, "kl": 0.154052734375, "learning_rate": 4.859212812750037e-06, "loss": 0.0278, "num_tokens": 31884995.0, "reward": 0.3125000298023224, "reward_std": 0.16271665692329407, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.21250000596046448, "rewards/unicoder_reward_fn/std": 0.30246874690055847, "step": 362 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1413.0, "completions/max_terminated_length": 1413.0, "completions/mean_length": 547.6625366210938, "completions/mean_terminated_length": 547.6625366210938, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.1408068269976726, "grad_norm": 0.6433162689208984, "kl": 0.198974609375, "learning_rate": 4.858226693555592e-06, "loss": 0.0619, "num_tokens": 31977384.0, "reward": 0.24249999225139618, "reward_std": 0.14983738958835602, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14250001311302185, "rewards/unicoder_reward_fn/std": 0.21154944598674774, "step": 363 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1370.0, "completions/max_terminated_length": 1370.0, "completions/mean_length": 619.2250366210938, "completions/mean_terminated_length": 619.2250366210938, "completions/min_length": 259.0, "completions/min_terminated_length": 259.0, "epoch": 0.14119472459270752, "grad_norm": 0.5190523266792297, "kl": 0.14013671875, "learning_rate": 4.857237245183434e-06, "loss": 0.014, "num_tokens": 32069394.0, "reward": 0.25999999046325684, "reward_std": 0.1460186243057251, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1600000113248825, "rewards/unicoder_reward_fn/std": 0.23254236578941345, "step": 364 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1297.0, "completions/max_terminated_length": 1297.0, "completions/mean_length": 529.8125, "completions/mean_terminated_length": 529.8125, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.14158262218774242, "grad_norm": 0.600216805934906, "kl": 0.176025390625, "learning_rate": 4.856244469196034e-06, "loss": 0.0435, "num_tokens": 32160219.0, "reward": 0.3725000023841858, "reward_std": 0.1529356688261032, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2724999785423279, "rewards/unicoder_reward_fn/std": 0.3056327700614929, "step": 365 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2305.0, "completions/max_terminated_length": 2305.0, "completions/mean_length": 621.1749877929688, "completions/mean_terminated_length": 621.1749877929688, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.14197051978277736, "grad_norm": 0.6332196593284607, "kl": 0.20263671875, "learning_rate": 4.855248367161123e-06, "loss": -0.0415, "num_tokens": 32263785.0, "reward": 0.35999998450279236, "reward_std": 0.11800310760736465, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2600000202655792, "rewards/unicoder_reward_fn/std": 0.2853379249572754, "step": 366 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1935.0, "completions/max_terminated_length": 1935.0, "completions/mean_length": 543.5125122070312, "completions/mean_terminated_length": 543.5125122070312, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "epoch": 0.14235841737781227, "grad_norm": 0.6577582955360413, "kl": 0.1865234375, "learning_rate": 4.854248940651681e-06, "loss": -0.0004, "num_tokens": 32361562.0, "reward": 0.30000001192092896, "reward_std": 0.13461732864379883, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.20623202621936798, "step": 367 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2965.0, "completions/max_terminated_length": 2965.0, "completions/mean_length": 615.2000122070312, "completions/mean_terminated_length": 615.2000122070312, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.14274631497284718, "grad_norm": 0.7452603578567505, "kl": 0.212158203125, "learning_rate": 4.85324619124594e-06, "loss": -0.0095, "num_tokens": 32459618.0, "reward": 0.2762500047683716, "reward_std": 0.08642581105232239, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.1855372041463852, "step": 368 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1474.0, "completions/max_terminated_length": 1474.0, "completions/mean_length": 453.5, "completions/mean_terminated_length": 453.5, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.14313421256788209, "grad_norm": 0.6633742451667786, "kl": 0.1865234375, "learning_rate": 4.852240120527378e-06, "loss": -0.0028, "num_tokens": 32536986.0, "reward": 0.3262500464916229, "reward_std": 0.0882900282740593, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.22750000655651093, "rewards/unicoder_reward_fn/std": 0.2787744402885437, "step": 369 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1283.0, "completions/max_terminated_length": 1283.0, "completions/mean_length": 591.7625122070312, "completions/mean_terminated_length": 591.7625122070312, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "epoch": 0.143522110162917, "grad_norm": 0.7854828238487244, "kl": 0.1708984375, "learning_rate": 4.8512307300847165e-06, "loss": -0.0026, "num_tokens": 32632199.0, "reward": 0.4124999940395355, "reward_std": 0.18089398741722107, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3125000298023224, "rewards/unicoder_reward_fn/std": 0.2913847267627716, "step": 370 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1216.0, "completions/max_terminated_length": 1216.0, "completions/mean_length": 486.4624938964844, "completions/mean_terminated_length": 486.4624938964844, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.1439100077579519, "grad_norm": 0.6145583987236023, "kl": 0.182861328125, "learning_rate": 4.850218021511922e-06, "loss": 0.0378, "num_tokens": 32710900.0, "reward": 0.33250001072883606, "reward_std": 0.08690296858549118, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23250000178813934, "rewards/unicoder_reward_fn/std": 0.2048618048429489, "step": 371 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1578.0, "completions/max_terminated_length": 1578.0, "completions/mean_length": 547.0750122070312, "completions/mean_terminated_length": 547.0750122070312, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "epoch": 0.1442979053529868, "grad_norm": 0.6976794004440308, "kl": 0.23095703125, "learning_rate": 4.8492019964082e-06, "loss": -0.0086, "num_tokens": 32810994.0, "reward": 0.3375000059604645, "reward_std": 0.11441679298877716, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23749999701976776, "rewards/unicoder_reward_fn/std": 0.25674447417259216, "step": 372 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1250.0, "completions/max_terminated_length": 1250.0, "completions/mean_length": 523.3375244140625, "completions/mean_terminated_length": 523.3375244140625, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.14468580294802172, "grad_norm": 0.7697983384132385, "kl": 0.19580078125, "learning_rate": 4.848182656377991e-06, "loss": 0.0367, "num_tokens": 32902461.0, "reward": 0.2562499940395355, "reward_std": 0.1625344604253769, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.1874377727508545, "step": 373 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2536.0, "completions/max_terminated_length": 2536.0, "completions/mean_length": 609.0875244140625, "completions/mean_terminated_length": 609.0875244140625, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.14507370054305663, "grad_norm": 0.6462472677230835, "kl": 0.166259765625, "learning_rate": 4.847160003030975e-06, "loss": 0.0251, "num_tokens": 33007644.0, "reward": 0.2912500202655792, "reward_std": 0.15327030420303345, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.25545936822891235, "step": 374 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1506.0, "completions/max_terminated_length": 1506.0, "completions/mean_length": 592.1625366210938, "completions/mean_terminated_length": 592.1625366210938, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.14546159813809154, "grad_norm": 0.6295875310897827, "kl": 0.202880859375, "learning_rate": 4.846134037982059e-06, "loss": -0.0137, "num_tokens": 33100737.0, "reward": 0.3225000202655792, "reward_std": 0.11245019733905792, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2224999964237213, "rewards/unicoder_reward_fn/std": 0.2110702097415924, "step": 375 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1758.0, "completions/max_terminated_length": 1758.0, "completions/mean_length": 731.0375366210938, "completions/mean_terminated_length": 731.0375366210938, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.14584949573312644, "grad_norm": 0.4804399907588959, "kl": 0.1689453125, "learning_rate": 4.845104762851382e-06, "loss": 0.0375, "num_tokens": 33208116.0, "reward": 0.2500000298023224, "reward_std": 0.07161671668291092, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.24339371919631958, "step": 376 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3874.0, "completions/max_terminated_length": 3874.0, "completions/mean_length": 645.2000122070312, "completions/mean_terminated_length": 645.2000122070312, "completions/min_length": 67.0, "completions/min_terminated_length": 67.0, "epoch": 0.14623739332816135, "grad_norm": 0.6313014626502991, "kl": 0.205810546875, "learning_rate": 4.844072179264311e-06, "loss": -0.0324, "num_tokens": 33298500.0, "reward": 0.48000001907348633, "reward_std": 0.16493050754070282, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3800000250339508, "rewards/unicoder_reward_fn/std": 0.2517784833908081, "step": 377 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1093.0, "completions/max_terminated_length": 1093.0, "completions/mean_length": 632.75, "completions/mean_terminated_length": 632.75, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.1466252909231963, "grad_norm": 0.6017581820487976, "kl": 0.182861328125, "learning_rate": 4.843036288851434e-06, "loss": -0.0233, "num_tokens": 33402040.0, "reward": 0.3499999940395355, "reward_std": 0.17126278579235077, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.25, "rewards/unicoder_reward_fn/std": 0.32685691118240356, "step": 378 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1413.0, "completions/max_terminated_length": 1413.0, "completions/mean_length": 689.4375, "completions/mean_terminated_length": 689.4375, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "epoch": 0.1470131885182312, "grad_norm": 0.45649734139442444, "kl": 0.182861328125, "learning_rate": 4.841997093248566e-06, "loss": 0.0326, "num_tokens": 33508235.0, "reward": 0.2775000035762787, "reward_std": 0.07328703254461288, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.25851327180862427, "step": 379 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1435.0, "completions/max_terminated_length": 1435.0, "completions/mean_length": 550.1375122070312, "completions/mean_terminated_length": 550.1375122070312, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.1474010861132661, "grad_norm": 0.6699570417404175, "kl": 0.206298828125, "learning_rate": 4.840954594096738e-06, "loss": 0.0071, "num_tokens": 33593390.0, "reward": 0.375, "reward_std": 0.09772800654172897, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.3548934757709503, "step": 380 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1073.0, "completions/mean_length": 654.625, "completions/mean_terminated_length": 611.0632934570312, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.147788983708301, "grad_norm": 0.6532240509986877, "kl": 0.18994140625, "learning_rate": 4.8399087930421975e-06, "loss": 0.0761, "num_tokens": 33692464.0, "reward": 0.3837500214576721, "reward_std": 0.13079170882701874, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.2850000262260437, "rewards/unicoder_reward_fn/std": 0.29813769459724426, "step": 381 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2735.0, "completions/max_terminated_length": 2735.0, "completions/mean_length": 595.5125122070312, "completions/mean_terminated_length": 595.5125122070312, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.14817688130333592, "grad_norm": 0.7225553393363953, "kl": 0.21435546875, "learning_rate": 4.838859691736409e-06, "loss": -0.0535, "num_tokens": 33800361.0, "reward": 0.3199999928474426, "reward_std": 0.13292710483074188, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2199999988079071, "rewards/unicoder_reward_fn/std": 0.2517784833908081, "step": 382 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1349.0, "completions/max_terminated_length": 1349.0, "completions/mean_length": 660.0875244140625, "completions/mean_terminated_length": 660.0875244140625, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "epoch": 0.14856477889837083, "grad_norm": 0.6393365859985352, "kl": 0.173828125, "learning_rate": 4.837807291836044e-06, "loss": -0.0105, "num_tokens": 33899088.0, "reward": 0.35874998569488525, "reward_std": 0.22099052369594574, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.25999999046325684, "rewards/unicoder_reward_fn/std": 0.314058780670166, "step": 383 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1747.0, "completions/max_terminated_length": 1747.0, "completions/mean_length": 674.5125122070312, "completions/mean_terminated_length": 674.5125122070312, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "epoch": 0.14895267649340574, "grad_norm": 0.601732075214386, "kl": 0.190673828125, "learning_rate": 4.836751595002989e-06, "loss": 0.0287, "num_tokens": 33992337.0, "reward": 0.2900000214576721, "reward_std": 0.1389540284872055, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1899999976158142, "rewards/unicoder_reward_fn/std": 0.24628892540931702, "step": 384 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2258.0, "completions/mean_length": 820.5875244140625, "completions/mean_terminated_length": 736.6026000976562, "completions/min_length": 275.0, "completions/min_terminated_length": 275.0, "epoch": 0.14934057408844065, "grad_norm": 0.5381890535354614, "kl": 0.170654296875, "learning_rate": 4.835692602904329e-06, "loss": 0.0027, "num_tokens": 34113576.0, "reward": 0.32750001549720764, "reward_std": 0.09808167070150375, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.23000001907348633, "rewards/unicoder_reward_fn/std": 0.2757728099822998, "step": 385 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1567.0, "completions/max_terminated_length": 1567.0, "completions/mean_length": 654.0625, "completions/mean_terminated_length": 654.0625, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.14972847168347556, "grad_norm": 0.6360284090042114, "kl": 0.204345703125, "learning_rate": 4.834630317212361e-06, "loss": -0.0162, "num_tokens": 34202197.0, "reward": 0.4375, "reward_std": 0.12398987263441086, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3375000059604645, "rewards/unicoder_reward_fn/std": 0.3238905072212219, "step": 386 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3122.0, "completions/max_terminated_length": 3122.0, "completions/mean_length": 676.6375122070312, "completions/mean_terminated_length": 676.6375122070312, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.15011636927851046, "grad_norm": 0.6481001377105713, "kl": 0.206787109375, "learning_rate": 4.833564739604578e-06, "loss": 0.0481, "num_tokens": 34301168.0, "reward": 0.42500001192092896, "reward_std": 0.1469370573759079, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.32499998807907104, "rewards/unicoder_reward_fn/std": 0.29745760560035706, "step": 387 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2782.0, "completions/max_terminated_length": 2782.0, "completions/mean_length": 807.5, "completions/mean_terminated_length": 807.5, "completions/min_length": 333.0, "completions/min_terminated_length": 333.0, "epoch": 0.15050426687354537, "grad_norm": 0.5976157784461975, "kl": 0.1865234375, "learning_rate": 4.832495871763672e-06, "loss": 0.0081, "num_tokens": 34414328.0, "reward": 0.3525000214576721, "reward_std": 0.15128082036972046, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2524999976158142, "rewards/unicoder_reward_fn/std": 0.27695223689079285, "step": 388 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3572.0, "completions/max_terminated_length": 3572.0, "completions/mean_length": 729.2250366210938, "completions/mean_terminated_length": 729.2250366210938, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "epoch": 0.15089216446858028, "grad_norm": 0.6867687106132507, "kl": 0.18994140625, "learning_rate": 4.831423715377531e-06, "loss": 0.1008, "num_tokens": 34521442.0, "reward": 0.4462500512599945, "reward_std": 0.21356628835201263, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.3474999964237213, "rewards/unicoder_reward_fn/std": 0.3031376004219055, "step": 389 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1589.0, "completions/max_terminated_length": 1589.0, "completions/mean_length": 762.9000244140625, "completions/mean_terminated_length": 762.9000244140625, "completions/min_length": 362.0, "completions/min_terminated_length": 362.0, "epoch": 0.15128006206361522, "grad_norm": 0.5581037402153015, "kl": 0.217529296875, "learning_rate": 4.830348272139236e-06, "loss": 0.0193, "num_tokens": 34629690.0, "reward": 0.4150000214576721, "reward_std": 0.12936784327030182, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3149999976158142, "rewards/unicoder_reward_fn/std": 0.27517545223236084, "step": 390 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1239.0, "completions/max_terminated_length": 1239.0, "completions/mean_length": 677.4500122070312, "completions/mean_terminated_length": 677.4500122070312, "completions/min_length": 303.0, "completions/min_terminated_length": 303.0, "epoch": 0.15166795965865013, "grad_norm": 0.5196791887283325, "kl": 0.195556640625, "learning_rate": 4.829269543747061e-06, "loss": -0.0022, "num_tokens": 34731230.0, "reward": 0.33250001072883606, "reward_std": 0.08610769361257553, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23249998688697815, "rewards/unicoder_reward_fn/std": 0.26897603273391724, "step": 391 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1283.0, "completions/max_terminated_length": 1283.0, "completions/mean_length": 605.6625366210938, "completions/mean_terminated_length": 605.6625366210938, "completions/min_length": 259.0, "completions/min_terminated_length": 259.0, "epoch": 0.15205585725368503, "grad_norm": 0.6688417196273804, "kl": 0.224609375, "learning_rate": 4.828187531904462e-06, "loss": -0.01, "num_tokens": 34821003.0, "reward": 0.3787500262260437, "reward_std": 0.17359109222888947, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.2800000309944153, "rewards/unicoder_reward_fn/std": 0.2821257412433624, "step": 392 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1461.0, "completions/max_terminated_length": 1461.0, "completions/mean_length": 724.9375, "completions/mean_terminated_length": 724.9375, "completions/min_length": 280.0, "completions/min_terminated_length": 280.0, "epoch": 0.15244375484871994, "grad_norm": 0.5300759077072144, "kl": 0.198974609375, "learning_rate": 4.827102238320086e-06, "loss": 0.0159, "num_tokens": 34941454.0, "reward": 0.19500000774860382, "reward_std": 0.08708720654249191, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.17422524094581604, "step": 393 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1780.0, "completions/max_terminated_length": 1780.0, "completions/mean_length": 794.3125, "completions/mean_terminated_length": 794.3125, "completions/min_length": 406.0, "completions/min_terminated_length": 406.0, "epoch": 0.15283165244375485, "grad_norm": 0.564262866973877, "kl": 0.21435546875, "learning_rate": 4.826013664707759e-06, "loss": 0.0045, "num_tokens": 35052599.0, "reward": 0.36250001192092896, "reward_std": 0.11250827461481094, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.26249998807907104, "rewards/unicoder_reward_fn/std": 0.28031402826309204, "step": 394 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2046.0, "completions/max_terminated_length": 2046.0, "completions/mean_length": 805.2999877929688, "completions/mean_terminated_length": 805.2999877929688, "completions/min_length": 202.0, "completions/min_terminated_length": 202.0, "epoch": 0.15321955003878976, "grad_norm": 0.4456382989883423, "kl": 0.21484375, "learning_rate": 4.824921812786486e-06, "loss": 0.0257, "num_tokens": 35166271.0, "reward": 0.29250001907348633, "reward_std": 0.07757189869880676, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1925000101327896, "rewards/unicoder_reward_fn/std": 0.2924254834651947, "step": 395 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2239.0, "completions/max_terminated_length": 2239.0, "completions/mean_length": 730.3875122070312, "completions/mean_terminated_length": 730.3875122070312, "completions/min_length": 233.0, "completions/min_terminated_length": 233.0, "epoch": 0.15360744763382467, "grad_norm": 0.43069660663604736, "kl": 0.1943359375, "learning_rate": 4.823826684280453e-06, "loss": -0.0154, "num_tokens": 35283262.0, "reward": 0.3150000274181366, "reward_std": 0.07362651079893112, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2150000035762787, "rewards/unicoder_reward_fn/std": 0.27883121371269226, "step": 396 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1719.0, "completions/max_terminated_length": 1719.0, "completions/mean_length": 754.3875122070312, "completions/mean_terminated_length": 754.3875122070312, "completions/min_length": 332.0, "completions/min_terminated_length": 332.0, "epoch": 0.15399534522885958, "grad_norm": 0.6392853856086731, "kl": 0.226806640625, "learning_rate": 4.822728280919014e-06, "loss": -0.0374, "num_tokens": 35390773.0, "reward": 0.38750001788139343, "reward_std": 0.12967856228351593, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.23780377209186554, "step": 397 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1897.0, "completions/max_terminated_length": 1897.0, "completions/mean_length": 689.625, "completions/mean_terminated_length": 689.625, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "epoch": 0.15438324282389448, "grad_norm": 0.6223946809768677, "kl": 0.223388671875, "learning_rate": 4.821626604436701e-06, "loss": 0.0189, "num_tokens": 35498815.0, "reward": 0.30750003457069397, "reward_std": 0.12081482261419296, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.2924254834651947, "step": 398 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1528.0, "completions/max_terminated_length": 1528.0, "completions/mean_length": 678.875, "completions/mean_terminated_length": 678.875, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.1547711404189294, "grad_norm": 0.6619320511817932, "kl": 0.2109375, "learning_rate": 4.820521656573209e-06, "loss": 0.0236, "num_tokens": 35598901.0, "reward": 0.3700000047683716, "reward_std": 0.15073540806770325, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.27000001072883606, "rewards/unicoder_reward_fn/std": 0.3199683725833893, "step": 399 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1578.0, "completions/max_terminated_length": 1578.0, "completions/mean_length": 787.5750122070312, "completions/mean_terminated_length": 787.5750122070312, "completions/min_length": 329.0, "completions/min_terminated_length": 329.0, "epoch": 0.1551590380139643, "grad_norm": 0.5480032563209534, "kl": 0.2021484375, "learning_rate": 4.819413439073402e-06, "loss": -0.03, "num_tokens": 35712499.0, "reward": 0.2849999964237213, "reward_std": 0.10818672180175781, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1850000023841858, "rewards/unicoder_reward_fn/std": 0.25011390447616577, "step": 400 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1434.0, "completions/max_terminated_length": 1434.0, "completions/mean_length": 781.1500244140625, "completions/mean_terminated_length": 781.1500244140625, "completions/min_length": 320.0, "completions/min_terminated_length": 320.0, "epoch": 0.15554693560899924, "grad_norm": 0.5019673109054565, "kl": 0.1982421875, "learning_rate": 4.818301953687309e-06, "loss": -0.0246, "num_tokens": 35835719.0, "reward": 0.23000001907348633, "reward_std": 0.07519414275884628, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.18029510974884033, "step": 401 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2523.0, "completions/max_terminated_length": 2523.0, "completions/mean_length": 860.0750122070312, "completions/mean_terminated_length": 860.0750122070312, "completions/min_length": 376.0, "completions/min_terminated_length": 376.0, "epoch": 0.15593483320403415, "grad_norm": 0.6035289168357849, "kl": 0.208740234375, "learning_rate": 4.817187202170114e-06, "loss": 0.0597, "num_tokens": 35950549.0, "reward": 0.4025000035762787, "reward_std": 0.2043372392654419, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.30250000953674316, "rewards/unicoder_reward_fn/std": 0.30020037293434143, "step": 402 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1114.0, "completions/max_terminated_length": 1114.0, "completions/mean_length": 671.6500244140625, "completions/mean_terminated_length": 671.6500244140625, "completions/min_length": 296.0, "completions/min_terminated_length": 296.0, "epoch": 0.15632273079906905, "grad_norm": 0.552253007888794, "kl": 0.19384765625, "learning_rate": 4.816069186282165e-06, "loss": 0.0027, "num_tokens": 36049337.0, "reward": 0.33500000834465027, "reward_std": 0.10818202793598175, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23499998450279236, "rewards/unicoder_reward_fn/std": 0.2658161520957947, "step": 403 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1248.0, "completions/max_terminated_length": 1248.0, "completions/mean_length": 691.6375122070312, "completions/mean_terminated_length": 691.6375122070312, "completions/min_length": 296.0, "completions/min_terminated_length": 296.0, "epoch": 0.15671062839410396, "grad_norm": 0.6150469779968262, "kl": 0.19775390625, "learning_rate": 4.8149479077889575e-06, "loss": 0.0257, "num_tokens": 36146004.0, "reward": 0.4775000214576721, "reward_std": 0.1591476947069168, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3774999976158142, "rewards/unicoder_reward_fn/std": 0.30020037293434143, "step": 404 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1730.0, "completions/max_terminated_length": 1730.0, "completions/mean_length": 751.5499877929688, "completions/mean_terminated_length": 751.5499877929688, "completions/min_length": 346.0, "completions/min_terminated_length": 346.0, "epoch": 0.15709852598913887, "grad_norm": 0.48276424407958984, "kl": 0.193115234375, "learning_rate": 4.813823368461148e-06, "loss": 0.0129, "num_tokens": 36249672.0, "reward": 0.30250000953674316, "reward_std": 0.1026988998055458, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.23491178452968597, "step": 405 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1192.0, "completions/max_terminated_length": 1192.0, "completions/mean_length": 660.6375122070312, "completions/mean_terminated_length": 660.6375122070312, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "epoch": 0.15748642358417378, "grad_norm": 0.5968731641769409, "kl": 0.20263671875, "learning_rate": 4.812695570074533e-06, "loss": 0.0267, "num_tokens": 36348779.0, "reward": 0.41750001907348633, "reward_std": 0.12250276654958725, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3175000250339508, "rewards/unicoder_reward_fn/std": 0.3080255687236786, "step": 406 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3371.0, "completions/mean_length": 832.7875366210938, "completions/mean_terminated_length": 749.1154174804688, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.1578743211792087, "grad_norm": 0.5995007157325745, "kl": 0.182373046875, "learning_rate": 4.811564514410064e-06, "loss": 0.1484, "num_tokens": 36458898.0, "reward": 0.29500001668930054, "reward_std": 0.11505033820867538, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.19750000536441803, "rewards/unicoder_reward_fn/std": 0.28727221488952637, "step": 407 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1413.0, "completions/max_terminated_length": 1413.0, "completions/mean_length": 673.9750366210938, "completions/mean_terminated_length": 673.9750366210938, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "epoch": 0.1582622187742436, "grad_norm": 0.6277256011962891, "kl": 0.19287109375, "learning_rate": 4.8104302032538305e-06, "loss": 0.0177, "num_tokens": 36556808.0, "reward": 0.4437500536441803, "reward_std": 0.10238873958587646, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.3450000286102295, "rewards/unicoder_reward_fn/std": 0.2916127145290375, "step": 408 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2153.0, "completions/max_terminated_length": 2153.0, "completions/mean_length": 725.0250244140625, "completions/mean_terminated_length": 725.0250244140625, "completions/min_length": 386.0, "completions/min_terminated_length": 386.0, "epoch": 0.1586501163692785, "grad_norm": 0.5282853245735168, "kl": 0.209228515625, "learning_rate": 4.809292638397065e-06, "loss": 0.0025, "num_tokens": 36668514.0, "reward": 0.2587500214576721, "reward_std": 0.07552998512983322, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.1599999964237213, "rewards/unicoder_reward_fn/std": 0.21441709995269775, "step": 409 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1200.0, "completions/max_terminated_length": 1200.0, "completions/mean_length": 617.7250366210938, "completions/mean_terminated_length": 617.7250366210938, "completions/min_length": 262.0, "completions/min_terminated_length": 262.0, "epoch": 0.1590380139643134, "grad_norm": 0.6184208989143372, "kl": 0.209228515625, "learning_rate": 4.8081518216361374e-06, "loss": 0.0158, "num_tokens": 36769572.0, "reward": 0.3400000035762787, "reward_std": 0.12013296037912369, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24000000953674316, "rewards/unicoder_reward_fn/std": 0.24525883793830872, "step": 410 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1255.0, "completions/max_terminated_length": 1255.0, "completions/mean_length": 651.0125122070312, "completions/mean_terminated_length": 651.0125122070312, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "epoch": 0.15942591155934832, "grad_norm": 0.5885857343673706, "kl": 0.20361328125, "learning_rate": 4.807007754772552e-06, "loss": 0.016, "num_tokens": 36866829.0, "reward": 0.36250001192092896, "reward_std": 0.1503179520368576, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.26249998807907104, "rewards/unicoder_reward_fn/std": 0.28920450806617737, "step": 411 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1280.0, "completions/max_terminated_length": 1280.0, "completions/mean_length": 617.9625244140625, "completions/mean_terminated_length": 617.9625244140625, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.15981380915438323, "grad_norm": 0.5812550783157349, "kl": 0.191162109375, "learning_rate": 4.80586043961295e-06, "loss": 0.0297, "num_tokens": 36960338.0, "reward": 0.36500000953674316, "reward_std": 0.068752221763134, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.26500001549720764, "rewards/unicoder_reward_fn/std": 0.29472148418426514, "step": 412 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1363.0, "completions/max_terminated_length": 1363.0, "completions/mean_length": 572.4125366210938, "completions/mean_terminated_length": 572.4125366210938, "completions/min_length": 201.0, "completions/min_terminated_length": 201.0, "epoch": 0.16020170674941817, "grad_norm": 0.6088348627090454, "kl": 0.209228515625, "learning_rate": 4.804709877969095e-06, "loss": 0.014, "num_tokens": 37049203.0, "reward": 0.43250003457069397, "reward_std": 0.1985878348350525, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.33250001072883606, "rewards/unicoder_reward_fn/std": 0.33062082529067993, "step": 413 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 870.0, "completions/max_terminated_length": 870.0, "completions/mean_length": 514.4500122070312, "completions/mean_terminated_length": 514.4500122070312, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "epoch": 0.16058960434445307, "grad_norm": 0.646271288394928, "kl": 0.20458984375, "learning_rate": 4.803556071657882e-06, "loss": 0.0167, "num_tokens": 37127103.0, "reward": 0.4937500059604645, "reward_std": 0.10067512840032578, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.39499998092651367, "rewards/unicoder_reward_fn/std": 0.33066868782043457, "step": 414 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1221.0, "completions/max_terminated_length": 1221.0, "completions/mean_length": 542.375, "completions/mean_terminated_length": 542.375, "completions/min_length": 336.0, "completions/min_terminated_length": 336.0, "epoch": 0.16097750193948798, "grad_norm": 0.7084649801254272, "kl": 0.219970703125, "learning_rate": 4.802399022501329e-06, "loss": -0.0165, "num_tokens": 37219277.0, "reward": 0.45000001788139343, "reward_std": 0.13657459616661072, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3499999940395355, "rewards/unicoder_reward_fn/std": 0.2454652041196823, "step": 415 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 940.0, "completions/max_terminated_length": 940.0, "completions/mean_length": 509.1125183105469, "completions/mean_terminated_length": 509.1125183105469, "completions/min_length": 264.0, "completions/min_terminated_length": 264.0, "epoch": 0.1613653995345229, "grad_norm": 0.6960617303848267, "kl": 0.21728515625, "learning_rate": 4.801238732326575e-06, "loss": 0.0033, "num_tokens": 37313022.0, "reward": 0.4700000286102295, "reward_std": 0.13144013285636902, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.37000003457069397, "rewards/unicoder_reward_fn/std": 0.33692580461502075, "step": 416 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 934.0, "completions/max_terminated_length": 934.0, "completions/mean_length": 493.0, "completions/mean_terminated_length": 493.0, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "epoch": 0.1617532971295578, "grad_norm": 0.6826635599136353, "kl": 0.225341796875, "learning_rate": 4.8000752029658755e-06, "loss": 0.0162, "num_tokens": 37390574.0, "reward": 0.25999999046325684, "reward_std": 0.12483979761600494, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1600000113248825, "rewards/unicoder_reward_fn/std": 0.19974668323993683, "step": 417 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 899.0, "completions/max_terminated_length": 899.0, "completions/mean_length": 505.6000061035156, "completions/mean_terminated_length": 505.6000061035156, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "epoch": 0.1621411947245927, "grad_norm": 0.7144949436187744, "kl": 0.23095703125, "learning_rate": 4.798908436256602e-06, "loss": 0.0218, "num_tokens": 37478390.0, "reward": 0.4000000059604645, "reward_std": 0.11540473997592926, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.30000001192092896, "rewards/unicoder_reward_fn/std": 0.27558913826942444, "step": 418 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 840.0, "completions/max_terminated_length": 840.0, "completions/mean_length": 483.1625061035156, "completions/mean_terminated_length": 483.1625061035156, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "epoch": 0.16252909231962762, "grad_norm": 0.6581419110298157, "kl": 0.223388671875, "learning_rate": 4.7977384340412405e-06, "loss": 0.0173, "num_tokens": 37578267.0, "reward": 0.41750001907348633, "reward_std": 0.08841761946678162, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3175000250339508, "rewards/unicoder_reward_fn/std": 0.22657325863838196, "step": 419 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 897.0, "completions/max_terminated_length": 897.0, "completions/mean_length": 528.7125244140625, "completions/mean_terminated_length": 528.7125244140625, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.16291698991466252, "grad_norm": 0.7073608040809631, "kl": 0.22412109375, "learning_rate": 4.796565198167383e-06, "loss": -0.0103, "num_tokens": 37681604.0, "reward": 0.36500000953674316, "reward_std": 0.10469266027212143, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.26500001549720764, "rewards/unicoder_reward_fn/std": 0.2733292281627655, "step": 420 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 853.0, "completions/max_terminated_length": 853.0, "completions/mean_length": 503.75, "completions/mean_terminated_length": 503.75, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.16330488750969743, "grad_norm": 0.6489725112915039, "kl": 0.2255859375, "learning_rate": 4.795388730487728e-06, "loss": 0.0256, "num_tokens": 37771264.0, "reward": 0.3499999940395355, "reward_std": 0.08715878427028656, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.25, "rewards/unicoder_reward_fn/std": 0.23057441413402557, "step": 421 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1092.0, "completions/max_terminated_length": 1092.0, "completions/mean_length": 512.7625122070312, "completions/mean_terminated_length": 512.7625122070312, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.16369278510473234, "grad_norm": 0.5546074509620667, "kl": 0.2138671875, "learning_rate": 4.7942090328600784e-06, "loss": -0.0089, "num_tokens": 37858669.0, "reward": 0.3725000321865082, "reward_std": 0.10811425745487213, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2724999785423279, "rewards/unicoder_reward_fn/std": 0.2850605249404907, "step": 422 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1013.0, "completions/max_terminated_length": 1013.0, "completions/mean_length": 519.7750244140625, "completions/mean_terminated_length": 519.7750244140625, "completions/min_length": 264.0, "completions/min_terminated_length": 264.0, "epoch": 0.16408068269976725, "grad_norm": 0.7664618492126465, "kl": 0.226806640625, "learning_rate": 4.79302610714734e-06, "loss": 0.0626, "num_tokens": 37952659.0, "reward": 0.3425000011920929, "reward_std": 0.19765309989452362, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24250002205371857, "rewards/unicoder_reward_fn/std": 0.2996939420700073, "step": 423 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1004.0, "completions/max_terminated_length": 1004.0, "completions/mean_length": 446.0625, "completions/mean_terminated_length": 446.0625, "completions/min_length": 252.0, "completions/min_terminated_length": 252.0, "epoch": 0.16446858029480219, "grad_norm": 0.7964127063751221, "kl": 0.244384765625, "learning_rate": 4.791839955217514e-06, "loss": 0.0242, "num_tokens": 38039400.0, "reward": 0.41999998688697815, "reward_std": 0.16017186641693115, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3199999928474426, "rewards/unicoder_reward_fn/std": 0.2978403568267822, "step": 424 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 803.0, "completions/max_terminated_length": 803.0, "completions/mean_length": 437.5625, "completions/mean_terminated_length": 437.5625, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.1648564778898371, "grad_norm": 0.7590237855911255, "kl": 0.233642578125, "learning_rate": 4.790650578943693e-06, "loss": 0.0043, "num_tokens": 38128029.0, "reward": 0.49500003457069397, "reward_std": 0.15651078522205353, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.39500001072883606, "rewards/unicoder_reward_fn/std": 0.32913389801979065, "step": 425 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 674.0, "completions/max_terminated_length": 674.0, "completions/mean_length": 391.8000183105469, "completions/mean_terminated_length": 391.8000183105469, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.165244375484872, "grad_norm": 0.6445778608322144, "kl": 0.22509765625, "learning_rate": 4.789457980204067e-06, "loss": 0.0179, "num_tokens": 38199693.0, "reward": 0.4400000274181366, "reward_std": 0.13466466963291168, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3399999737739563, "rewards/unicoder_reward_fn/std": 0.2707701325416565, "step": 426 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 744.0, "completions/max_terminated_length": 744.0, "completions/mean_length": 413.125, "completions/mean_terminated_length": 413.125, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.1656322730799069, "grad_norm": 0.812999963760376, "kl": 0.25244140625, "learning_rate": 4.788262160881912e-06, "loss": 0.0312, "num_tokens": 38270887.0, "reward": 0.3425000309944153, "reward_std": 0.11648782342672348, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24250002205371857, "rewards/unicoder_reward_fn/std": 0.21510964632034302, "step": 427 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 608.0, "completions/max_terminated_length": 608.0, "completions/mean_length": 411.0874938964844, "completions/mean_terminated_length": 411.0874938964844, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.16602017067494182, "grad_norm": 0.8191460967063904, "kl": 0.2529296875, "learning_rate": 4.78706312286559e-06, "loss": 0.0022, "num_tokens": 38345726.0, "reward": 0.3850000202655792, "reward_std": 0.11537822335958481, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2849999964237213, "rewards/unicoder_reward_fn/std": 0.2222440242767334, "step": 428 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 817.0, "completions/max_terminated_length": 817.0, "completions/mean_length": 441.1750183105469, "completions/mean_terminated_length": 441.1750183105469, "completions/min_length": 196.0, "completions/min_terminated_length": 196.0, "epoch": 0.16640806826997673, "grad_norm": 0.7690134048461914, "kl": 0.237060546875, "learning_rate": 4.7858608680485444e-06, "loss": 0.0239, "num_tokens": 38438044.0, "reward": 0.3174999952316284, "reward_std": 0.14086806774139404, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2175000160932541, "rewards/unicoder_reward_fn/std": 0.2627868950366974, "step": 429 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2769.0, "completions/max_terminated_length": 2769.0, "completions/mean_length": 465.4750061035156, "completions/mean_terminated_length": 465.4750061035156, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "epoch": 0.16679596586501164, "grad_norm": 0.625775933265686, "kl": 0.2060546875, "learning_rate": 4.784655398329301e-06, "loss": -0.0016, "num_tokens": 38521594.0, "reward": 0.4449999928474426, "reward_std": 0.10855213552713394, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3449999690055847, "rewards/unicoder_reward_fn/std": 0.35145270824432373, "step": 430 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2370.0, "completions/max_terminated_length": 2370.0, "completions/mean_length": 427.7875061035156, "completions/mean_terminated_length": 427.7875061035156, "completions/min_length": 210.0, "completions/min_terminated_length": 210.0, "epoch": 0.16718386346004654, "grad_norm": 0.6834935545921326, "kl": 0.249755859375, "learning_rate": 4.78344671561146e-06, "loss": 0.0288, "num_tokens": 38601873.0, "reward": 0.4749999940395355, "reward_std": 0.08456059545278549, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.375, "rewards/unicoder_reward_fn/std": 0.2631359100341797, "step": 431 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 732.0, "completions/max_terminated_length": 732.0, "completions/mean_length": 401.6000061035156, "completions/mean_terminated_length": 401.6000061035156, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "epoch": 0.16757176105508145, "grad_norm": 0.8147160410881042, "kl": 0.236572265625, "learning_rate": 4.782234821803698e-06, "loss": 0.0319, "num_tokens": 38680121.0, "reward": 0.4075000286102295, "reward_std": 0.09751089662313461, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3075000047683716, "rewards/unicoder_reward_fn/std": 0.2584153413772583, "step": 432 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2127.0, "completions/max_terminated_length": 2127.0, "completions/mean_length": 436.3000183105469, "completions/mean_terminated_length": 436.3000183105469, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.16795965865011636, "grad_norm": 0.558454692363739, "kl": 0.247314453125, "learning_rate": 4.781019718819759e-06, "loss": 0.0041, "num_tokens": 38759257.0, "reward": 0.29749998450279236, "reward_std": 0.0419151745736599, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19749999046325684, "rewards/unicoder_reward_fn/std": 0.1948547214269638, "step": 433 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 777.0, "completions/max_terminated_length": 777.0, "completions/mean_length": 391.70001220703125, "completions/mean_terminated_length": 391.70001220703125, "completions/min_length": 218.0, "completions/min_terminated_length": 218.0, "epoch": 0.16834755624515127, "grad_norm": 0.7098185420036316, "kl": 0.26171875, "learning_rate": 4.779801408578457e-06, "loss": 0.0273, "num_tokens": 38839289.0, "reward": 0.3700000047683716, "reward_std": 0.11255146563053131, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.27000001072883606, "rewards/unicoder_reward_fn/std": 0.30206045508384705, "step": 434 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 826.0, "completions/max_terminated_length": 826.0, "completions/mean_length": 344.0, "completions/mean_terminated_length": 344.0, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.16873545384018618, "grad_norm": 0.94935542345047, "kl": 0.27294921875, "learning_rate": 4.77857989300367e-06, "loss": 0.0424, "num_tokens": 38909265.0, "reward": 0.4375, "reward_std": 0.14574384689331055, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3375000059604645, "rewards/unicoder_reward_fn/std": 0.29612478613853455, "step": 435 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1559.0, "completions/max_terminated_length": 1559.0, "completions/mean_length": 416.4624938964844, "completions/mean_terminated_length": 416.4624938964844, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.16912335143522111, "grad_norm": 0.768760621547699, "kl": 0.2490234375, "learning_rate": 4.777355174024339e-06, "loss": 0.0344, "num_tokens": 38999678.0, "reward": 0.3575000464916229, "reward_std": 0.15818674862384796, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2574999928474426, "rewards/unicoder_reward_fn/std": 0.2971489131450653, "step": 436 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 718.0, "completions/max_terminated_length": 718.0, "completions/mean_length": 399.5375061035156, "completions/mean_terminated_length": 399.5375061035156, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.16951124903025602, "grad_norm": 0.6689597368240356, "kl": 0.239501953125, "learning_rate": 4.7761272535744615e-06, "loss": 0.0044, "num_tokens": 39085585.0, "reward": 0.5250000357627869, "reward_std": 0.08403722196817398, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.42500001192092896, "rewards/unicoder_reward_fn/std": 0.3491399884223938, "step": 437 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 779.0, "completions/max_terminated_length": 779.0, "completions/mean_length": 369.875, "completions/mean_terminated_length": 369.875, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.16989914662529093, "grad_norm": 0.8592768907546997, "kl": 0.2548828125, "learning_rate": 4.774896133593091e-06, "loss": -0.0032, "num_tokens": 39162655.0, "reward": 0.4100000560283661, "reward_std": 0.1500496119260788, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3100000023841858, "rewards/unicoder_reward_fn/std": 0.25036683678627014, "step": 438 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 575.0, "completions/max_terminated_length": 575.0, "completions/mean_length": 353.25, "completions/mean_terminated_length": 353.25, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.17028704422032584, "grad_norm": 0.7510709762573242, "kl": 0.28759765625, "learning_rate": 4.773661816024335e-06, "loss": 0.0065, "num_tokens": 39249267.0, "reward": 0.4675000309944153, "reward_std": 0.1376277059316635, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3675000071525574, "rewards/unicoder_reward_fn/std": 0.28718408942222595, "step": 439 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 796.0, "completions/max_terminated_length": 796.0, "completions/mean_length": 353.5625, "completions/mean_terminated_length": 353.5625, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.17067494181536075, "grad_norm": 0.688008725643158, "kl": 0.27880859375, "learning_rate": 4.772424302817349e-06, "loss": -0.0155, "num_tokens": 39323016.0, "reward": 0.33250004053115845, "reward_std": 0.0819007083773613, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23250000178813934, "rewards/unicoder_reward_fn/std": 0.3606516420841217, "step": 440 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1009.0, "completions/max_terminated_length": 1009.0, "completions/mean_length": 409.5, "completions/mean_terminated_length": 409.5, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.17106283941039566, "grad_norm": 0.8687078356742859, "kl": 0.24169921875, "learning_rate": 4.771183595926336e-06, "loss": 0.0102, "num_tokens": 39411912.0, "reward": 0.3425000309944153, "reward_std": 0.06512318551540375, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24250002205371857, "rewards/unicoder_reward_fn/std": 0.23099946975708008, "step": 441 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 506.0, "completions/max_terminated_length": 506.0, "completions/mean_length": 335.38751220703125, "completions/mean_terminated_length": 335.38751220703125, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.17145073700543056, "grad_norm": 0.7888396382331848, "kl": 0.24072265625, "learning_rate": 4.769939697310539e-06, "loss": 0.0222, "num_tokens": 39468295.0, "reward": 0.45500001311302185, "reward_std": 0.11806197464466095, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.35500001907348633, "rewards/unicoder_reward_fn/std": 0.301850825548172, "step": 442 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 764.0, "completions/max_terminated_length": 764.0, "completions/mean_length": 408.625, "completions/mean_terminated_length": 408.625, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "epoch": 0.17183863460046547, "grad_norm": 0.603972852230072, "kl": 0.24609375, "learning_rate": 4.768692608934247e-06, "loss": -0.0062, "num_tokens": 39550465.0, "reward": 0.45000001788139343, "reward_std": 0.12040966749191284, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3499999940395355, "rewards/unicoder_reward_fn/std": 0.36630210280418396, "step": 443 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 662.0, "completions/max_terminated_length": 662.0, "completions/mean_length": 371.0249938964844, "completions/mean_terminated_length": 371.0249938964844, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.17222653219550038, "grad_norm": 0.9261247515678406, "kl": 0.2724609375, "learning_rate": 4.767442332766781e-06, "loss": -0.0067, "num_tokens": 39626067.0, "reward": 0.5300000309944153, "reward_std": 0.1683577597141266, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4300000071525574, "rewards/unicoder_reward_fn/std": 0.31357473134994507, "step": 444 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 857.0, "completions/max_terminated_length": 857.0, "completions/mean_length": 434.1875, "completions/mean_terminated_length": 434.1875, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.1726144297905353, "grad_norm": 0.8190153241157532, "kl": 0.25439453125, "learning_rate": 4.766188870782497e-06, "loss": 0.0301, "num_tokens": 39709434.0, "reward": 0.2775000035762787, "reward_std": 0.11409395933151245, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.20865754783153534, "step": 445 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 696.0, "completions/max_terminated_length": 696.0, "completions/mean_length": 412.2875061035156, "completions/mean_terminated_length": 412.2875061035156, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.1730023273855702, "grad_norm": 0.7095459699630737, "kl": 0.2607421875, "learning_rate": 4.764932224960786e-06, "loss": -0.011, "num_tokens": 39794177.0, "reward": 0.4050000309944153, "reward_std": 0.09651970863342285, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3050000071525574, "rewards/unicoder_reward_fn/std": 0.32135000824928284, "step": 446 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 798.0, "completions/max_terminated_length": 798.0, "completions/mean_length": 415.4375, "completions/mean_terminated_length": 415.4375, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.1733902249806051, "grad_norm": 0.8166552186012268, "kl": 0.244140625, "learning_rate": 4.763672397286059e-06, "loss": 0.0156, "num_tokens": 39880668.0, "reward": 0.6000000834465027, "reward_std": 0.1566130667924881, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.5, "rewards/unicoder_reward_fn/std": 0.38186749815940857, "step": 447 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 764.0, "completions/max_terminated_length": 764.0, "completions/mean_length": 402.1499938964844, "completions/mean_terminated_length": 402.1499938964844, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "epoch": 0.17377812257564004, "grad_norm": 0.8449990153312683, "kl": 0.28955078125, "learning_rate": 4.762409389747758e-06, "loss": -0.004, "num_tokens": 39955384.0, "reward": 0.3100000023841858, "reward_std": 0.11056756973266602, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20999999344348907, "rewards/unicoder_reward_fn/std": 0.1879940778017044, "step": 448 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 755.0, "completions/max_terminated_length": 755.0, "completions/mean_length": 421.70001220703125, "completions/mean_terminated_length": 421.70001220703125, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.17416602017067495, "grad_norm": 0.7230152487754822, "kl": 0.24951171875, "learning_rate": 4.761143204340344e-06, "loss": -0.0208, "num_tokens": 40034944.0, "reward": 0.5724999904632568, "reward_std": 0.1480284482240677, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4725000262260437, "rewards/unicoder_reward_fn/std": 0.3310799300670624, "step": 449 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1180.0, "completions/max_terminated_length": 1180.0, "completions/mean_length": 447.7124938964844, "completions/mean_terminated_length": 447.7124938964844, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.17455391776570986, "grad_norm": 0.7611681222915649, "kl": 0.248046875, "learning_rate": 4.759873843063297e-06, "loss": -0.0118, "num_tokens": 40115241.0, "reward": 0.4050000309944153, "reward_std": 0.1617054045200348, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3050000071525574, "rewards/unicoder_reward_fn/std": 0.35432231426239014, "step": 450 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 878.0, "completions/max_terminated_length": 878.0, "completions/mean_length": 499.88751220703125, "completions/mean_terminated_length": 499.88751220703125, "completions/min_length": 268.0, "completions/min_terminated_length": 268.0, "epoch": 0.17494181536074477, "grad_norm": 0.6876310706138611, "kl": 0.250244140625, "learning_rate": 4.7586013079211105e-06, "loss": 0.0249, "num_tokens": 40201392.0, "reward": 0.3075000047683716, "reward_std": 0.11647140979766846, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.20974668860435486, "step": 451 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 796.0, "completions/max_terminated_length": 796.0, "completions/mean_length": 490.26251220703125, "completions/mean_terminated_length": 490.26251220703125, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.17532971295577968, "grad_norm": 0.598539412021637, "kl": 0.24365234375, "learning_rate": 4.757325600923293e-06, "loss": 0.0287, "num_tokens": 40287613.0, "reward": 0.3150000274181366, "reward_std": 0.08358418941497803, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2150000035762787, "rewards/unicoder_reward_fn/std": 0.20320221781730652, "step": 452 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 711.0, "completions/max_terminated_length": 711.0, "completions/mean_length": 468.7875061035156, "completions/mean_terminated_length": 468.7875061035156, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.17571761055081458, "grad_norm": 0.6817914247512817, "kl": 0.24755859375, "learning_rate": 4.756046724084358e-06, "loss": 0.0167, "num_tokens": 40372012.0, "reward": 0.22750000655651093, "reward_std": 0.07866684347391129, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1274999976158142, "rewards/unicoder_reward_fn/std": 0.2300110012292862, "step": 453 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1339.0, "completions/max_terminated_length": 1339.0, "completions/mean_length": 476.5, "completions/mean_terminated_length": 476.5, "completions/min_length": 135.0, "completions/min_terminated_length": 135.0, "epoch": 0.1761055081458495, "grad_norm": 0.7396994829177856, "kl": 0.244140625, "learning_rate": 4.754764679423827e-06, "loss": 0.0589, "num_tokens": 40462676.0, "reward": 0.46500006318092346, "reward_std": 0.13452966511249542, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.36500000953674316, "rewards/unicoder_reward_fn/std": 0.28952163457870483, "step": 454 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1045.0, "completions/max_terminated_length": 1045.0, "completions/mean_length": 475.88751220703125, "completions/mean_terminated_length": 475.88751220703125, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "epoch": 0.1764934057408844, "grad_norm": 0.6780167818069458, "kl": 0.2578125, "learning_rate": 4.7534794689662265e-06, "loss": 0.0117, "num_tokens": 40554451.0, "reward": 0.35500001907348633, "reward_std": 0.10790055990219116, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2550000250339508, "rewards/unicoder_reward_fn/std": 0.24437987804412842, "step": 455 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 905.0, "completions/max_terminated_length": 905.0, "completions/mean_length": 464.5375061035156, "completions/mean_terminated_length": 464.5375061035156, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.1768813033359193, "grad_norm": 0.7254911065101624, "kl": 0.2666015625, "learning_rate": 4.752191094741075e-06, "loss": 0.0386, "num_tokens": 40639182.0, "reward": 0.5275000333786011, "reward_std": 0.1468040943145752, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.42750000953674316, "rewards/unicoder_reward_fn/std": 0.29977840185165405, "step": 456 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1749.0, "completions/max_terminated_length": 1749.0, "completions/mean_length": 541.1124877929688, "completions/mean_terminated_length": 541.1124877929688, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.17726920093095422, "grad_norm": 0.5487136244773865, "kl": 0.217041015625, "learning_rate": 4.750899558782894e-06, "loss": 0.0393, "num_tokens": 40735943.0, "reward": 0.20500002801418304, "reward_std": 0.07813965529203415, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.10500000417232513, "rewards/unicoder_reward_fn/std": 0.12314446270465851, "step": 457 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 832.0, "completions/max_terminated_length": 832.0, "completions/mean_length": 439.125, "completions/mean_terminated_length": 439.125, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.17765709852598913, "grad_norm": 0.6805989742279053, "kl": 0.2412109375, "learning_rate": 4.749604863131193e-06, "loss": 0.0346, "num_tokens": 40819481.0, "reward": 0.5200000405311584, "reward_std": 0.15318764746189117, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.42000001668930054, "rewards/unicoder_reward_fn/std": 0.28921544551849365, "step": 458 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1721.0, "completions/max_terminated_length": 1721.0, "completions/mean_length": 504.6875, "completions/mean_terminated_length": 504.6875, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.17804499612102406, "grad_norm": 0.6929457783699036, "kl": 0.26904296875, "learning_rate": 4.748307009830475e-06, "loss": 0.0859, "num_tokens": 40903656.0, "reward": 0.39500001072883606, "reward_std": 0.12428174167871475, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.29500001668930054, "rewards/unicoder_reward_fn/std": 0.22720091044902802, "step": 459 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3598.0, "completions/max_terminated_length": 3598.0, "completions/mean_length": 475.7749938964844, "completions/mean_terminated_length": 475.7749938964844, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "epoch": 0.17843289371605897, "grad_norm": 0.7279007434844971, "kl": 0.26318359375, "learning_rate": 4.747006000930227e-06, "loss": 0.0921, "num_tokens": 40992662.0, "reward": 0.5350000262260437, "reward_std": 0.13770365715026855, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4350000321865082, "rewards/unicoder_reward_fn/std": 0.31626778841018677, "step": 460 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 883.0, "completions/max_terminated_length": 883.0, "completions/mean_length": 460.82501220703125, "completions/mean_terminated_length": 460.82501220703125, "completions/min_length": 239.0, "completions/min_terminated_length": 239.0, "epoch": 0.17882079131109388, "grad_norm": 0.6905207633972168, "kl": 0.2705078125, "learning_rate": 4.74570183848492e-06, "loss": 0.0038, "num_tokens": 41081800.0, "reward": 0.26250001788139343, "reward_std": 0.10659918934106827, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.20399491488933563, "step": 461 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1085.0, "completions/max_terminated_length": 1085.0, "completions/mean_length": 451.2749938964844, "completions/mean_terminated_length": 451.2749938964844, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.1792086889061288, "grad_norm": 0.5091977715492249, "kl": 0.233154296875, "learning_rate": 4.744394524554003e-06, "loss": -0.0055, "num_tokens": 41167790.0, "reward": 0.4350000321865082, "reward_std": 0.11977429687976837, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.33500000834465027, "rewards/unicoder_reward_fn/std": 0.34240400791168213, "step": 462 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1334.0, "completions/max_terminated_length": 1334.0, "completions/mean_length": 445.6875, "completions/mean_terminated_length": 445.6875, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.1795965865011637, "grad_norm": 0.6285807490348816, "kl": 0.26318359375, "learning_rate": 4.743084061201905e-06, "loss": 0.0297, "num_tokens": 41255837.0, "reward": 0.53125, "reward_std": 0.14136174321174622, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.43250003457069397, "rewards/unicoder_reward_fn/std": 0.33744198083877563, "step": 463 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 785.0, "completions/max_terminated_length": 785.0, "completions/mean_length": 357.2375183105469, "completions/mean_terminated_length": 357.2375183105469, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.1799844840961986, "grad_norm": 0.8075671195983887, "kl": 0.2900390625, "learning_rate": 4.741770450498024e-06, "loss": 0.0166, "num_tokens": 41322288.0, "reward": 0.49000006914138794, "reward_std": 0.10857420414686203, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.39000001549720764, "rewards/unicoder_reward_fn/std": 0.2641441822052002, "step": 464 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1071.0, "completions/max_terminated_length": 1071.0, "completions/mean_length": 392.8625183105469, "completions/mean_terminated_length": 392.8625183105469, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.1803723816912335, "grad_norm": 0.8102830052375793, "kl": 0.27685546875, "learning_rate": 4.740453694516734e-06, "loss": -0.0262, "num_tokens": 41399133.0, "reward": 0.38499999046325684, "reward_std": 0.14653538167476654, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2850000262260437, "rewards/unicoder_reward_fn/std": 0.282439649105072, "step": 465 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 603.0, "completions/max_terminated_length": 603.0, "completions/mean_length": 382.6125183105469, "completions/mean_terminated_length": 382.6125183105469, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.18076027928626842, "grad_norm": 0.858945906162262, "kl": 0.2783203125, "learning_rate": 4.73913379533737e-06, "loss": -0.0327, "num_tokens": 41475038.0, "reward": 0.48750001192092896, "reward_std": 0.13458256423473358, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.38749998807907104, "rewards/unicoder_reward_fn/std": 0.3633093237876892, "step": 466 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 832.0, "completions/max_terminated_length": 832.0, "completions/mean_length": 395.07501220703125, "completions/mean_terminated_length": 395.07501220703125, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.18114817688130333, "grad_norm": 0.8170744180679321, "kl": 0.26611328125, "learning_rate": 4.737810755044234e-06, "loss": -0.014, "num_tokens": 41554028.0, "reward": 0.42750000953674316, "reward_std": 0.11832928657531738, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.32750001549720764, "rewards/unicoder_reward_fn/std": 0.2164003551006317, "step": 467 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 709.0, "completions/max_terminated_length": 709.0, "completions/mean_length": 367.9875183105469, "completions/mean_terminated_length": 367.9875183105469, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.18153607447633824, "grad_norm": 0.6901174187660217, "kl": 0.28173828125, "learning_rate": 4.7364845757265865e-06, "loss": 0.0143, "num_tokens": 41630091.0, "reward": 0.30000001192092896, "reward_std": 0.08000855147838593, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.2134704440832138, "step": 468 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 728.0, "completions/max_terminated_length": 728.0, "completions/mean_length": 441.45001220703125, "completions/mean_terminated_length": 441.45001220703125, "completions/min_length": 217.0, "completions/min_terminated_length": 217.0, "epoch": 0.18192397207137315, "grad_norm": 0.6072164177894592, "kl": 0.27099609375, "learning_rate": 4.735155259478645e-06, "loss": 0.0053, "num_tokens": 41723903.0, "reward": 0.4675000309944153, "reward_std": 0.176797553896904, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3675000071525574, "rewards/unicoder_reward_fn/std": 0.30761435627937317, "step": 469 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 745.0, "completions/max_terminated_length": 745.0, "completions/mean_length": 382.5874938964844, "completions/mean_terminated_length": 382.5874938964844, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "epoch": 0.18231186966640806, "grad_norm": 0.6954526901245117, "kl": 0.249755859375, "learning_rate": 4.733822808399583e-06, "loss": -0.0002, "num_tokens": 41788230.0, "reward": 0.4949999749660492, "reward_std": 0.12056422233581543, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.39500001072883606, "rewards/unicoder_reward_fn/std": 0.2933439016342163, "step": 470 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 779.0, "completions/max_terminated_length": 779.0, "completions/mean_length": 413.0, "completions/mean_terminated_length": 413.0, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "epoch": 0.182699767261443, "grad_norm": 0.7071959972381592, "kl": 0.2734375, "learning_rate": 4.732487224593523e-06, "loss": 0.036, "num_tokens": 41872542.0, "reward": 0.48500004410743713, "reward_std": 0.10090887546539307, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3850000202655792, "rewards/unicoder_reward_fn/std": 0.31142789125442505, "step": 471 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 809.0, "completions/max_terminated_length": 809.0, "completions/mean_length": 413.4375, "completions/mean_terminated_length": 413.4375, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.1830876648564779, "grad_norm": 0.7965189814567566, "kl": 0.2744140625, "learning_rate": 4.7311485101695314e-06, "loss": 0.0222, "num_tokens": 41954761.0, "reward": 0.46000003814697266, "reward_std": 0.1814199686050415, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.35999998450279236, "rewards/unicoder_reward_fn/std": 0.2744845747947693, "step": 472 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1030.0, "completions/max_terminated_length": 1030.0, "completions/mean_length": 415.5, "completions/mean_terminated_length": 415.5, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.1834755624515128, "grad_norm": 0.47499600052833557, "kl": 0.2607421875, "learning_rate": 4.7298066672416235e-06, "loss": 0.0146, "num_tokens": 42036377.0, "reward": 0.41750001907348633, "reward_std": 0.09729490429162979, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3175000250339508, "rewards/unicoder_reward_fn/std": 0.33781692385673523, "step": 473 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 775.0, "completions/max_terminated_length": 775.0, "completions/mean_length": 410.9624938964844, "completions/mean_terminated_length": 410.9624938964844, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.18386346004654772, "grad_norm": 0.5670905709266663, "kl": 0.2802734375, "learning_rate": 4.728461697928752e-06, "loss": -0.0115, "num_tokens": 42128438.0, "reward": 0.41999998688697815, "reward_std": 0.07611346244812012, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3199999928474426, "rewards/unicoder_reward_fn/std": 0.29953551292419434, "step": 474 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 919.0, "completions/max_terminated_length": 919.0, "completions/mean_length": 424.8125, "completions/mean_terminated_length": 424.8125, "completions/min_length": 224.0, "completions/min_terminated_length": 224.0, "epoch": 0.18425135764158262, "grad_norm": 0.5764901041984558, "kl": 0.266845703125, "learning_rate": 4.727113604354806e-06, "loss": 0.0175, "num_tokens": 42211807.0, "reward": 0.4012499749660492, "reward_std": 0.09406192600727081, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.30250000953674316, "rewards/unicoder_reward_fn/std": 0.3117833435535431, "step": 475 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 777.0, "completions/max_terminated_length": 777.0, "completions/mean_length": 387.3000183105469, "completions/mean_terminated_length": 387.3000183105469, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.18463925523661753, "grad_norm": 0.7215948104858398, "kl": 0.3056640625, "learning_rate": 4.7257623886486105e-06, "loss": -0.0014, "num_tokens": 42300103.0, "reward": 0.36000001430511475, "reward_std": 0.06524034589529037, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.25999999046325684, "rewards/unicoder_reward_fn/std": 0.24935360252857208, "step": 476 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1054.0, "completions/max_terminated_length": 1054.0, "completions/mean_length": 405.88751220703125, "completions/mean_terminated_length": 405.88751220703125, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.18502715283165244, "grad_norm": 0.8317935466766357, "kl": 0.28564453125, "learning_rate": 4.724408052943918e-06, "loss": 0.0894, "num_tokens": 42385414.0, "reward": 0.51500004529953, "reward_std": 0.11474591493606567, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4150000214576721, "rewards/unicoder_reward_fn/std": 0.2521301507949829, "step": 477 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 556.0, "completions/mean_length": 392.1125183105469, "completions/mean_terminated_length": 345.22784423828125, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.18541505042668735, "grad_norm": 0.8636894226074219, "kl": 0.31494140625, "learning_rate": 4.723050599379411e-06, "loss": 0.065, "num_tokens": 42466767.0, "reward": 0.39124998450279236, "reward_std": 0.12330351024866104, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.29250001907348633, "rewards/unicoder_reward_fn/std": 0.23587971925735474, "step": 478 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1373.0, "completions/max_terminated_length": 1373.0, "completions/mean_length": 402.63751220703125, "completions/mean_terminated_length": 402.63751220703125, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.18580294802172226, "grad_norm": 0.6353549361228943, "kl": 0.27490234375, "learning_rate": 4.721690030098693e-06, "loss": 0.0095, "num_tokens": 42552850.0, "reward": 0.4462500214576721, "reward_std": 0.09241002798080444, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.3474999964237213, "rewards/unicoder_reward_fn/std": 0.2787744700908661, "step": 479 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1290.0, "completions/max_terminated_length": 1290.0, "completions/mean_length": 431.20001220703125, "completions/mean_terminated_length": 431.20001220703125, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.18619084561675717, "grad_norm": 0.4563101530075073, "kl": 0.2880859375, "learning_rate": 4.72032634725029e-06, "loss": 0.0304, "num_tokens": 42637226.0, "reward": 0.3475000560283661, "reward_std": 0.05277739092707634, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24750001728534698, "rewards/unicoder_reward_fn/std": 0.2510610520839691, "step": 480 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 927.0, "completions/max_terminated_length": 927.0, "completions/mean_length": 422.1499938964844, "completions/mean_terminated_length": 422.1499938964844, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.18657874321179208, "grad_norm": 0.6243096590042114, "kl": 0.29248046875, "learning_rate": 4.718959552987641e-06, "loss": 0.0269, "num_tokens": 42724894.0, "reward": 0.4525000751018524, "reward_std": 0.09185662120580673, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3525000214576721, "rewards/unicoder_reward_fn/std": 0.3186105191707611, "step": 481 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1187.0, "completions/max_terminated_length": 1187.0, "completions/mean_length": 442.57501220703125, "completions/mean_terminated_length": 442.57501220703125, "completions/min_length": 241.0, "completions/min_terminated_length": 241.0, "epoch": 0.186966640806827, "grad_norm": 0.40090879797935486, "kl": 0.2607421875, "learning_rate": 4.7175896494691025e-06, "loss": 0.045, "num_tokens": 42805628.0, "reward": 0.36500000953674316, "reward_std": 0.03156419098377228, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.26500001549720764, "rewards/unicoder_reward_fn/std": 0.2658161520957947, "step": 482 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 800.0, "completions/max_terminated_length": 800.0, "completions/mean_length": 382.70001220703125, "completions/mean_terminated_length": 382.70001220703125, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.18735453840186192, "grad_norm": 0.6534484624862671, "kl": 0.29638671875, "learning_rate": 4.716216638857939e-06, "loss": -0.0259, "num_tokens": 42876140.0, "reward": 0.5250000357627869, "reward_std": 0.08719932287931442, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.42500001192092896, "rewards/unicoder_reward_fn/std": 0.3619566857814789, "step": 483 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1053.0, "completions/mean_length": 465.6000061035156, "completions/mean_terminated_length": 419.64556884765625, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.18774243599689683, "grad_norm": 0.5177233815193176, "kl": 0.28466796875, "learning_rate": 4.7148405233223205e-06, "loss": 0.0425, "num_tokens": 42958284.0, "reward": 0.3725000023841858, "reward_std": 0.05019275099039078, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.27250000834465027, "rewards/unicoder_reward_fn/std": 0.28859108686447144, "step": 484 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 646.0, "completions/max_terminated_length": 646.0, "completions/mean_length": 398.38751220703125, "completions/mean_terminated_length": 398.38751220703125, "completions/min_length": 211.0, "completions/min_terminated_length": 211.0, "epoch": 0.18813033359193174, "grad_norm": 0.7649958729743958, "kl": 0.30615234375, "learning_rate": 4.7134613050353225e-06, "loss": -0.0092, "num_tokens": 43043179.0, "reward": 0.5100000500679016, "reward_std": 0.07312648743391037, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4099999964237213, "rewards/unicoder_reward_fn/std": 0.24628892540931702, "step": 485 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 883.0, "completions/max_terminated_length": 883.0, "completions/mean_length": 374.6750183105469, "completions/mean_terminated_length": 374.6750183105469, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.18851823118696664, "grad_norm": 0.6282624006271362, "kl": 0.29931640625, "learning_rate": 4.712078986174919e-06, "loss": -0.0165, "num_tokens": 43119801.0, "reward": 0.30000001192092896, "reward_std": 0.04678413271903992, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.2792394757270813, "step": 486 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 684.0, "completions/max_terminated_length": 684.0, "completions/mean_length": 369.9375, "completions/mean_terminated_length": 369.9375, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.18890612878200155, "grad_norm": 0.7222946286201477, "kl": 0.32080078125, "learning_rate": 4.710693568923979e-06, "loss": 0.0124, "num_tokens": 43193804.0, "reward": 0.3175000250339508, "reward_std": 0.0615093819797039, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2175000011920929, "rewards/unicoder_reward_fn/std": 0.16896933317184448, "step": 487 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 595.0, "completions/max_terminated_length": 595.0, "completions/mean_length": 378.1125183105469, "completions/mean_terminated_length": 378.1125183105469, "completions/min_length": 220.0, "completions/min_terminated_length": 220.0, "epoch": 0.18929402637703646, "grad_norm": 0.6388532519340515, "kl": 0.28955078125, "learning_rate": 4.709305055470266e-06, "loss": 0.0144, "num_tokens": 43262165.0, "reward": 0.30000001192092896, "reward_std": 0.04340285062789917, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.1492597758769989, "step": 488 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1058.0, "completions/max_terminated_length": 1058.0, "completions/mean_length": 424.4250183105469, "completions/mean_terminated_length": 424.4250183105469, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.18968192397207137, "grad_norm": 0.7211001515388489, "kl": 0.267578125, "learning_rate": 4.707913448006433e-06, "loss": 0.0126, "num_tokens": 43350863.0, "reward": 0.4000000059604645, "reward_std": 0.1009402647614479, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.30000001192092896, "rewards/unicoder_reward_fn/std": 0.22725661098957062, "step": 489 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1006.0, "completions/max_terminated_length": 1006.0, "completions/mean_length": 421.7250061035156, "completions/mean_terminated_length": 421.7250061035156, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.19006982156710628, "grad_norm": 0.9457219243049622, "kl": 0.326171875, "learning_rate": 4.7065187487300175e-06, "loss": 0.0648, "num_tokens": 43432865.0, "reward": 0.46000000834465027, "reward_std": 0.12343428283929825, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.36000001430511475, "rewards/unicoder_reward_fn/std": 0.2817665636539459, "step": 490 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 653.0, "completions/max_terminated_length": 653.0, "completions/mean_length": 409.5625, "completions/mean_terminated_length": 409.5625, "completions/min_length": 224.0, "completions/min_terminated_length": 224.0, "epoch": 0.1904577191621412, "grad_norm": 0.6139965057373047, "kl": 0.2822265625, "learning_rate": 4.70512095984344e-06, "loss": -0.0169, "num_tokens": 43509862.0, "reward": 0.5074999928474426, "reward_std": 0.07307395339012146, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4074999690055847, "rewards/unicoder_reward_fn/std": 0.30761435627937317, "step": 491 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 885.0, "completions/max_terminated_length": 885.0, "completions/mean_length": 464.01251220703125, "completions/mean_terminated_length": 464.01251220703125, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.1908456167571761, "grad_norm": 0.6002998352050781, "kl": 0.28466796875, "learning_rate": 4.703720083554e-06, "loss": 0.0347, "num_tokens": 43595991.0, "reward": 0.3575000464916229, "reward_std": 0.09666387736797333, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.257500022649765, "rewards/unicoder_reward_fn/std": 0.22768789529800415, "step": 492 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3503.0, "completions/mean_length": 490.26251220703125, "completions/mean_terminated_length": 444.6202697753906, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.191233514352211, "grad_norm": 0.7461252212524414, "kl": 0.335205078125, "learning_rate": 4.702316122073872e-06, "loss": 0.0762, "num_tokens": 43675460.0, "reward": 0.44875001907348633, "reward_std": 0.0792120024561882, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.3499999940395355, "rewards/unicoder_reward_fn/std": 0.27833136916160583, "step": 493 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1196.0, "completions/max_terminated_length": 1196.0, "completions/mean_length": 460.5625, "completions/mean_terminated_length": 460.5625, "completions/min_length": 196.0, "completions/min_terminated_length": 196.0, "epoch": 0.19162141194724594, "grad_norm": 0.632739782333374, "kl": 0.264404296875, "learning_rate": 4.700909077620103e-06, "loss": 0.0083, "num_tokens": 43755257.0, "reward": 0.4449999928474426, "reward_std": 0.1371445208787918, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3449999988079071, "rewards/unicoder_reward_fn/std": 0.30848753452301025, "step": 494 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 838.0, "completions/max_terminated_length": 838.0, "completions/mean_length": 470.51251220703125, "completions/mean_terminated_length": 470.51251220703125, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.19200930954228085, "grad_norm": 0.6458467841148376, "kl": 0.263427734375, "learning_rate": 4.699498952414609e-06, "loss": -0.0076, "num_tokens": 43841010.0, "reward": 0.5300000309944153, "reward_std": 0.12402965873479843, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4300000071525574, "rewards/unicoder_reward_fn/std": 0.2918297052383423, "step": 495 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 952.0, "completions/max_terminated_length": 952.0, "completions/mean_length": 470.4875183105469, "completions/mean_terminated_length": 470.4875183105469, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.19239720713731576, "grad_norm": 0.7101290225982666, "kl": 0.2939453125, "learning_rate": 4.69808574868417e-06, "loss": 0.0358, "num_tokens": 43933745.0, "reward": 0.26250001788139343, "reward_std": 0.09164317697286606, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.14615321159362793, "step": 496 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1387.0, "completions/max_terminated_length": 1387.0, "completions/mean_length": 533.3375244140625, "completions/mean_terminated_length": 533.3375244140625, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "epoch": 0.19278510473235067, "grad_norm": 0.4603882431983948, "kl": 0.265625, "learning_rate": 4.6966694686604266e-06, "loss": 0.0115, "num_tokens": 44023052.0, "reward": 0.3499999940395355, "reward_std": 0.08758517354726791, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.25, "rewards/unicoder_reward_fn/std": 0.25950294733047485, "step": 497 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 882.0, "completions/max_terminated_length": 882.0, "completions/mean_length": 473.4750061035156, "completions/mean_terminated_length": 473.4750061035156, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.19317300232738557, "grad_norm": 0.7246650457382202, "kl": 0.29345703125, "learning_rate": 4.695250114579878e-06, "loss": 0.0269, "num_tokens": 44119714.0, "reward": 0.4100000560283661, "reward_std": 0.12123612314462662, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3100000023841858, "rewards/unicoder_reward_fn/std": 0.31646785140037537, "step": 498 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 847.0, "completions/max_terminated_length": 847.0, "completions/mean_length": 507.2250061035156, "completions/mean_terminated_length": 507.2250061035156, "completions/min_length": 289.0, "completions/min_terminated_length": 289.0, "epoch": 0.19356089992242048, "grad_norm": 0.644055187702179, "kl": 0.24658203125, "learning_rate": 4.6938276886838795e-06, "loss": -0.0183, "num_tokens": 44204940.0, "reward": 0.5287500619888306, "reward_std": 0.08073337376117706, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.4300000071525574, "rewards/unicoder_reward_fn/std": 0.2664344906806946, "step": 499 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 838.0, "completions/max_terminated_length": 838.0, "completions/mean_length": 508.25, "completions/mean_terminated_length": 508.25, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "epoch": 0.1939487975174554, "grad_norm": 0.7662819623947144, "kl": 0.283935546875, "learning_rate": 4.692402193218636e-06, "loss": 0.0124, "num_tokens": 44286952.0, "reward": 0.29250001907348633, "reward_std": 0.12429986149072647, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.19985756278038025, "step": 500 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1706.0, "completions/max_terminated_length": 1706.0, "completions/mean_length": 542.6625366210938, "completions/mean_terminated_length": 542.6625366210938, "completions/min_length": 318.0, "completions/min_terminated_length": 318.0, "epoch": 0.1943366951124903, "grad_norm": 0.5580618381500244, "kl": 0.25634765625, "learning_rate": 4.690973630435198e-06, "loss": 0.043, "num_tokens": 44377541.0, "reward": 0.445000022649765, "reward_std": 0.0906301736831665, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3449999988079071, "rewards/unicoder_reward_fn/std": 0.2933439016342163, "step": 501 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 873.0, "completions/max_terminated_length": 873.0, "completions/mean_length": 511.75, "completions/mean_terminated_length": 511.75, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "epoch": 0.1947245927075252, "grad_norm": 0.6597313284873962, "kl": 0.27294921875, "learning_rate": 4.6895420025894615e-06, "loss": 0.0279, "num_tokens": 44464161.0, "reward": 0.3800000250339508, "reward_std": 0.09727052599191666, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2800000309944153, "rewards/unicoder_reward_fn/std": 0.2125195562839508, "step": 502 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1290.0, "completions/max_terminated_length": 1290.0, "completions/mean_length": 585.5875244140625, "completions/mean_terminated_length": 585.5875244140625, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.19511249030256012, "grad_norm": 0.5496182441711426, "kl": 0.2333984375, "learning_rate": 4.688107311942162e-06, "loss": 0.0072, "num_tokens": 44557568.0, "reward": 0.30750003457069397, "reward_std": 0.09109651297330856, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.21000000834465027, "rewards/unicoder_reward_fn/std": 0.2059863656759262, "step": 503 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1288.0, "completions/max_terminated_length": 1288.0, "completions/mean_length": 533.6749877929688, "completions/mean_terminated_length": 533.6749877929688, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.19550038789759502, "grad_norm": 0.4550407826900482, "kl": 0.27392578125, "learning_rate": 4.686669560758874e-06, "loss": -0.0053, "num_tokens": 44653398.0, "reward": 0.4150000214576721, "reward_std": 0.04390309005975723, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3149999976158142, "rewards/unicoder_reward_fn/std": 0.30318978428840637, "step": 504 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1096.0, "completions/max_terminated_length": 1096.0, "completions/mean_length": 560.8250122070312, "completions/mean_terminated_length": 560.8250122070312, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "epoch": 0.19588828549262993, "grad_norm": 0.5740103721618652, "kl": 0.26416015625, "learning_rate": 4.68522875131e-06, "loss": 0.0162, "num_tokens": 44742136.0, "reward": 0.35500001907348633, "reward_std": 0.07916396111249924, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2549999952316284, "rewards/unicoder_reward_fn/std": 0.2134111076593399, "step": 505 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2217.0, "completions/max_terminated_length": 2217.0, "completions/mean_length": 679.5375366210938, "completions/mean_terminated_length": 679.5375366210938, "completions/min_length": 279.0, "completions/min_terminated_length": 279.0, "epoch": 0.19627618308766487, "grad_norm": 0.4942134618759155, "kl": 0.219970703125, "learning_rate": 4.683784885870776e-06, "loss": 0.0229, "num_tokens": 44843395.0, "reward": 0.3174999952316284, "reward_std": 0.13707736134529114, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.2199999988079071, "rewards/unicoder_reward_fn/std": 0.2978403568267822, "step": 506 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1074.0, "completions/max_terminated_length": 1074.0, "completions/mean_length": 548.7625122070312, "completions/mean_terminated_length": 548.7625122070312, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.19666408068269978, "grad_norm": 0.6465409398078918, "kl": 0.255859375, "learning_rate": 4.682337966721263e-06, "loss": -0.0074, "num_tokens": 44936688.0, "reward": 0.3387500047683716, "reward_std": 0.06302955001592636, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.24000000953674316, "rewards/unicoder_reward_fn/std": 0.11205785721540451, "step": 507 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1054.0, "completions/max_terminated_length": 1054.0, "completions/mean_length": 543.6875, "completions/mean_terminated_length": 543.6875, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "epoch": 0.19705197827773469, "grad_norm": 0.5764495730400085, "kl": 0.24755859375, "learning_rate": 4.680887996146343e-06, "loss": 0.0066, "num_tokens": 45029255.0, "reward": 0.32749998569488525, "reward_std": 0.09134350717067719, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.22749999165534973, "rewards/unicoder_reward_fn/std": 0.21286176145076752, "step": 508 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 926.0, "completions/max_terminated_length": 926.0, "completions/mean_length": 559.7999877929688, "completions/mean_terminated_length": 559.7999877929688, "completions/min_length": 243.0, "completions/min_terminated_length": 243.0, "epoch": 0.1974398758727696, "grad_norm": 0.6762300729751587, "kl": 0.306396484375, "learning_rate": 4.679434976435718e-06, "loss": -0.0008, "num_tokens": 45112759.0, "reward": 0.4112500250339508, "reward_std": 0.08038812130689621, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.3125, "rewards/unicoder_reward_fn/std": 0.20829325914382935, "step": 509 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1145.0, "completions/max_terminated_length": 1145.0, "completions/mean_length": 616.125, "completions/mean_terminated_length": 616.125, "completions/min_length": 370.0, "completions/min_terminated_length": 370.0, "epoch": 0.1978277734678045, "grad_norm": 0.654362142086029, "kl": 0.234619140625, "learning_rate": 4.677978909883905e-06, "loss": 0.0074, "num_tokens": 45206657.0, "reward": 0.5825000405311584, "reward_std": 0.18214182555675507, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.48249998688697815, "rewards/unicoder_reward_fn/std": 0.3096649944782257, "step": 510 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1624.0, "completions/max_terminated_length": 1624.0, "completions/mean_length": 629.6500244140625, "completions/mean_terminated_length": 629.6500244140625, "completions/min_length": 283.0, "completions/min_terminated_length": 283.0, "epoch": 0.1982156710628394, "grad_norm": 0.48883697390556335, "kl": 0.223388671875, "learning_rate": 4.676519798790229e-06, "loss": 0.0067, "num_tokens": 45301341.0, "reward": 0.39000001549720764, "reward_std": 0.09158191084861755, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2900000214576721, "rewards/unicoder_reward_fn/std": 0.2622203230857849, "step": 511 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1426.0, "completions/max_terminated_length": 1426.0, "completions/mean_length": 646.4000244140625, "completions/mean_terminated_length": 646.4000244140625, "completions/min_length": 314.0, "completions/min_terminated_length": 314.0, "epoch": 0.19860356865787432, "grad_norm": 0.380204975605011, "kl": 0.239501953125, "learning_rate": 4.67505764545883e-06, "loss": 0.0222, "num_tokens": 45408661.0, "reward": 0.4625000059604645, "reward_std": 0.05465741083025932, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.36250001192092896, "rewards/unicoder_reward_fn/std": 0.2900785803794861, "step": 512 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1373.0, "completions/max_terminated_length": 1373.0, "completions/mean_length": 593.375, "completions/mean_terminated_length": 593.375, "completions/min_length": 295.0, "completions/min_terminated_length": 295.0, "epoch": 0.19899146625290923, "grad_norm": 0.7452918887138367, "kl": 0.243896484375, "learning_rate": 4.673592452198645e-06, "loss": 0.0308, "num_tokens": 45502179.0, "reward": 0.4412500560283661, "reward_std": 0.17311826348304749, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.3449999988079071, "rewards/unicoder_reward_fn/std": 0.2967759370803833, "step": 513 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1495.0, "completions/max_terminated_length": 1495.0, "completions/mean_length": 555.3375244140625, "completions/mean_terminated_length": 555.3375244140625, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.19937936384794414, "grad_norm": 0.6594035625457764, "kl": 0.248291015625, "learning_rate": 4.672124221323414e-06, "loss": -0.0177, "num_tokens": 45598606.0, "reward": 0.4400000274181366, "reward_std": 0.10445361584424973, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.3449999988079071, "rewards/unicoder_reward_fn/std": 0.26236510276794434, "step": 514 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1441.0, "completions/max_terminated_length": 1441.0, "completions/mean_length": 626.2750244140625, "completions/mean_terminated_length": 626.2750244140625, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "epoch": 0.19976726144297904, "grad_norm": 0.5217410326004028, "kl": 0.242431640625, "learning_rate": 4.670652955151675e-06, "loss": 0.0093, "num_tokens": 45698476.0, "reward": 0.30125001072883606, "reward_std": 0.05822935327887535, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.20500001311302185, "rewards/unicoder_reward_fn/std": 0.21102523803710938, "step": 515 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1241.0, "completions/max_terminated_length": 1241.0, "completions/mean_length": 619.8624877929688, "completions/mean_terminated_length": 619.8624877929688, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.20015515903801395, "grad_norm": 0.5444376468658447, "kl": 0.224853515625, "learning_rate": 4.669178656006757e-06, "loss": 0.0357, "num_tokens": 45796953.0, "reward": 0.5075000524520874, "reward_std": 0.09318845719099045, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.4124999940395355, "rewards/unicoder_reward_fn/std": 0.2834573984146118, "step": 516 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2014.0, "completions/max_terminated_length": 2014.0, "completions/mean_length": 666.6124877929688, "completions/mean_terminated_length": 666.6124877929688, "completions/min_length": 312.0, "completions/min_terminated_length": 312.0, "epoch": 0.2005430566330489, "grad_norm": 0.6134974360466003, "kl": 0.220703125, "learning_rate": 4.667701326216782e-06, "loss": -0.0144, "num_tokens": 45900138.0, "reward": 0.42125001549720764, "reward_std": 0.09540340304374695, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.33250004053115845, "rewards/unicoder_reward_fn/std": 0.2736416459083557, "step": 517 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1268.0, "completions/max_terminated_length": 1268.0, "completions/mean_length": 628.3624877929688, "completions/mean_terminated_length": 628.3624877929688, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "epoch": 0.2009309542280838, "grad_norm": 0.6776772737503052, "kl": 0.237548828125, "learning_rate": 4.666220968114653e-06, "loss": -0.017, "num_tokens": 46006983.0, "reward": 0.36375001072883606, "reward_std": 0.11187949031591415, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.27810388803482056, "step": 518 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2617.0, "completions/max_terminated_length": 2617.0, "completions/mean_length": 648.8375244140625, "completions/mean_terminated_length": 648.8375244140625, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.2013188518231187, "grad_norm": 0.6543768048286438, "kl": 0.213134765625, "learning_rate": 4.664737584038057e-06, "loss": -0.001, "num_tokens": 46102850.0, "reward": 0.5525000095367432, "reward_std": 0.1736598014831543, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.4625000059604645, "rewards/unicoder_reward_fn/std": 0.2557564973831177, "step": 519 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2297.0, "completions/max_terminated_length": 2297.0, "completions/mean_length": 655.9625244140625, "completions/mean_terminated_length": 655.9625244140625, "completions/min_length": 363.0, "completions/min_terminated_length": 363.0, "epoch": 0.2017067494181536, "grad_norm": 0.5052390098571777, "kl": 0.240234375, "learning_rate": 4.663251176329462e-06, "loss": -0.0139, "num_tokens": 46200463.0, "reward": 0.2837500274181366, "reward_std": 0.059011008590459824, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.1812579184770584, "step": 520 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1289.0, "completions/max_terminated_length": 1289.0, "completions/mean_length": 613.7000122070312, "completions/mean_terminated_length": 613.7000122070312, "completions/min_length": 193.0, "completions/min_terminated_length": 193.0, "epoch": 0.20209464701318852, "grad_norm": 0.653666079044342, "kl": 0.24365234375, "learning_rate": 4.661761747336108e-06, "loss": -0.0016, "num_tokens": 46293631.0, "reward": 0.5525000095367432, "reward_std": 0.10606699436903, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.4675000309944153, "rewards/unicoder_reward_fn/std": 0.3228727877140045, "step": 521 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1271.0, "completions/mean_length": 621.9500122070312, "completions/mean_terminated_length": 577.9746704101562, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "epoch": 0.20248254460822343, "grad_norm": 0.4952937066555023, "kl": 0.238525390625, "learning_rate": 4.660269299410005e-06, "loss": 0.0333, "num_tokens": 46400283.0, "reward": 0.35625001788139343, "reward_std": 0.0814070925116539, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.26500001549720764, "rewards/unicoder_reward_fn/std": 0.2677142024040222, "step": 522 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1867.0, "completions/max_terminated_length": 1867.0, "completions/mean_length": 607.5499877929688, "completions/mean_terminated_length": 607.5499877929688, "completions/min_length": 305.0, "completions/min_terminated_length": 305.0, "epoch": 0.20287044220325834, "grad_norm": 0.6798290610313416, "kl": 0.222412109375, "learning_rate": 4.6587738349079345e-06, "loss": 0.0307, "num_tokens": 46491063.0, "reward": 0.42375001311302185, "reward_std": 0.12331034243106842, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.33250001072883606, "rewards/unicoder_reward_fn/std": 0.2967226207256317, "step": 523 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1879.0, "completions/max_terminated_length": 1879.0, "completions/mean_length": 578.5625, "completions/mean_terminated_length": 578.5625, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "epoch": 0.20325833979829325, "grad_norm": 0.7584303617477417, "kl": 0.250732421875, "learning_rate": 4.657275356191437e-06, "loss": 0.0012, "num_tokens": 46594156.0, "reward": 0.471250057220459, "reward_std": 0.13301411271095276, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.38499999046325684, "rewards/unicoder_reward_fn/std": 0.3468118906021118, "step": 524 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2020.0, "completions/max_terminated_length": 2020.0, "completions/mean_length": 658.7750244140625, "completions/mean_terminated_length": 658.7750244140625, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "epoch": 0.20364623739332816, "grad_norm": 0.5372108221054077, "kl": 0.218017578125, "learning_rate": 4.655773865626817e-06, "loss": -0.0275, "num_tokens": 46688930.0, "reward": 0.3737500011920929, "reward_std": 0.046703726053237915, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.2775000035762787, "rewards/unicoder_reward_fn/std": 0.2993558347225189, "step": 525 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1021.0, "completions/max_terminated_length": 1021.0, "completions/mean_length": 643.6124877929688, "completions/mean_terminated_length": 643.6124877929688, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "epoch": 0.20403413498836306, "grad_norm": 0.603861927986145, "kl": 0.24609375, "learning_rate": 4.654269365585131e-06, "loss": 0.0152, "num_tokens": 46791299.0, "reward": 0.35500001907348633, "reward_std": 0.0594940260052681, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.25999999046325684, "rewards/unicoder_reward_fn/std": 0.221388041973114, "step": 526 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1210.0, "completions/max_terminated_length": 1210.0, "completions/mean_length": 606.4000244140625, "completions/mean_terminated_length": 606.4000244140625, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "epoch": 0.20442203258339797, "grad_norm": 0.5200044512748718, "kl": 0.246337890625, "learning_rate": 4.652761858442193e-06, "loss": 0.0142, "num_tokens": 46890251.0, "reward": 0.4037500321865082, "reward_std": 0.0447278767824173, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.3074999749660492, "rewards/unicoder_reward_fn/std": 0.3165378272533417, "step": 527 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1045.0, "completions/max_terminated_length": 1045.0, "completions/mean_length": 611.0750122070312, "completions/mean_terminated_length": 611.0750122070312, "completions/min_length": 311.0, "completions/min_terminated_length": 311.0, "epoch": 0.20480993017843288, "grad_norm": 0.47360873222351074, "kl": 0.232421875, "learning_rate": 4.651251346578558e-06, "loss": 0.0034, "num_tokens": 46987961.0, "reward": 0.3725000321865082, "reward_std": 0.09246983379125595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.23893223702907562, "step": 528 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1168.0, "completions/max_terminated_length": 1168.0, "completions/mean_length": 608.6875, "completions/mean_terminated_length": 608.6875, "completions/min_length": 284.0, "completions/min_terminated_length": 284.0, "epoch": 0.20519782777346782, "grad_norm": 0.686039388179779, "kl": 0.236328125, "learning_rate": 4.649737832379535e-06, "loss": 0.0687, "num_tokens": 47083496.0, "reward": 0.2874999940395355, "reward_std": 0.1640658676624298, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.1900000125169754, "rewards/unicoder_reward_fn/std": 0.2247924506664276, "step": 529 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1033.0, "completions/mean_length": 693.5499877929688, "completions/mean_terminated_length": 650.4810180664062, "completions/min_length": 264.0, "completions/min_terminated_length": 264.0, "epoch": 0.20558572536850273, "grad_norm": 0.6703917980194092, "kl": 0.22705078125, "learning_rate": 4.648221318235165e-06, "loss": 0.0671, "num_tokens": 47190924.0, "reward": 0.5475000739097595, "reward_std": 0.19510726630687714, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.44999998807907104, "rewards/unicoder_reward_fn/std": 0.2907868027687073, "step": 530 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1674.0, "completions/max_terminated_length": 1674.0, "completions/mean_length": 627.8624877929688, "completions/mean_terminated_length": 627.8624877929688, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "epoch": 0.20597362296353763, "grad_norm": 0.6032615900039673, "kl": 0.24658203125, "learning_rate": 4.646701806540235e-06, "loss": 0.0435, "num_tokens": 47282937.0, "reward": 0.33250004053115845, "reward_std": 0.08088989555835724, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23250000178813934, "rewards/unicoder_reward_fn/std": 0.1812579184770584, "step": 531 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1065.0, "completions/max_terminated_length": 1065.0, "completions/mean_length": 551.1375122070312, "completions/mean_terminated_length": 551.1375122070312, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "epoch": 0.20636152055857254, "grad_norm": 0.5498509407043457, "kl": 0.230712890625, "learning_rate": 4.645179299694258e-06, "loss": -0.0258, "num_tokens": 47365932.0, "reward": 0.36500000953674316, "reward_std": 0.053515732288360596, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.26500001549720764, "rewards/unicoder_reward_fn/std": 0.20812606811523438, "step": 532 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1327.0, "completions/max_terminated_length": 1327.0, "completions/mean_length": 658.9375, "completions/mean_terminated_length": 658.9375, "completions/min_length": 337.0, "completions/min_terminated_length": 337.0, "epoch": 0.20674941815360745, "grad_norm": 0.4525653123855591, "kl": 0.20947265625, "learning_rate": 4.643653800101483e-06, "loss": 0.0238, "num_tokens": 47477383.0, "reward": 0.3375000059604645, "reward_std": 0.06439042836427689, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23749999701976776, "rewards/unicoder_reward_fn/std": 0.21604911983013153, "step": 533 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2658.0, "completions/max_terminated_length": 2658.0, "completions/mean_length": 654.7000122070312, "completions/mean_terminated_length": 654.7000122070312, "completions/min_length": 362.0, "completions/min_terminated_length": 362.0, "epoch": 0.20713731574864236, "grad_norm": 0.6145139932632446, "kl": 0.221923828125, "learning_rate": 4.642125310170881e-06, "loss": 0.0129, "num_tokens": 47578135.0, "reward": 0.4400000274181366, "reward_std": 0.07582943886518478, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3400000035762787, "rewards/unicoder_reward_fn/std": 0.221388041973114, "step": 534 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1489.0, "completions/max_terminated_length": 1489.0, "completions/mean_length": 592.8875122070312, "completions/mean_terminated_length": 592.8875122070312, "completions/min_length": 316.0, "completions/min_terminated_length": 316.0, "epoch": 0.20752521334367727, "grad_norm": 0.5769472718238831, "kl": 0.220947265625, "learning_rate": 4.6405938323161455e-06, "loss": -0.0074, "num_tokens": 47663782.0, "reward": 0.3887500464916229, "reward_std": 0.06994383782148361, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.2900000214576721, "rewards/unicoder_reward_fn/std": 0.3499005436897278, "step": 535 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1457.0, "completions/max_terminated_length": 1457.0, "completions/mean_length": 550.5750122070312, "completions/mean_terminated_length": 550.5750122070312, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.20791311093871218, "grad_norm": 0.6774815917015076, "kl": 0.2333984375, "learning_rate": 4.639059368955689e-06, "loss": 0.0432, "num_tokens": 47747924.0, "reward": 0.47874999046325684, "reward_std": 0.10166894644498825, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.3800000250339508, "rewards/unicoder_reward_fn/std": 0.29442068934440613, "step": 536 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1132.0, "completions/max_terminated_length": 1132.0, "completions/mean_length": 558.6875, "completions/mean_terminated_length": 558.6875, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "epoch": 0.20830100853374708, "grad_norm": 0.44763725996017456, "kl": 0.23681640625, "learning_rate": 4.63752192251264e-06, "loss": -0.0124, "num_tokens": 47844915.0, "reward": 0.35750001668930054, "reward_std": 0.07479684054851532, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2574999928474426, "rewards/unicoder_reward_fn/std": 0.24275581538677216, "step": 537 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1132.0, "completions/max_terminated_length": 1132.0, "completions/mean_length": 552.9625244140625, "completions/mean_terminated_length": 552.9625244140625, "completions/min_length": 340.0, "completions/min_terminated_length": 340.0, "epoch": 0.208688906128782, "grad_norm": 0.5157929062843323, "kl": 0.236572265625, "learning_rate": 4.635981495414835e-06, "loss": 0.0062, "num_tokens": 47936840.0, "reward": 0.3775000274181366, "reward_std": 0.07945513725280762, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2775000035762787, "rewards/unicoder_reward_fn/std": 0.23491178452968597, "step": 538 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1201.0, "completions/max_terminated_length": 1201.0, "completions/mean_length": 568.0250244140625, "completions/mean_terminated_length": 568.0250244140625, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "epoch": 0.2090768037238169, "grad_norm": 0.617868185043335, "kl": 0.22802734375, "learning_rate": 4.634438090094819e-06, "loss": 0.0019, "num_tokens": 48026394.0, "reward": 0.5074999928474426, "reward_std": 0.13892637193202972, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4074999690055847, "rewards/unicoder_reward_fn/std": 0.28894177079200745, "step": 539 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1529.0, "completions/max_terminated_length": 1529.0, "completions/mean_length": 543.6749877929688, "completions/mean_terminated_length": 543.6749877929688, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "epoch": 0.20946470131885184, "grad_norm": 0.5653985142707825, "kl": 0.2236328125, "learning_rate": 4.63289170898984e-06, "loss": 0.0222, "num_tokens": 48127984.0, "reward": 0.3375000059604645, "reward_std": 0.1218092069029808, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23749999701976776, "rewards/unicoder_reward_fn/std": 0.21369268000125885, "step": 540 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1390.0, "completions/max_terminated_length": 1390.0, "completions/mean_length": 592.3125, "completions/mean_terminated_length": 592.3125, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "epoch": 0.20985259891388675, "grad_norm": 0.5873072743415833, "kl": 0.21826171875, "learning_rate": 4.6313423545418446e-06, "loss": 0.0458, "num_tokens": 48221969.0, "reward": 0.39875003695487976, "reward_std": 0.09053926914930344, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.30000001192092896, "rewards/unicoder_reward_fn/std": 0.23813621699810028, "step": 541 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2550.0, "completions/max_terminated_length": 2550.0, "completions/mean_length": 630.8500366210938, "completions/mean_terminated_length": 630.8500366210938, "completions/min_length": 211.0, "completions/min_terminated_length": 211.0, "epoch": 0.21024049650892165, "grad_norm": 0.5880686044692993, "kl": 0.2255859375, "learning_rate": 4.629790029197476e-06, "loss": -0.0506, "num_tokens": 48322125.0, "reward": 0.518750011920929, "reward_std": 0.13563796877861023, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.41999998688697815, "rewards/unicoder_reward_fn/std": 0.3045644462108612, "step": 542 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1863.0, "completions/max_terminated_length": 1863.0, "completions/mean_length": 616.25, "completions/mean_terminated_length": 616.25, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.21062839410395656, "grad_norm": 0.4993744194507599, "kl": 0.21240234375, "learning_rate": 4.628234735408065e-06, "loss": -0.0015, "num_tokens": 48424865.0, "reward": 0.4650000035762787, "reward_std": 0.11819902807474136, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.3675000071525574, "rewards/unicoder_reward_fn/std": 0.3125132918357849, "step": 543 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3569.0, "completions/mean_length": 736.5250244140625, "completions/mean_terminated_length": 694.0, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.21101629169899147, "grad_norm": 0.675476610660553, "kl": 0.261474609375, "learning_rate": 4.626676475629636e-06, "loss": 0.0494, "num_tokens": 48535819.0, "reward": 0.4675000309944153, "reward_std": 0.13506607711315155, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3675000071525574, "rewards/unicoder_reward_fn/std": 0.28004294633865356, "step": 544 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1465.0, "completions/max_terminated_length": 1465.0, "completions/mean_length": 601.6124877929688, "completions/mean_terminated_length": 601.6124877929688, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.21140418929402638, "grad_norm": 0.48193347454071045, "kl": 0.24267578125, "learning_rate": 4.625115252322891e-06, "loss": 0.023, "num_tokens": 48635628.0, "reward": 0.4100000560283661, "reward_std": 0.07986899465322495, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3100000023841858, "rewards/unicoder_reward_fn/std": 0.2602822184562683, "step": 545 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1176.0, "completions/max_terminated_length": 1176.0, "completions/mean_length": 534.9750366210938, "completions/mean_terminated_length": 534.9750366210938, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "epoch": 0.2117920868890613, "grad_norm": 0.7251015305519104, "kl": 0.239990234375, "learning_rate": 4.6235510679532176e-06, "loss": 0.0305, "num_tokens": 48726026.0, "reward": 0.5050000548362732, "reward_std": 0.1503053605556488, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4050000309944153, "rewards/unicoder_reward_fn/std": 0.3778352737426758, "step": 546 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1345.0, "completions/max_terminated_length": 1345.0, "completions/mean_length": 552.3250122070312, "completions/mean_terminated_length": 552.3250122070312, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.2121799844840962, "grad_norm": 0.6834052205085754, "kl": 0.256591796875, "learning_rate": 4.621983924990674e-06, "loss": 0.0187, "num_tokens": 48818724.0, "reward": 0.38999998569488525, "reward_std": 0.12879151105880737, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2900000214576721, "rewards/unicoder_reward_fn/std": 0.31163105368614197, "step": 547 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1263.0, "completions/max_terminated_length": 1263.0, "completions/mean_length": 543.4500122070312, "completions/mean_terminated_length": 543.4500122070312, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.2125678820791311, "grad_norm": 0.596498966217041, "kl": 0.23779296875, "learning_rate": 4.620413825909995e-06, "loss": 0.031, "num_tokens": 48904384.0, "reward": 0.39875003695487976, "reward_std": 0.09679745882749557, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.30000001192092896, "rewards/unicoder_reward_fn/std": 0.24649441242218018, "step": 548 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1286.0, "completions/max_terminated_length": 1286.0, "completions/mean_length": 532.9000244140625, "completions/mean_terminated_length": 532.9000244140625, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "epoch": 0.212955779674166, "grad_norm": 0.5057166218757629, "kl": 0.2529296875, "learning_rate": 4.618840773190579e-06, "loss": -0.0088, "num_tokens": 48992960.0, "reward": 0.49250003695487976, "reward_std": 0.09439100325107574, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.39250001311302185, "rewards/unicoder_reward_fn/std": 0.24944241344928741, "step": 549 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1849.0, "completions/max_terminated_length": 1849.0, "completions/mean_length": 546.7374877929688, "completions/mean_terminated_length": 546.7374877929688, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.21334367726920092, "grad_norm": 0.6398712396621704, "kl": 0.239013671875, "learning_rate": 4.617264769316493e-06, "loss": -0.0123, "num_tokens": 49087187.0, "reward": 0.35499998927116394, "reward_std": 0.155701145529747, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2549999952316284, "rewards/unicoder_reward_fn/std": 0.22041389346122742, "step": 550 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1955.0, "completions/max_terminated_length": 1955.0, "completions/mean_length": 560.4500122070312, "completions/mean_terminated_length": 560.4500122070312, "completions/min_length": 279.0, "completions/min_terminated_length": 279.0, "epoch": 0.21373157486423583, "grad_norm": 0.6791058778762817, "kl": 0.234130859375, "learning_rate": 4.61568581677646e-06, "loss": -0.0232, "num_tokens": 49179615.0, "reward": 0.4437499940395355, "reward_std": 0.1427289992570877, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.3450000286102295, "rewards/unicoder_reward_fn/std": 0.26809218525886536, "step": 551 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3489.0, "completions/max_terminated_length": 3489.0, "completions/mean_length": 690.0625, "completions/mean_terminated_length": 690.0625, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.21411947245927077, "grad_norm": 0.676084578037262, "kl": 0.229736328125, "learning_rate": 4.614103918063865e-06, "loss": 0.0044, "num_tokens": 49284820.0, "reward": 0.42125001549720764, "reward_std": 0.1293049305677414, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.3225000202655792, "rewards/unicoder_reward_fn/std": 0.28550422191619873, "step": 552 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2107.0, "completions/max_terminated_length": 2107.0, "completions/mean_length": 722.0375366210938, "completions/mean_terminated_length": 722.0375366210938, "completions/min_length": 361.0, "completions/min_terminated_length": 361.0, "epoch": 0.21450737005430567, "grad_norm": 0.5595484972000122, "kl": 0.222412109375, "learning_rate": 4.612519075676738e-06, "loss": 0.0289, "num_tokens": 49390023.0, "reward": 0.42750000953674316, "reward_std": 0.08162443339824677, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.32750001549720764, "rewards/unicoder_reward_fn/std": 0.20682960748672485, "step": 553 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2118.0, "completions/mean_length": 888.5125122070312, "completions/mean_terminated_length": 763.5454711914062, "completions/min_length": 290.0, "completions/min_terminated_length": 290.0, "epoch": 0.21489526764934058, "grad_norm": 0.4965727925300598, "kl": 0.208740234375, "learning_rate": 4.610931292117764e-06, "loss": -0.001, "num_tokens": 49501800.0, "reward": 0.4375, "reward_std": 0.11654721945524216, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.33750003576278687, "rewards/unicoder_reward_fn/std": 0.2027500867843628, "step": 554 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2556.0, "completions/max_terminated_length": 2556.0, "completions/mean_length": 688.6500244140625, "completions/mean_terminated_length": 688.6500244140625, "completions/min_length": 309.0, "completions/min_terminated_length": 309.0, "epoch": 0.2152831652443755, "grad_norm": 0.584362268447876, "kl": 0.2294921875, "learning_rate": 4.6093405698942704e-06, "loss": 0.0186, "num_tokens": 49605044.0, "reward": 0.4425000250339508, "reward_std": 0.1066458448767662, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3425000309944153, "rewards/unicoder_reward_fn/std": 0.25890472531318665, "step": 555 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2079.0, "completions/mean_length": 740.4750366210938, "completions/mean_terminated_length": 654.4359130859375, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.2156710628394104, "grad_norm": 0.5104061365127563, "kl": 0.241943359375, "learning_rate": 4.607746911518223e-06, "loss": -0.0609, "num_tokens": 49705818.0, "reward": 0.3700000047683716, "reward_std": 0.09923424571752548, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.27000001072883606, "rewards/unicoder_reward_fn/std": 0.21132494509220123, "step": 556 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3786.0, "completions/mean_length": 1006.4874877929688, "completions/mean_terminated_length": 886.1168823242188, "completions/min_length": 268.0, "completions/min_terminated_length": 268.0, "epoch": 0.2160589604344453, "grad_norm": 0.4205986559391022, "kl": 0.202880859375, "learning_rate": 4.606150319506227e-06, "loss": -0.0145, "num_tokens": 49845617.0, "reward": 0.39500001072883606, "reward_std": 0.13249912858009338, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.29499998688697815, "rewards/unicoder_reward_fn/std": 0.23379100859165192, "step": 557 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 4015.0, "completions/mean_length": 1197.3125, "completions/mean_terminated_length": 1044.75, "completions/min_length": 294.0, "completions/min_terminated_length": 294.0, "epoch": 0.21644685802948022, "grad_norm": 0.46892765164375305, "kl": 0.21630859375, "learning_rate": 4.604550796379518e-06, "loss": 0.0119, "num_tokens": 49991850.0, "reward": 0.4449999928474426, "reward_std": 0.10663008689880371, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3449999988079071, "rewards/unicoder_reward_fn/std": 0.25452861189842224, "step": 558 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2556.0, "completions/mean_length": 873.9500122070312, "completions/mean_terminated_length": 833.16455078125, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.21683475562451512, "grad_norm": 0.999720573425293, "kl": 0.331787109375, "learning_rate": 4.602948344663962e-06, "loss": -0.0178, "num_tokens": 50102422.0, "reward": 0.4675000309944153, "reward_std": 0.07661308348178864, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3675000071525574, "rewards/unicoder_reward_fn/std": 0.2782290279865265, "step": 559 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3310.0, "completions/mean_length": 1248.425048828125, "completions/mean_terminated_length": 1137.48046875, "completions/min_length": 331.0, "completions/min_terminated_length": 331.0, "epoch": 0.21722265321955003, "grad_norm": 0.30969592928886414, "kl": 0.20458984375, "learning_rate": 4.601342966890049e-06, "loss": -0.0005, "num_tokens": 50247896.0, "reward": 0.4000000059604645, "reward_std": 0.05547168478369713, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.30000001192092896, "rewards/unicoder_reward_fn/std": 0.2134704291820526, "step": 560 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3480.0, "completions/mean_length": 1334.112548828125, "completions/mean_terminated_length": 1263.294921875, "completions/min_length": 303.0, "completions/min_terminated_length": 303.0, "epoch": 0.21761055081458494, "grad_norm": 0.3814685344696045, "kl": 0.21826171875, "learning_rate": 4.599734665592891e-06, "loss": 0.0156, "num_tokens": 50409529.0, "reward": 0.4337500035762787, "reward_std": 0.10134851932525635, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.3349999785423279, "rewards/unicoder_reward_fn/std": 0.2619788348674774, "step": 561 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3305.0, "completions/mean_length": 1553.25, "completions/mean_terminated_length": 1347.0810546875, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "epoch": 0.21799844840961985, "grad_norm": 0.44054940342903137, "kl": 0.203369140625, "learning_rate": 4.598123443312213e-06, "loss": 0.114, "num_tokens": 50582245.0, "reward": 0.48000001907348633, "reward_std": 0.1736411601305008, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.3850000202655792, "rewards/unicoder_reward_fn/std": 0.29643452167510986, "step": 562 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3913.0, "completions/mean_length": 1674.2874755859375, "completions/mean_terminated_length": 1405.2083740234375, "completions/min_length": 406.0, "completions/min_terminated_length": 406.0, "epoch": 0.21838634600465476, "grad_norm": 0.37509965896606445, "kl": 0.218505859375, "learning_rate": 4.596509302592356e-06, "loss": 0.0734, "num_tokens": 50765020.0, "reward": 0.4650000035762787, "reward_std": 0.1666102111339569, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.3700000047683716, "rewards/unicoder_reward_fn/std": 0.29008948802948, "step": 563 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3758.0, "completions/mean_length": 1885.4500732421875, "completions/mean_terminated_length": 1605.2393798828125, "completions/min_length": 414.0, "completions/min_terminated_length": 414.0, "epoch": 0.2187742435996897, "grad_norm": 0.3203367590904236, "kl": 0.1904296875, "learning_rate": 4.594892245982268e-06, "loss": 0.0218, "num_tokens": 50964600.0, "reward": 0.4387500286102295, "reward_std": 0.16107919812202454, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.3499999940395355, "rewards/unicoder_reward_fn/std": 0.32217612862586975, "step": 564 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3919.0, "completions/mean_length": 1650.4000244140625, "completions/mean_terminated_length": 1452.108154296875, "completions/min_length": 375.0, "completions/min_terminated_length": 375.0, "epoch": 0.2191621411947246, "grad_norm": 0.44307515025138855, "kl": 0.231201171875, "learning_rate": 4.593272276035503e-06, "loss": 0.1663, "num_tokens": 51151768.0, "reward": 0.30250003933906555, "reward_std": 0.1416790783405304, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.21000000834465027, "rewards/unicoder_reward_fn/std": 0.24833625555038452, "step": 565 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3369.0, "completions/mean_length": 1141.9375, "completions/mean_terminated_length": 1104.5443115234375, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.2195500387897595, "grad_norm": 0.4439795911312103, "kl": 0.235107421875, "learning_rate": 4.591649395310214e-06, "loss": -0.0408, "num_tokens": 51280867.0, "reward": 0.36625000834465027, "reward_std": 0.15803836286067963, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.27000001072883606, "rewards/unicoder_reward_fn/std": 0.27942073345184326, "step": 566 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2766.0, "completions/max_terminated_length": 2766.0, "completions/mean_length": 1119.362548828125, "completions/mean_terminated_length": 1119.362548828125, "completions/min_length": 323.0, "completions/min_terminated_length": 323.0, "epoch": 0.21993793638479442, "grad_norm": 0.36537063121795654, "kl": 0.26318359375, "learning_rate": 4.590023606369152e-06, "loss": -0.0356, "num_tokens": 51421912.0, "reward": 0.4074999988079071, "reward_std": 0.07853256165981293, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.30750003457069397, "rewards/unicoder_reward_fn/std": 0.2156972885131836, "step": 567 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3428.0, "completions/mean_length": 1264.1500244140625, "completions/mean_terminated_length": 1228.3038330078125, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.22032583397982933, "grad_norm": 0.48278558254241943, "kl": 0.219970703125, "learning_rate": 4.588394911779658e-06, "loss": -0.0458, "num_tokens": 51580596.0, "reward": 0.4375, "reward_std": 0.13448910415172577, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3375000059604645, "rewards/unicoder_reward_fn/std": 0.3127562403678894, "step": 568 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3263.0, "completions/mean_length": 1179.3375244140625, "completions/mean_terminated_length": 1104.55126953125, "completions/min_length": 325.0, "completions/min_terminated_length": 325.0, "epoch": 0.22071373157486424, "grad_norm": 0.395449161529541, "kl": 0.23974609375, "learning_rate": 4.586763314113665e-06, "loss": 0.0527, "num_tokens": 51724615.0, "reward": 0.26750001311302185, "reward_std": 0.07161181420087814, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.1426689624786377, "step": 569 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2531.0, "completions/mean_length": 1044.5999755859375, "completions/mean_terminated_length": 1005.9747314453125, "completions/min_length": 273.0, "completions/min_terminated_length": 273.0, "epoch": 0.22110162916989914, "grad_norm": 0.4688416123390198, "kl": 0.243896484375, "learning_rate": 4.585128815947686e-06, "loss": 0.0308, "num_tokens": 51856527.0, "reward": 0.4737500250339508, "reward_std": 0.18266718089580536, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.375, "rewards/unicoder_reward_fn/std": 0.288821280002594, "step": 570 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2613.0, "completions/max_terminated_length": 2613.0, "completions/mean_length": 862.7125244140625, "completions/mean_terminated_length": 862.7125244140625, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "epoch": 0.22148952676493405, "grad_norm": 0.44258853793144226, "kl": 0.298828125, "learning_rate": 4.58349141986282e-06, "loss": 0.0002, "num_tokens": 51972456.0, "reward": 0.33000001311302185, "reward_std": 0.05604233220219612, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.22999998927116394, "rewards/unicoder_reward_fn/std": 0.18029512465000153, "step": 571 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2801.0, "completions/max_terminated_length": 2801.0, "completions/mean_length": 1059.9749755859375, "completions/mean_terminated_length": 1059.9749755859375, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "epoch": 0.22187742435996896, "grad_norm": 0.36763256788253784, "kl": 0.229248046875, "learning_rate": 4.581851128444735e-06, "loss": -0.0496, "num_tokens": 52108038.0, "reward": 0.5100000500679016, "reward_std": 0.12340810149908066, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.4124999940395355, "rewards/unicoder_reward_fn/std": 0.32193049788475037, "step": 572 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3417.0, "completions/max_terminated_length": 3417.0, "completions/mean_length": 1016.5875244140625, "completions/mean_terminated_length": 1016.5875244140625, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.22226532195500387, "grad_norm": 0.4397162199020386, "kl": 0.23583984375, "learning_rate": 4.580207944283679e-06, "loss": -0.0921, "num_tokens": 52232221.0, "reward": 0.2775000035762787, "reward_std": 0.11064688116312027, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.18824639916419983, "step": 573 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4079.0, "completions/max_terminated_length": 4079.0, "completions/mean_length": 1167.7125244140625, "completions/mean_terminated_length": 1167.7125244140625, "completions/min_length": 359.0, "completions/min_terminated_length": 359.0, "epoch": 0.22265321955003878, "grad_norm": 0.36382895708084106, "kl": 0.234375, "learning_rate": 4.578561869974462e-06, "loss": -0.0072, "num_tokens": 52370646.0, "reward": 0.38750001788139343, "reward_std": 0.10792883485555649, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.22240060567855835, "step": 574 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3294.0, "completions/mean_length": 1088.2874755859375, "completions/mean_terminated_length": 1050.2152099609375, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "epoch": 0.22304111714507371, "grad_norm": 0.5025940537452698, "kl": 0.24365234375, "learning_rate": 4.57691290811646e-06, "loss": 0.021, "num_tokens": 52499341.0, "reward": 0.3487500250339508, "reward_std": 0.15361309051513672, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.25, "rewards/unicoder_reward_fn/std": 0.2349252700805664, "step": 575 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3098.0, "completions/max_terminated_length": 3098.0, "completions/mean_length": 1150.050048828125, "completions/mean_terminated_length": 1150.050048828125, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.22342901474010862, "grad_norm": 0.4728158414363861, "kl": 0.234619140625, "learning_rate": 4.575261061313608e-06, "loss": -0.0121, "num_tokens": 52653793.0, "reward": 0.39750000834465027, "reward_std": 0.12582862377166748, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.29750001430511475, "rewards/unicoder_reward_fn/std": 0.2718781530857086, "step": 576 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3433.0, "completions/max_terminated_length": 3433.0, "completions/mean_length": 1143.4749755859375, "completions/mean_terminated_length": 1143.4749755859375, "completions/min_length": 334.0, "completions/min_terminated_length": 334.0, "epoch": 0.22381691233514353, "grad_norm": 0.455452561378479, "kl": 0.25390625, "learning_rate": 4.573606332174401e-06, "loss": 0.0811, "num_tokens": 52801247.0, "reward": 0.3800000250339508, "reward_std": 0.12387456744909286, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.2825000286102295, "rewards/unicoder_reward_fn/std": 0.2768608033657074, "step": 577 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2233.0, "completions/max_terminated_length": 2233.0, "completions/mean_length": 1007.9375, "completions/mean_terminated_length": 1007.9375, "completions/min_length": 320.0, "completions/min_terminated_length": 320.0, "epoch": 0.22420480993017844, "grad_norm": 0.5193778276443481, "kl": 0.245849609375, "learning_rate": 4.571948723311878e-06, "loss": -0.0026, "num_tokens": 52926362.0, "reward": 0.49500003457069397, "reward_std": 0.15581417083740234, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.39500001072883606, "rewards/unicoder_reward_fn/std": 0.25452861189842224, "step": 578 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3034.0, "completions/max_terminated_length": 3034.0, "completions/mean_length": 1188.9000244140625, "completions/mean_terminated_length": 1188.9000244140625, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "epoch": 0.22459270752521335, "grad_norm": 0.428622305393219, "kl": 0.23974609375, "learning_rate": 4.570288237343632e-06, "loss": 0.042, "num_tokens": 53066130.0, "reward": 0.36125001311302185, "reward_std": 0.13538940250873566, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.26249998807907104, "rewards/unicoder_reward_fn/std": 0.2456585168838501, "step": 579 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2353.0, "completions/max_terminated_length": 2353.0, "completions/mean_length": 1076.6375732421875, "completions/mean_terminated_length": 1076.6375732421875, "completions/min_length": 414.0, "completions/min_terminated_length": 414.0, "epoch": 0.22498060512024826, "grad_norm": 0.43520498275756836, "kl": 0.23681640625, "learning_rate": 4.568624876891796e-06, "loss": 0.0184, "num_tokens": 53198253.0, "reward": 0.5175000429153442, "reward_std": 0.12504105269908905, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.41999998688697815, "rewards/unicoder_reward_fn/std": 0.31114327907562256, "step": 580 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2584.0, "completions/max_terminated_length": 2584.0, "completions/mean_length": 855.6749877929688, "completions/mean_terminated_length": 855.6749877929688, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "epoch": 0.22536850271528316, "grad_norm": 0.5108845233917236, "kl": 0.2958984375, "learning_rate": 4.566958644583045e-06, "loss": 0.0467, "num_tokens": 53311659.0, "reward": 0.4425000250339508, "reward_std": 0.11567151546478271, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.3450000286102295, "rewards/unicoder_reward_fn/std": 0.26428791880607605, "step": 581 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3004.0, "completions/max_terminated_length": 3004.0, "completions/mean_length": 969.9125366210938, "completions/mean_terminated_length": 969.9125366210938, "completions/min_length": 334.0, "completions/min_terminated_length": 334.0, "epoch": 0.22575640031031807, "grad_norm": 0.495329886674881, "kl": 0.26904296875, "learning_rate": 4.565289543048586e-06, "loss": 0.0236, "num_tokens": 53441740.0, "reward": 0.33125001192092896, "reward_std": 0.10817108303308487, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.23499998450279236, "rewards/unicoder_reward_fn/std": 0.24396513402462006, "step": 582 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2174.0, "completions/max_terminated_length": 2174.0, "completions/mean_length": 960.6875, "completions/mean_terminated_length": 960.6875, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "epoch": 0.22614429790535298, "grad_norm": 0.46304237842559814, "kl": 0.271484375, "learning_rate": 4.563617574924156e-06, "loss": 0.0504, "num_tokens": 53569811.0, "reward": 0.29249998927116394, "reward_std": 0.10037253051996231, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.1812579184770584, "step": 583 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1892.0, "completions/max_terminated_length": 1892.0, "completions/mean_length": 866.7125244140625, "completions/mean_terminated_length": 866.7125244140625, "completions/min_length": 354.0, "completions/min_terminated_length": 354.0, "epoch": 0.2265321955003879, "grad_norm": 0.5264309048652649, "kl": 0.260498046875, "learning_rate": 4.561942742850024e-06, "loss": -0.0047, "num_tokens": 53696252.0, "reward": 0.3712500333786011, "reward_std": 0.15411750972270966, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.2724999785423279, "rewards/unicoder_reward_fn/std": 0.2343723177909851, "step": 584 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1856.0, "completions/max_terminated_length": 1856.0, "completions/mean_length": 860.5, "completions/mean_terminated_length": 860.5, "completions/min_length": 352.0, "completions/min_terminated_length": 352.0, "epoch": 0.2269200930954228, "grad_norm": 0.6602587699890137, "kl": 0.314453125, "learning_rate": 4.560265049470979e-06, "loss": -0.0309, "num_tokens": 53819412.0, "reward": 0.45124998688697815, "reward_std": 0.14869852364063263, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.35500001907348633, "rewards/unicoder_reward_fn/std": 0.3306686580181122, "step": 585 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1430.0, "completions/max_terminated_length": 1430.0, "completions/mean_length": 750.5499877929688, "completions/mean_terminated_length": 750.5499877929688, "completions/min_length": 327.0, "completions/min_terminated_length": 327.0, "epoch": 0.2273079906904577, "grad_norm": 0.5509629249572754, "kl": 0.2802734375, "learning_rate": 4.558584497436324e-06, "loss": -0.0338, "num_tokens": 53923440.0, "reward": 0.5287500023841858, "reward_std": 0.23699842393398285, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.4300000071525574, "rewards/unicoder_reward_fn/std": 0.3086926341056824, "step": 586 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2291.0, "completions/max_terminated_length": 2291.0, "completions/mean_length": 850.3500366210938, "completions/mean_terminated_length": 850.3500366210938, "completions/min_length": 231.0, "completions/min_terminated_length": 231.0, "epoch": 0.22769588828549264, "grad_norm": 0.5053699612617493, "kl": 0.279296875, "learning_rate": 4.556901089399884e-06, "loss": -0.0085, "num_tokens": 54047700.0, "reward": 0.41750001907348633, "reward_std": 0.14163833856582642, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3175000250339508, "rewards/unicoder_reward_fn/std": 0.325605571269989, "step": 587 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2175.0, "completions/max_terminated_length": 2175.0, "completions/mean_length": 884.6875, "completions/mean_terminated_length": 884.6875, "completions/min_length": 254.0, "completions/min_terminated_length": 254.0, "epoch": 0.22808378588052755, "grad_norm": 0.5607403516769409, "kl": 0.3388671875, "learning_rate": 4.555214828019988e-06, "loss": 0.0132, "num_tokens": 54165491.0, "reward": 0.3487500250339508, "reward_std": 0.1256590336561203, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.2524999976158142, "rewards/unicoder_reward_fn/std": 0.222172811627388, "step": 588 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2135.0, "completions/max_terminated_length": 2135.0, "completions/mean_length": 956.3125, "completions/mean_terminated_length": 956.3125, "completions/min_length": 381.0, "completions/min_terminated_length": 381.0, "epoch": 0.22847168347556246, "grad_norm": 0.42481759190559387, "kl": 0.29296875, "learning_rate": 4.553525715959472e-06, "loss": 0.0218, "num_tokens": 54288612.0, "reward": 0.2825000286102295, "reward_std": 0.12134983390569687, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.23886600136756897, "step": 589 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2747.0, "completions/max_terminated_length": 2747.0, "completions/mean_length": 940.2125244140625, "completions/mean_terminated_length": 940.2125244140625, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.22885958107059737, "grad_norm": 0.4048309922218323, "kl": 0.2880859375, "learning_rate": 4.551833755885677e-06, "loss": 0.0202, "num_tokens": 54417909.0, "reward": 0.35500001907348633, "reward_std": 0.06517866253852844, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2549999952316284, "rewards/unicoder_reward_fn/std": 0.20119895040988922, "step": 590 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1698.0, "completions/max_terminated_length": 1698.0, "completions/mean_length": 980.2999877929688, "completions/mean_terminated_length": 980.2999877929688, "completions/min_length": 374.0, "completions/min_terminated_length": 374.0, "epoch": 0.22924747866563228, "grad_norm": 0.4572540521621704, "kl": 0.28271484375, "learning_rate": 4.550138950470438e-06, "loss": 0.011, "num_tokens": 54545237.0, "reward": 0.3137499988079071, "reward_std": 0.12668417394161224, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.2175000011920929, "rewards/unicoder_reward_fn/std": 0.25694161653518677, "step": 591 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2074.0, "completions/max_terminated_length": 2074.0, "completions/mean_length": 997.5875244140625, "completions/mean_terminated_length": 997.5875244140625, "completions/min_length": 328.0, "completions/min_terminated_length": 328.0, "epoch": 0.22963537626066718, "grad_norm": 0.5475422739982605, "kl": 0.265380859375, "learning_rate": 4.548441302390084e-06, "loss": 0.0153, "num_tokens": 54669476.0, "reward": 0.5562500357627869, "reward_std": 0.18770956993103027, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.45750004053115845, "rewards/unicoder_reward_fn/std": 0.31693750619888306, "step": 592 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2282.0, "completions/max_terminated_length": 2282.0, "completions/mean_length": 1030.5125732421875, "completions/mean_terminated_length": 1030.5125732421875, "completions/min_length": 391.0, "completions/min_terminated_length": 391.0, "epoch": 0.2300232738557021, "grad_norm": 0.5022130608558655, "kl": 0.27685546875, "learning_rate": 4.546740814325432e-06, "loss": 0.0122, "num_tokens": 54796901.0, "reward": 0.3050000071525574, "reward_std": 0.10839176177978516, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.2150000035762787, "rewards/unicoder_reward_fn/std": 0.18493928015232086, "step": 593 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2811.0, "completions/max_terminated_length": 2811.0, "completions/mean_length": 1121.2249755859375, "completions/mean_terminated_length": 1121.2249755859375, "completions/min_length": 327.0, "completions/min_terminated_length": 327.0, "epoch": 0.230411171450737, "grad_norm": 0.4496687948703766, "kl": 0.244873046875, "learning_rate": 4.545037488961785e-06, "loss": -0.0214, "num_tokens": 54933719.0, "reward": 0.36625003814697266, "reward_std": 0.1621946543455124, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.27000001072883606, "rewards/unicoder_reward_fn/std": 0.2664344906806946, "step": 594 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2346.0, "completions/mean_length": 1002.8500366210938, "completions/mean_terminated_length": 963.6962280273438, "completions/min_length": 335.0, "completions/min_terminated_length": 335.0, "epoch": 0.2307990690457719, "grad_norm": 0.47021061182022095, "kl": 0.25390625, "learning_rate": 4.543331328988927e-06, "loss": 0.0856, "num_tokens": 55064227.0, "reward": 0.49250003695487976, "reward_std": 0.13935014605522156, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.39750000834465027, "rewards/unicoder_reward_fn/std": 0.2746574580669403, "step": 595 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1646.0, "completions/max_terminated_length": 1646.0, "completions/mean_length": 942.3875122070312, "completions/mean_terminated_length": 942.3875122070312, "completions/min_length": 366.0, "completions/min_terminated_length": 366.0, "epoch": 0.23118696664080682, "grad_norm": 0.49636074900627136, "kl": 0.26953125, "learning_rate": 4.541622337101119e-06, "loss": -0.0504, "num_tokens": 55182570.0, "reward": 0.3112500309944153, "reward_std": 0.12327909469604492, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.2150000035762787, "rewards/unicoder_reward_fn/std": 0.24396513402462006, "step": 596 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2197.0, "completions/max_terminated_length": 2197.0, "completions/mean_length": 996.2875366210938, "completions/mean_terminated_length": 996.2875366210938, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "epoch": 0.23157486423584173, "grad_norm": 0.44792500138282776, "kl": 0.246337890625, "learning_rate": 4.539910515997089e-06, "loss": 0.0126, "num_tokens": 55315009.0, "reward": 0.382500022649765, "reward_std": 0.12260345369577408, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2824999988079071, "rewards/unicoder_reward_fn/std": 0.265661358833313, "step": 597 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2415.0, "completions/max_terminated_length": 2415.0, "completions/mean_length": 940.0250244140625, "completions/mean_terminated_length": 940.0250244140625, "completions/min_length": 300.0, "completions/min_terminated_length": 300.0, "epoch": 0.23196276183087663, "grad_norm": 0.4964325726032257, "kl": 0.27099609375, "learning_rate": 4.538195868380039e-06, "loss": -0.0167, "num_tokens": 55437763.0, "reward": 0.5399999618530273, "reward_std": 0.10494407266378403, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4399999678134918, "rewards/unicoder_reward_fn/std": 0.33587220311164856, "step": 598 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2089.0, "completions/max_terminated_length": 2089.0, "completions/mean_length": 922.9249877929688, "completions/mean_terminated_length": 922.9249877929688, "completions/min_length": 233.0, "completions/min_terminated_length": 233.0, "epoch": 0.23235065942591157, "grad_norm": 0.5443563461303711, "kl": 0.28857421875, "learning_rate": 4.53647839695763e-06, "loss": 0.0219, "num_tokens": 55562661.0, "reward": 0.45749998092651367, "reward_std": 0.14723430573940277, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.35750001668930054, "rewards/unicoder_reward_fn/std": 0.30137869715690613, "step": 599 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1961.0, "completions/max_terminated_length": 1961.0, "completions/mean_length": 930.7000122070312, "completions/mean_terminated_length": 930.7000122070312, "completions/min_length": 387.0, "completions/min_terminated_length": 387.0, "epoch": 0.23273855702094648, "grad_norm": 0.4868752956390381, "kl": 0.246337890625, "learning_rate": 4.534758104441985e-06, "loss": 0.0392, "num_tokens": 55679269.0, "reward": 0.41999998688697815, "reward_std": 0.15485750138759613, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3199999928474426, "rewards/unicoder_reward_fn/std": 0.23940856754779816, "step": 600 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2354.0, "completions/max_terminated_length": 2354.0, "completions/mean_length": 994.3624877929688, "completions/mean_terminated_length": 994.3624877929688, "completions/min_length": 391.0, "completions/min_terminated_length": 391.0, "epoch": 0.2331264546159814, "grad_norm": 0.4523355960845947, "kl": 0.267333984375, "learning_rate": 4.533034993549679e-06, "loss": -0.0061, "num_tokens": 55818978.0, "reward": 0.288750022649765, "reward_std": 0.08938281238079071, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.1900000125169754, "rewards/unicoder_reward_fn/std": 0.2754053473472595, "step": 601 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2316.0, "completions/max_terminated_length": 2316.0, "completions/mean_length": 999.3250122070312, "completions/mean_terminated_length": 999.3250122070312, "completions/min_length": 424.0, "completions/min_terminated_length": 424.0, "epoch": 0.2335143522110163, "grad_norm": 0.4261605441570282, "kl": 0.24853515625, "learning_rate": 4.531309067001741e-06, "loss": 0.0039, "num_tokens": 55946732.0, "reward": 0.38750001788139343, "reward_std": 0.11525621265172958, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.27895602583885193, "step": 602 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3370.0, "completions/max_terminated_length": 3370.0, "completions/mean_length": 1092.300048828125, "completions/mean_terminated_length": 1092.300048828125, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.2339022498060512, "grad_norm": 0.49250656366348267, "kl": 0.2548828125, "learning_rate": 4.529580327523644e-06, "loss": 0.0653, "num_tokens": 56088516.0, "reward": 0.3375000059604645, "reward_std": 0.13280482590198517, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23749999701976776, "rewards/unicoder_reward_fn/std": 0.32777538895606995, "step": 603 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2235.0, "completions/max_terminated_length": 2235.0, "completions/mean_length": 1071.8375244140625, "completions/mean_terminated_length": 1071.8375244140625, "completions/min_length": 496.0, "completions/min_terminated_length": 496.0, "epoch": 0.2342901474010861, "grad_norm": 0.42302700877189636, "kl": 0.2578125, "learning_rate": 4.527848777845304e-06, "loss": -0.0059, "num_tokens": 56232415.0, "reward": 0.351250022649765, "reward_std": 0.08661413192749023, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.2524999976158142, "rewards/unicoder_reward_fn/std": 0.2732713222503662, "step": 604 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2229.0, "completions/max_terminated_length": 2229.0, "completions/mean_length": 1054.9500732421875, "completions/mean_terminated_length": 1054.9500732421875, "completions/min_length": 365.0, "completions/min_terminated_length": 365.0, "epoch": 0.23467804499612102, "grad_norm": 0.45729318261146545, "kl": 0.26025390625, "learning_rate": 4.526114420701074e-06, "loss": -0.0262, "num_tokens": 56374275.0, "reward": 0.3850000202655792, "reward_std": 0.13101422786712646, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2849999964237213, "rewards/unicoder_reward_fn/std": 0.23335744440555573, "step": 605 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2302.0, "completions/max_terminated_length": 2302.0, "completions/mean_length": 1038.25, "completions/mean_terminated_length": 1038.25, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "epoch": 0.23506594259115593, "grad_norm": 0.39974886178970337, "kl": 0.253173828125, "learning_rate": 4.524377258829739e-06, "loss": 0.0406, "num_tokens": 56495151.0, "reward": 0.4825000464916229, "reward_std": 0.12195227295160294, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3824999928474426, "rewards/unicoder_reward_fn/std": 0.3185310363769531, "step": 606 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2432.0, "completions/max_terminated_length": 2432.0, "completions/mean_length": 1053.25, "completions/mean_terminated_length": 1053.25, "completions/min_length": 401.0, "completions/min_terminated_length": 401.0, "epoch": 0.23545384018619084, "grad_norm": 0.3837812840938568, "kl": 0.26416015625, "learning_rate": 4.522637294974517e-06, "loss": -0.0036, "num_tokens": 56623531.0, "reward": 0.2874999940395355, "reward_std": 0.07112202793359756, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.22125934064388275, "step": 607 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2501.0, "completions/max_terminated_length": 2501.0, "completions/mean_length": 1075.7750244140625, "completions/mean_terminated_length": 1075.7750244140625, "completions/min_length": 319.0, "completions/min_terminated_length": 319.0, "epoch": 0.23584173778122575, "grad_norm": 0.35022711753845215, "kl": 0.254638671875, "learning_rate": 4.520894531883048e-06, "loss": -0.0185, "num_tokens": 56764257.0, "reward": 0.36625003814697266, "reward_std": 0.06246553733944893, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.26750001311302185, "rewards/unicoder_reward_fn/std": 0.2984241545200348, "step": 608 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2455.0, "completions/max_terminated_length": 2455.0, "completions/mean_length": 1040.7874755859375, "completions/mean_terminated_length": 1040.7874755859375, "completions/min_length": 389.0, "completions/min_terminated_length": 389.0, "epoch": 0.23622963537626065, "grad_norm": 0.45447787642478943, "kl": 0.259033203125, "learning_rate": 4.519148972307393e-06, "loss": 0.0065, "num_tokens": 56896936.0, "reward": 0.48000001907348633, "reward_std": 0.11904807388782501, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3799999952316284, "rewards/unicoder_reward_fn/std": 0.21488885581493378, "step": 609 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2175.0, "completions/max_terminated_length": 2175.0, "completions/mean_length": 1033.1624755859375, "completions/mean_terminated_length": 1033.1624755859375, "completions/min_length": 533.0, "completions/min_terminated_length": 533.0, "epoch": 0.2366175329712956, "grad_norm": 0.46440085768699646, "kl": 0.2421875, "learning_rate": 4.517400619004027e-06, "loss": 0.0613, "num_tokens": 57035517.0, "reward": 0.3449999988079071, "reward_std": 0.15534070134162903, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.24500000476837158, "rewards/unicoder_reward_fn/std": 0.2967759370803833, "step": 610 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2072.0, "completions/max_terminated_length": 2072.0, "completions/mean_length": 1043.4375, "completions/mean_terminated_length": 1043.4375, "completions/min_length": 291.0, "completions/min_terminated_length": 291.0, "epoch": 0.2370054305663305, "grad_norm": 0.5074459910392761, "kl": 0.260498046875, "learning_rate": 4.5156494747338415e-06, "loss": 0.0144, "num_tokens": 57165568.0, "reward": 0.4437499940395355, "reward_std": 0.1380719244480133, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.3449999988079071, "rewards/unicoder_reward_fn/std": 0.26809218525886536, "step": 611 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2162.0, "completions/max_terminated_length": 2162.0, "completions/mean_length": 882.3500366210938, "completions/mean_terminated_length": 882.3500366210938, "completions/min_length": 308.0, "completions/min_terminated_length": 308.0, "epoch": 0.2373933281613654, "grad_norm": 0.4763687252998352, "kl": 0.2666015625, "learning_rate": 4.51389554226213e-06, "loss": -0.0493, "num_tokens": 57281340.0, "reward": 0.5250000357627869, "reward_std": 0.12583580613136292, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.42500001192092896, "rewards/unicoder_reward_fn/std": 0.30585014820098877, "step": 612 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2272.0, "completions/max_terminated_length": 2272.0, "completions/mean_length": 1001.5499877929688, "completions/mean_terminated_length": 1001.5499877929688, "completions/min_length": 459.0, "completions/min_terminated_length": 459.0, "epoch": 0.23778122575640032, "grad_norm": 0.421146422624588, "kl": 0.250244140625, "learning_rate": 4.512138824358592e-06, "loss": 0.02, "num_tokens": 57408016.0, "reward": 0.4400000274181366, "reward_std": 0.15979459881782532, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3400000035762787, "rewards/unicoder_reward_fn/std": 0.2853379249572754, "step": 613 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1449.0, "completions/max_terminated_length": 1449.0, "completions/mean_length": 882.3125, "completions/mean_terminated_length": 882.3125, "completions/min_length": 345.0, "completions/min_terminated_length": 345.0, "epoch": 0.23816912335143522, "grad_norm": 0.4509447515010834, "kl": 0.27734375, "learning_rate": 4.510379323797323e-06, "loss": 0.0152, "num_tokens": 57525937.0, "reward": 0.4150000214576721, "reward_std": 0.12693791091442108, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3149999976158142, "rewards/unicoder_reward_fn/std": 0.3540363907814026, "step": 614 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4056.0, "completions/max_terminated_length": 4056.0, "completions/mean_length": 902.0250244140625, "completions/mean_terminated_length": 902.0250244140625, "completions/min_length": 371.0, "completions/min_terminated_length": 371.0, "epoch": 0.23855702094647013, "grad_norm": 0.4828765094280243, "kl": 0.250244140625, "learning_rate": 4.508617043356818e-06, "loss": 0.0007, "num_tokens": 57649875.0, "reward": 0.3475000262260437, "reward_std": 0.15080364048480988, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.25, "rewards/unicoder_reward_fn/std": 0.2454652041196823, "step": 615 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1718.0, "completions/mean_length": 923.7125244140625, "completions/mean_terminated_length": 883.5570068359375, "completions/min_length": 366.0, "completions/min_terminated_length": 366.0, "epoch": 0.23894491854150504, "grad_norm": 0.47426700592041016, "kl": 0.267578125, "learning_rate": 4.506851985819957e-06, "loss": 0.053, "num_tokens": 57776892.0, "reward": 0.26625001430511475, "reward_std": 0.08367211371660233, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.16750000417232513, "rewards/unicoder_reward_fn/std": 0.1812579184770584, "step": 616 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1819.0, "completions/max_terminated_length": 1819.0, "completions/mean_length": 853.6124877929688, "completions/mean_terminated_length": 853.6124877929688, "completions/min_length": 417.0, "completions/min_terminated_length": 417.0, "epoch": 0.23933281613653995, "grad_norm": 0.49529561400413513, "kl": 0.26171875, "learning_rate": 4.5050841539740064e-06, "loss": 0.0374, "num_tokens": 57885869.0, "reward": 0.4350000023841858, "reward_std": 0.09059371799230576, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.33500000834465027, "rewards/unicoder_reward_fn/std": 0.2153007984161377, "step": 617 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1494.0, "completions/max_terminated_length": 1494.0, "completions/mean_length": 774.7000122070312, "completions/mean_terminated_length": 774.7000122070312, "completions/min_length": 387.0, "completions/min_terminated_length": 387.0, "epoch": 0.23972071373157486, "grad_norm": 0.5469084978103638, "kl": 0.28564453125, "learning_rate": 4.503313550610613e-06, "loss": 0.0122, "num_tokens": 57989005.0, "reward": 0.4749999940395355, "reward_std": 0.1674930900335312, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.3774999976158142, "rewards/unicoder_reward_fn/std": 0.22724269330501556, "step": 618 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1748.0, "completions/max_terminated_length": 1748.0, "completions/mean_length": 959.9750366210938, "completions/mean_terminated_length": 959.9750366210938, "completions/min_length": 424.0, "completions/min_terminated_length": 424.0, "epoch": 0.24010861132660977, "grad_norm": 0.41138756275177, "kl": 0.257080078125, "learning_rate": 4.501540178525803e-06, "loss": 0.0187, "num_tokens": 58127171.0, "reward": 0.42500001192092896, "reward_std": 0.10133367031812668, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.32500001788139343, "rewards/unicoder_reward_fn/std": 0.33128058910369873, "step": 619 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2539.0, "completions/max_terminated_length": 2539.0, "completions/mean_length": 876.6124877929688, "completions/mean_terminated_length": 876.6124877929688, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.24049650892164467, "grad_norm": 0.5776572823524475, "kl": 0.259033203125, "learning_rate": 4.499764040519972e-06, "loss": 0.0246, "num_tokens": 58241676.0, "reward": 0.5, "reward_std": 0.16245850920677185, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4000000059604645, "rewards/unicoder_reward_fn/std": 0.33526867628097534, "step": 620 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2140.0, "completions/max_terminated_length": 2140.0, "completions/mean_length": 850.6875, "completions/mean_terminated_length": 850.6875, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "epoch": 0.24088440651667958, "grad_norm": 0.5105133056640625, "kl": 0.243408203125, "learning_rate": 4.497985139397885e-06, "loss": 0.031, "num_tokens": 58354435.0, "reward": 0.6487500071525574, "reward_std": 0.25176212191581726, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.550000011920929, "rewards/unicoder_reward_fn/std": 0.36905625462532043, "step": 621 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1514.0, "completions/max_terminated_length": 1514.0, "completions/mean_length": 822.7125244140625, "completions/mean_terminated_length": 822.7125244140625, "completions/min_length": 412.0, "completions/min_terminated_length": 412.0, "epoch": 0.24127230411171452, "grad_norm": 0.48941269516944885, "kl": 0.25537109375, "learning_rate": 4.496203477968669e-06, "loss": -0.0085, "num_tokens": 58465820.0, "reward": 0.43250003457069397, "reward_std": 0.16638019680976868, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.33250001072883606, "rewards/unicoder_reward_fn/std": 0.2967226207256317, "step": 622 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1646.0, "completions/max_terminated_length": 1646.0, "completions/mean_length": 806.2999877929688, "completions/mean_terminated_length": 806.2999877929688, "completions/min_length": 350.0, "completions/min_terminated_length": 350.0, "epoch": 0.24166020170674943, "grad_norm": 0.5780818462371826, "kl": 0.262451171875, "learning_rate": 4.494419059045811e-06, "loss": -0.0005, "num_tokens": 58574172.0, "reward": 0.42500001192092896, "reward_std": 0.16162356734275818, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.32499998807907104, "rewards/unicoder_reward_fn/std": 0.2870628237724304, "step": 623 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2617.0, "completions/mean_length": 878.9000244140625, "completions/mean_terminated_length": 838.17724609375, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.24204809930178434, "grad_norm": 0.5586786866188049, "kl": 0.248779296875, "learning_rate": 4.4926318854471516e-06, "loss": 0.0447, "num_tokens": 58690892.0, "reward": 0.42875000834465027, "reward_std": 0.20961745083332062, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.33250001072883606, "rewards/unicoder_reward_fn/std": 0.26613739132881165, "step": 624 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1949.0, "completions/max_terminated_length": 1949.0, "completions/mean_length": 841.1375122070312, "completions/mean_terminated_length": 841.1375122070312, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "epoch": 0.24243599689681924, "grad_norm": 0.555622935295105, "kl": 0.23486328125, "learning_rate": 4.490841959994883e-06, "loss": 0.0351, "num_tokens": 58809583.0, "reward": 0.38625001907348633, "reward_std": 0.21688099205493927, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.28076523542404175, "step": 625 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1518.0, "completions/max_terminated_length": 1518.0, "completions/mean_length": 785.3624877929688, "completions/mean_terminated_length": 785.3624877929688, "completions/min_length": 344.0, "completions/min_terminated_length": 344.0, "epoch": 0.24282389449185415, "grad_norm": 0.6320343613624573, "kl": 0.25146484375, "learning_rate": 4.4890492855155395e-06, "loss": 0.0335, "num_tokens": 58927316.0, "reward": 0.5250000357627869, "reward_std": 0.15847240388393402, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.4274999499320984, "rewards/unicoder_reward_fn/std": 0.33031436800956726, "step": 626 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1602.0, "completions/max_terminated_length": 1602.0, "completions/mean_length": 745.1500244140625, "completions/mean_terminated_length": 745.1500244140625, "completions/min_length": 390.0, "completions/min_terminated_length": 390.0, "epoch": 0.24321179208688906, "grad_norm": 0.578612744808197, "kl": 0.2822265625, "learning_rate": 4.487253864840001e-06, "loss": 0.0057, "num_tokens": 59039752.0, "reward": 0.32750001549720764, "reward_std": 0.10644640773534775, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.22750000655651093, "rewards/unicoder_reward_fn/std": 0.19025298953056335, "step": 627 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1320.0, "completions/max_terminated_length": 1320.0, "completions/mean_length": 724.0250244140625, "completions/mean_terminated_length": 724.0250244140625, "completions/min_length": 302.0, "completions/min_terminated_length": 302.0, "epoch": 0.24359968968192397, "grad_norm": 0.5264387726783752, "kl": 0.2646484375, "learning_rate": 4.485455700803481e-06, "loss": 0.0191, "num_tokens": 59150378.0, "reward": 0.4625000059604645, "reward_std": 0.14085224270820618, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.36500000953674316, "rewards/unicoder_reward_fn/std": 0.2658161520957947, "step": 628 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1389.0, "completions/max_terminated_length": 1389.0, "completions/mean_length": 702.4375, "completions/mean_terminated_length": 702.4375, "completions/min_length": 300.0, "completions/min_terminated_length": 300.0, "epoch": 0.24398758727695888, "grad_norm": 0.6117992997169495, "kl": 0.259521484375, "learning_rate": 4.483654796245526e-06, "loss": -0.0548, "num_tokens": 59256725.0, "reward": 0.4337500035762787, "reward_std": 0.12570030987262726, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.33500000834465027, "rewards/unicoder_reward_fn/std": 0.25413045287132263, "step": 629 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1741.0, "completions/max_terminated_length": 1741.0, "completions/mean_length": 758.4750366210938, "completions/mean_terminated_length": 758.4750366210938, "completions/min_length": 371.0, "completions/min_terminated_length": 371.0, "epoch": 0.2443754848719938, "grad_norm": 0.5407521724700928, "kl": 0.27978515625, "learning_rate": 4.481851154010011e-06, "loss": 0.0316, "num_tokens": 59370147.0, "reward": 0.39625000953674316, "reward_std": 0.13571690022945404, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.29749998450279236, "rewards/unicoder_reward_fn/std": 0.3245152533054352, "step": 630 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1306.0, "completions/max_terminated_length": 1306.0, "completions/mean_length": 762.5499877929688, "completions/mean_terminated_length": 762.5499877929688, "completions/min_length": 319.0, "completions/min_terminated_length": 319.0, "epoch": 0.2447633824670287, "grad_norm": 0.5865707397460938, "kl": 0.27392578125, "learning_rate": 4.480044776945131e-06, "loss": 0.0373, "num_tokens": 59473335.0, "reward": 0.5175000429153442, "reward_std": 0.1814797818660736, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.41750001907348633, "rewards/unicoder_reward_fn/std": 0.3137260675430298, "step": 631 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1654.0, "completions/max_terminated_length": 1654.0, "completions/mean_length": 744.7125244140625, "completions/mean_terminated_length": 744.7125244140625, "completions/min_length": 220.0, "completions/min_terminated_length": 220.0, "epoch": 0.2451512800620636, "grad_norm": 0.5741066336631775, "kl": 0.2841796875, "learning_rate": 4.478235667903404e-06, "loss": -0.0, "num_tokens": 59586136.0, "reward": 0.45500001311302185, "reward_std": 0.20355306565761566, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.35500001907348633, "rewards/unicoder_reward_fn/std": 0.2863564193248749, "step": 632 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1553.0, "completions/max_terminated_length": 1553.0, "completions/mean_length": 741.7875366210938, "completions/mean_terminated_length": 741.7875366210938, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.24553917765709854, "grad_norm": 0.5153830051422119, "kl": 0.246826171875, "learning_rate": 4.4764238297416595e-06, "loss": -0.0296, "num_tokens": 59687847.0, "reward": 0.4375, "reward_std": 0.13369779288768768, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3375000059604645, "rewards/unicoder_reward_fn/std": 0.3436126112937927, "step": 633 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2178.0, "completions/max_terminated_length": 2178.0, "completions/mean_length": 767.625, "completions/mean_terminated_length": 767.625, "completions/min_length": 389.0, "completions/min_terminated_length": 389.0, "epoch": 0.24592707525213345, "grad_norm": 0.5920447111129761, "kl": 0.2734375, "learning_rate": 4.474609265321038e-06, "loss": -0.0429, "num_tokens": 59790425.0, "reward": 0.33250004053115845, "reward_std": 0.1524578183889389, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.23499998450279236, "rewards/unicoder_reward_fn/std": 0.2199539691209793, "step": 634 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1722.0, "completions/max_terminated_length": 1722.0, "completions/mean_length": 792.4625244140625, "completions/mean_terminated_length": 792.4625244140625, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "epoch": 0.24631497284716836, "grad_norm": 0.5588251352310181, "kl": 0.2666015625, "learning_rate": 4.472791977506983e-06, "loss": 0.0141, "num_tokens": 59903894.0, "reward": 0.4100000560283661, "reward_std": 0.16177789866924286, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3100000023841858, "rewards/unicoder_reward_fn/std": 0.30671802163124084, "step": 635 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1569.0, "completions/max_terminated_length": 1569.0, "completions/mean_length": 825.0750122070312, "completions/mean_terminated_length": 825.0750122070312, "completions/min_length": 340.0, "completions/min_terminated_length": 340.0, "epoch": 0.24670287044220326, "grad_norm": 0.4729157090187073, "kl": 0.250244140625, "learning_rate": 4.47097196916924e-06, "loss": -0.0189, "num_tokens": 60014380.0, "reward": 0.48125001788139343, "reward_std": 0.14409561455249786, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.382500022649765, "rewards/unicoder_reward_fn/std": 0.31533586978912354, "step": 636 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1538.0, "completions/max_terminated_length": 1538.0, "completions/mean_length": 861.3375244140625, "completions/mean_terminated_length": 861.3375244140625, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.24709076803723817, "grad_norm": 0.44663330912590027, "kl": 0.27001953125, "learning_rate": 4.469149243181851e-06, "loss": 0.0424, "num_tokens": 60131143.0, "reward": 0.32750001549720764, "reward_std": 0.1142178401350975, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.22750000655651093, "rewards/unicoder_reward_fn/std": 0.2695401608943939, "step": 637 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1883.0, "completions/max_terminated_length": 1883.0, "completions/mean_length": 825.9500122070312, "completions/mean_terminated_length": 825.9500122070312, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "epoch": 0.24747866563227308, "grad_norm": 0.48949044942855835, "kl": 0.271484375, "learning_rate": 4.4673238024231515e-06, "loss": 0.0109, "num_tokens": 60247603.0, "reward": 0.4100000560283661, "reward_std": 0.15079300105571747, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3100000023841858, "rewards/unicoder_reward_fn/std": 0.2879873514175415, "step": 638 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1298.0, "completions/max_terminated_length": 1298.0, "completions/mean_length": 807.3250122070312, "completions/mean_terminated_length": 807.3250122070312, "completions/min_length": 364.0, "completions/min_terminated_length": 364.0, "epoch": 0.247866563227308, "grad_norm": 0.4766918122768402, "kl": 0.2939453125, "learning_rate": 4.465495649775758e-06, "loss": 0.0007, "num_tokens": 60364181.0, "reward": 0.38750001788139343, "reward_std": 0.10698860883712769, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.28256288170814514, "step": 639 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2866.0, "completions/max_terminated_length": 2866.0, "completions/mean_length": 938.7625122070312, "completions/mean_terminated_length": 938.7625122070312, "completions/min_length": 384.0, "completions/min_terminated_length": 384.0, "epoch": 0.2482544608223429, "grad_norm": 0.43838635087013245, "kl": 0.25, "learning_rate": 4.4636647881265746e-06, "loss": 0.0126, "num_tokens": 60491298.0, "reward": 0.38875001668930054, "reward_std": 0.14677675068378448, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.28999999165534973, "rewards/unicoder_reward_fn/std": 0.2602822184562683, "step": 640 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2179.0, "completions/max_terminated_length": 2179.0, "completions/mean_length": 825.0750122070312, "completions/mean_terminated_length": 825.0750122070312, "completions/min_length": 300.0, "completions/min_terminated_length": 300.0, "epoch": 0.2486423584173778, "grad_norm": 0.5052236318588257, "kl": 0.30224609375, "learning_rate": 4.461831220366783e-06, "loss": 0.0032, "num_tokens": 60607920.0, "reward": 0.44749999046325684, "reward_std": 0.1550348401069641, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3475000262260437, "rewards/unicoder_reward_fn/std": 0.3162177801132202, "step": 641 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2205.0, "completions/mean_length": 913.9249877929688, "completions/mean_terminated_length": 873.6455688476562, "completions/min_length": 371.0, "completions/min_terminated_length": 371.0, "epoch": 0.24903025601241272, "grad_norm": 0.38875219225883484, "kl": 0.28369140625, "learning_rate": 4.459994949391834e-06, "loss": 0.0313, "num_tokens": 60735810.0, "reward": 0.32375001907348633, "reward_std": 0.05031966045498848, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.22499999403953552, "rewards/unicoder_reward_fn/std": 0.21903124451637268, "step": 642 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2345.0, "completions/mean_length": 970.5, "completions/mean_terminated_length": 890.3590087890625, "completions/min_length": 476.0, "completions/min_terminated_length": 476.0, "epoch": 0.24941815360744762, "grad_norm": 0.5126664638519287, "kl": 0.2861328125, "learning_rate": 4.458155978101453e-06, "loss": 0.0606, "num_tokens": 60851162.0, "reward": 0.46250006556510925, "reward_std": 0.1905437558889389, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.36250001192092896, "rewards/unicoder_reward_fn/std": 0.27206432819366455, "step": 643 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2817.0, "completions/max_terminated_length": 2817.0, "completions/mean_length": 964.6749877929688, "completions/mean_terminated_length": 964.6749877929688, "completions/min_length": 421.0, "completions/min_terminated_length": 421.0, "epoch": 0.24980605120248253, "grad_norm": 0.388233482837677, "kl": 0.246337890625, "learning_rate": 4.456314309399624e-06, "loss": 0.0005, "num_tokens": 60966888.0, "reward": 0.3799999952316284, "reward_std": 0.11005055904388428, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2800000309944153, "rewards/unicoder_reward_fn/std": 0.2748532295227051, "step": 644 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3146.0, "completions/mean_length": 901.3250122070312, "completions/mean_terminated_length": 860.8861083984375, "completions/min_length": 361.0, "completions/min_terminated_length": 361.0, "epoch": 0.25019394879751744, "grad_norm": 0.4977852404117584, "kl": 0.3017578125, "learning_rate": 4.454469946194597e-06, "loss": 0.043, "num_tokens": 61078666.0, "reward": 0.3387500047683716, "reward_std": 0.1268996000289917, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.24000000953674316, "rewards/unicoder_reward_fn/std": 0.19461093842983246, "step": 645 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2435.0, "completions/max_terminated_length": 2435.0, "completions/mean_length": 794.5125122070312, "completions/mean_terminated_length": 794.5125122070312, "completions/min_length": 252.0, "completions/min_terminated_length": 252.0, "epoch": 0.2505818463925524, "grad_norm": 0.5561448931694031, "kl": 0.29296875, "learning_rate": 4.452622891398869e-06, "loss": -0.0108, "num_tokens": 61193787.0, "reward": 0.5012499690055847, "reward_std": 0.14116887748241425, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.4025000035762787, "rewards/unicoder_reward_fn/std": 0.2122662514448166, "step": 646 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1810.0, "completions/mean_length": 881.4249877929688, "completions/mean_terminated_length": 840.7341918945312, "completions/min_length": 347.0, "completions/min_terminated_length": 347.0, "epoch": 0.25096974398758726, "grad_norm": 0.5268040299415588, "kl": 0.2978515625, "learning_rate": 4.450773147929195e-06, "loss": 0.041, "num_tokens": 61316069.0, "reward": 0.5425000190734863, "reward_std": 0.13157624006271362, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.4425000250339508, "rewards/unicoder_reward_fn/std": 0.285858690738678, "step": 647 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2493.0, "completions/mean_length": 1122.7625732421875, "completions/mean_terminated_length": 837.6575317382812, "completions/min_length": 395.0, "completions/min_terminated_length": 395.0, "epoch": 0.2513576415826222, "grad_norm": 0.746401846408844, "kl": 0.3017578125, "learning_rate": 4.448920718706569e-06, "loss": 0.1056, "num_tokens": 61454282.0, "reward": 0.49000006914138794, "reward_std": 0.24962210655212402, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.39500001072883606, "rewards/unicoder_reward_fn/std": 0.32913389801979065, "step": 648 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1775.0, "completions/mean_length": 1130.612548828125, "completions/mean_terminated_length": 846.26025390625, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "epoch": 0.2517455391776571, "grad_norm": 1.3163787126541138, "kl": 0.30419921875, "learning_rate": 4.447065606656234e-06, "loss": 0.1916, "num_tokens": 61590811.0, "reward": 0.33000001311302185, "reward_std": 0.16868242621421814, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.23499998450279236, "rewards/unicoder_reward_fn/std": 0.22451071441173553, "step": 649 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1273.0, "completions/mean_length": 2863.887451171875, "completions/mean_terminated_length": 697.0689697265625, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "epoch": 0.252133436772692, "grad_norm": 2.7188496589660645, "kl": 0.359375, "learning_rate": 4.445207814707661e-06, "loss": 0.2202, "num_tokens": 61877074.0, "reward": 0.28125, "reward_std": 0.2453388273715973, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.3158172369003296, "step": 650 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 3014.0, "completions/mean_length": 3101.025146484375, "completions/mean_terminated_length": 779.4166870117188, "completions/min_length": 400.0, "completions/min_terminated_length": 400.0, "epoch": 0.25252133436772695, "grad_norm": 2.3346261978149414, "kl": 0.6484375, "learning_rate": 4.4433473457945605e-06, "loss": 0.0989, "num_tokens": 62177084.0, "reward": 0.32750001549720764, "reward_std": 0.25367411971092224, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.2474999874830246, "rewards/unicoder_reward_fn/std": 0.29034027457237244, "step": 651 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 960.0, "completions/mean_length": 3697.012451171875, "completions/mean_terminated_length": 549.4444580078125, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.2529092319627618, "grad_norm": 1.4652224779129028, "kl": 1.482421875, "learning_rate": 4.4414842028548644e-06, "loss": 0.0613, "num_tokens": 62515445.0, "reward": 0.36000001430511475, "reward_std": 0.18837594985961914, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.29749998450279236, "rewards/unicoder_reward_fn/std": 0.30687278509140015, "step": 652 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 761.0, "completions/mean_length": 3919.762451171875, "completions/mean_terminated_length": 571.25, "completions/min_length": 460.0, "completions/min_terminated_length": 460.0, "epoch": 0.25329712955779676, "grad_norm": 1.1846668720245361, "kl": 2.00390625, "learning_rate": 4.439618388830732e-06, "loss": 0.027, "num_tokens": 62878082.0, "reward": 0.36875003576278687, "reward_std": 0.20127306878566742, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.30000001192092896, "rewards/unicoder_reward_fn/std": 0.2899148762226105, "step": 653 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 746.0, "completions/mean_length": 3877.550048828125, "completions/mean_terminated_length": 600.7999877929688, "completions/min_length": 448.0, "completions/min_terminated_length": 448.0, "epoch": 0.25368502715283164, "grad_norm": 1.6897838115692139, "kl": 1.89453125, "learning_rate": 4.437749906668535e-06, "loss": 0.0077, "num_tokens": 63239846.0, "reward": 0.32375001907348633, "reward_std": 0.23918314278125763, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.2550000250339508, "rewards/unicoder_reward_fn/std": 0.31818312406539917, "step": 654 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 714.0, "completions/mean_length": 3695.97509765625, "completions/mean_terminated_length": 540.2222290039062, "completions/min_length": 448.0, "completions/min_terminated_length": 448.0, "epoch": 0.2540729247478666, "grad_norm": 1.665998935699463, "kl": 1.16796875, "learning_rate": 4.435878759318864e-06, "loss": 0.064, "num_tokens": 63581788.0, "reward": 0.30000001192092896, "reward_std": 0.1552770882844925, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.22750000655651093, "rewards/unicoder_reward_fn/std": 0.2500506341457367, "step": 655 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2061.0, "completions/mean_length": 3251.5, "completions/mean_terminated_length": 540.2105102539062, "completions/min_length": 261.0, "completions/min_terminated_length": 261.0, "epoch": 0.25446082234290146, "grad_norm": 3.859569549560547, "kl": 1.134765625, "learning_rate": 4.434004949736515e-06, "loss": 0.1549, "num_tokens": 63899260.0, "reward": 0.2912500202655792, "reward_std": 0.24413461983203888, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.22499999403953552, "rewards/unicoder_reward_fn/std": 0.30419015884399414, "step": 656 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8125, "completions/max_length": 4096.0, "completions/max_terminated_length": 613.0, "completions/mean_length": 3415.175048828125, "completions/mean_terminated_length": 464.933349609375, "completions/min_length": 323.0, "completions/min_terminated_length": 323.0, "epoch": 0.2548487199379364, "grad_norm": 4.217111110687256, "kl": 1.685546875, "learning_rate": 4.432128480880491e-06, "loss": 0.1828, "num_tokens": 64218970.0, "reward": 0.32375001907348633, "reward_std": 0.3008497655391693, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.26249998807907104, "rewards/unicoder_reward_fn/std": 0.3175758719444275, "step": 657 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 589.0, "completions/mean_length": 3824.337646484375, "completions/mean_terminated_length": 473.8333435058594, "completions/min_length": 305.0, "completions/min_terminated_length": 305.0, "epoch": 0.2552366175329713, "grad_norm": 3.944188117980957, "kl": 3.265625, "learning_rate": 4.430249355713991e-06, "loss": 0.0907, "num_tokens": 64575373.0, "reward": 0.1274999976158142, "reward_std": 0.18874917924404144, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.2564484775066376, "step": 658 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 640.0, "completions/mean_length": 3688.462646484375, "completions/mean_terminated_length": 473.4444580078125, "completions/min_length": 325.0, "completions/min_terminated_length": 325.0, "epoch": 0.2556245151280062, "grad_norm": 5.133075714111328, "kl": 4.8046875, "learning_rate": 4.4283675772044105e-06, "loss": 0.1492, "num_tokens": 64917490.0, "reward": 0.22874999046325684, "reward_std": 0.2384788990020752, "rewards/format_reward/mean": 0.2874999940395355, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.3724704086780548, "step": 659 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 1521.0, "completions/mean_length": 2935.824951171875, "completions/mean_terminated_length": 526.2307739257812, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "epoch": 0.2560124127230411, "grad_norm": 11.054366111755371, "kl": 5.9453125, "learning_rate": 4.426483148323334e-06, "loss": 0.4016, "num_tokens": 65193348.0, "reward": 0.1537499874830246, "reward_std": 0.1906917840242386, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.20829325914382935, "step": 660 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 738.0, "completions/mean_length": 3518.16259765625, "completions/mean_terminated_length": 540.0769653320312, "completions/min_length": 351.0, "completions/min_terminated_length": 351.0, "epoch": 0.25640031031807603, "grad_norm": 3.359651803970337, "kl": 9.515625, "learning_rate": 4.424596072046533e-06, "loss": 0.3182, "num_tokens": 65528049.0, "reward": 0.07000000774860382, "reward_std": 0.13239377737045288, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.15094637870788574, "step": 661 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 730.0, "completions/mean_length": 3737.262451171875, "completions/mean_terminated_length": 508.625, "completions/min_length": 334.0, "completions/min_terminated_length": 334.0, "epoch": 0.25678820791311097, "grad_norm": 8.893245697021484, "kl": 11.921875, "learning_rate": 4.422706351353957e-06, "loss": 0.2931, "num_tokens": 65884526.0, "reward": 0.04500000178813934, "reward_std": 0.10523207485675812, "rewards/format_reward/mean": 0.17499999701976776, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.11797340959310532, "step": 662 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 415.0, "completions/mean_length": 4003.550048828125, "completions/mean_terminated_length": 398.0, "completions/min_length": 381.0, "completions/min_terminated_length": 381.0, "epoch": 0.25717610550814585, "grad_norm": 12.084975242614746, "kl": 27.09375, "learning_rate": 4.4208139892297356e-06, "loss": 0.3249, "num_tokens": 66255426.0, "reward": 0.01875000260770321, "reward_std": 0.04082316905260086, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434914350509644, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.06284399330615997, "step": 663 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3310.0, "completions/mean_length": 3999.612548828125, "completions/mean_terminated_length": 1525.666748046875, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "epoch": 0.2575640031031808, "grad_norm": 11.456422805786133, "kl": 27.28125, "learning_rate": 4.418918988662164e-06, "loss": 0.2587, "num_tokens": 66620619.0, "reward": 0.07874999940395355, "reward_std": 0.11318010091781616, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434914350509644, "rewards/unicoder_reward_fn/mean": 0.06999999284744263, "rewards/unicoder_reward_fn/std": 0.24463872611522675, "step": 664 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1519.0, "completions/mean_length": 3975.012451171875, "completions/mean_terminated_length": 869.6666870117188, "completions/min_length": 518.0, "completions/min_terminated_length": 518.0, "epoch": 0.25795190069821566, "grad_norm": 5.625100612640381, "kl": 8.609375, "learning_rate": 4.4170213526437075e-06, "loss": 0.0757, "num_tokens": 66988068.0, "reward": 0.1600000113248825, "reward_std": 0.1905207484960556, "rewards/format_reward/mean": 0.22499999403953552, "rewards/format_reward/std": 0.4202169179916382, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.25575652718544006, "step": 665 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.85, "completions/max_length": 4096.0, "completions/max_terminated_length": 3520.0, "completions/mean_length": 3656.775146484375, "completions/mean_terminated_length": 1167.8333740234375, "completions/min_length": 419.0, "completions/min_terminated_length": 419.0, "epoch": 0.2583397982932506, "grad_norm": 11.633063316345215, "kl": 1.541015625, "learning_rate": 4.415121084170993e-06, "loss": -0.0306, "num_tokens": 67330570.0, "reward": 0.26500001549720764, "reward_std": 0.16694989800453186, "rewards/format_reward/mean": 0.22499999403953552, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.24250002205371857, "rewards/unicoder_reward_fn/std": 0.3271569013595581, "step": 666 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9125, "completions/max_length": 4096.0, "completions/max_terminated_length": 969.0, "completions/mean_length": 3793.72509765625, "completions/mean_terminated_length": 641.4285888671875, "completions/min_length": 432.0, "completions/min_terminated_length": 432.0, "epoch": 0.2587276958882855, "grad_norm": 0.9813224077224731, "kl": 0.40087890625, "learning_rate": 4.4132181862448044e-06, "loss": -0.0117, "num_tokens": 67683836.0, "reward": 0.24375000596046448, "reward_std": 0.18309643864631653, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.23250000178813934, "rewards/unicoder_reward_fn/std": 0.2924254834651947, "step": 667 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 550.0, "completions/mean_length": 4005.512451171875, "completions/mean_terminated_length": 476.5, "completions/min_length": 403.0, "completions/min_terminated_length": 403.0, "epoch": 0.2591155934833204, "grad_norm": 18868.78125, "kl": 4.360107421875, "learning_rate": 4.411312661870076e-06, "loss": 0.0378, "num_tokens": 68052157.0, "reward": 0.30250000953674316, "reward_std": 0.19683842360973358, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.29500001668930054, "rewards/unicoder_reward_fn/std": 0.3275919258594513, "step": 668 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 709.0, "completions/mean_length": 3867.362548828125, "completions/mean_terminated_length": 437.8000183105469, "completions/min_length": 284.0, "completions/min_terminated_length": 284.0, "epoch": 0.2595034910783553, "grad_norm": 102.83506774902344, "kl": 1.06787109375, "learning_rate": 4.409404514055892e-06, "loss": 0.0413, "num_tokens": 68399154.0, "reward": 0.19875000417232513, "reward_std": 0.13043589890003204, "rewards/format_reward/mean": 0.13750000298023224, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.1849999874830246, "rewards/unicoder_reward_fn/std": 0.3611163794994354, "step": 669 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 855.0, "completions/mean_length": 3964.300048828125, "completions/mean_terminated_length": 584.0, "completions/min_length": 446.0, "completions/min_terminated_length": 446.0, "epoch": 0.25989138867339023, "grad_norm": 9.922833442687988, "kl": 1.62890625, "learning_rate": 4.407493745815476e-06, "loss": 0.0116, "num_tokens": 68764770.0, "reward": 0.07750000059604645, "reward_std": 0.127890944480896, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.16868816316127777, "step": 670 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 443.0, "completions/mean_length": 4003.525146484375, "completions/mean_terminated_length": 397.0, "completions/min_length": 351.0, "completions/min_terminated_length": 351.0, "epoch": 0.2602792862684251, "grad_norm": 8.500365257263184, "kl": 12.875, "learning_rate": 4.405580360166195e-06, "loss": 0.1646, "num_tokens": 69138428.0, "reward": 0.02250000089406967, "reward_std": 0.05184227228164673, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.07919692248106003, "step": 671 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26066718386346005, "grad_norm": 13.289616584777832, "kl": 12.578125, "learning_rate": 4.4036643601295446e-06, "loss": 0.1259, "num_tokens": 69514572.0, "reward": 0.017500001937150955, "reward_std": 0.044613536447286606, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.09648899734020233, "step": 672 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 306.0, "completions/mean_length": 4000.150146484375, "completions/mean_terminated_length": 262.0, "completions/min_length": 218.0, "completions/min_terminated_length": 218.0, "epoch": 0.26105508145849493, "grad_norm": 6.774023056030273, "kl": 4.34375, "learning_rate": 4.40174574873115e-06, "loss": 0.0847, "num_tokens": 69883888.0, "reward": 0.010000000707805157, "reward_std": 0.025842303410172462, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.044721364974975586, "step": 673 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26144297905352987, "grad_norm": 3.1190783977508545, "kl": 3.875, "learning_rate": 4.399824529000761e-06, "loss": 0.0388, "num_tokens": 70258632.0, "reward": 0.011250000447034836, "reward_std": 0.03181980550289154, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.07044164091348648, "step": 674 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2618308766485648, "grad_norm": 11.234739303588867, "kl": 8.28125, "learning_rate": 4.397900703972245e-06, "loss": 0.0831, "num_tokens": 70632648.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 675 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2622187742435997, "grad_norm": 3.6912038326263428, "kl": 5.5390625, "learning_rate": 4.3959742766835854e-06, "loss": 0.0554, "num_tokens": 71010200.0, "reward": 0.042500000447034836, "reward_std": 0.07318723946809769, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.04249999672174454, "rewards/unicoder_reward_fn/std": 0.15159490704536438, "step": 676 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2626066718386346, "grad_norm": 11.123494148254395, "kl": 8.5, "learning_rate": 4.394045250176873e-06, "loss": 0.085, "num_tokens": 71384112.0, "reward": 0.03750000149011612, "reward_std": 0.051754917949438095, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.1911821961402893, "step": 677 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2629945694336695, "grad_norm": 4.204624652862549, "kl": 5.1015625, "learning_rate": 4.392113627498305e-06, "loss": 0.0511, "num_tokens": 71754360.0, "reward": 0.015000000596046448, "reward_std": 0.020701969042420387, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.07647287845611572, "step": 678 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26338246702870444, "grad_norm": 3.5734245777130127, "kl": 1.74609375, "learning_rate": 4.390179411698176e-06, "loss": 0.0174, "num_tokens": 72135912.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 679 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2637703646237393, "grad_norm": 0.30264148116111755, "kl": 0.181884765625, "learning_rate": 4.388242605830878e-06, "loss": 0.0018, "num_tokens": 72517968.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 680 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26415826221877425, "grad_norm": 0.014419930055737495, "kl": 0.04559326171875, "learning_rate": 4.386303212954891e-06, "loss": 0.0005, "num_tokens": 72888312.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 681 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26454615981380913, "grad_norm": 0.11353500187397003, "kl": 0.03411865234375, "learning_rate": 4.384361236132781e-06, "loss": 0.0003, "num_tokens": 73262712.0, "reward": 0.010000000707805157, "reward_std": 0.018516402691602707, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.06284399330615997, "step": 682 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 278.0, "completions/mean_length": 4048.275146484375, "completions/mean_terminated_length": 278.0, "completions/min_length": 278.0, "completions/min_terminated_length": 278.0, "epoch": 0.26493405740884407, "grad_norm": 0.1397808939218521, "kl": 0.03472900390625, "learning_rate": 4.382416678431195e-06, "loss": 0.0312, "num_tokens": 73629774.0, "reward": 0.028750000521540642, "reward_std": 0.05968106910586357, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.14838796854019165, "step": 683 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26532195500387895, "grad_norm": 0.2040976583957672, "kl": 0.033935546875, "learning_rate": 4.380469542920856e-06, "loss": 0.0003, "num_tokens": 74007886.0, "reward": 0.02500000037252903, "reward_std": 0.04629100486636162, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.15710999071598053, "step": 684 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2657098525989139, "grad_norm": 0.13536415994167328, "kl": 0.031768798828125, "learning_rate": 4.378519832676554e-06, "loss": 0.0003, "num_tokens": 74385566.0, "reward": 0.007500000298023224, "reward_std": 0.02121320366859436, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 685 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2660977501939488, "grad_norm": 0.18098877370357513, "kl": 0.027557373046875, "learning_rate": 4.376567550777149e-06, "loss": 0.0003, "num_tokens": 74758702.0, "reward": 0.02500000037252903, "reward_std": 0.04629100486636162, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.15710999071598053, "step": 686 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2664856477889837, "grad_norm": 0.007387795951217413, "kl": 0.027313232421875, "learning_rate": 4.374612700305562e-06, "loss": 0.0003, "num_tokens": 75147142.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 687 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26687354538401864, "grad_norm": 0.09031977504491806, "kl": 0.037567138671875, "learning_rate": 4.372655284348767e-06, "loss": 0.0004, "num_tokens": 75533030.0, "reward": 0.02500000037252903, "reward_std": 0.020701969042420387, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.09743547439575195, "step": 688 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2672614429790535, "grad_norm": 0.10768009722232819, "kl": 0.031280517578125, "learning_rate": 4.370695305997791e-06, "loss": 0.0003, "num_tokens": 75899182.0, "reward": 0.010000000707805157, "reward_std": 0.02828427031636238, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944272249937057, "step": 689 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26764934057408846, "grad_norm": 0.005661409348249435, "kl": 0.023101806640625, "learning_rate": 4.368732768347708e-06, "loss": 0.0002, "num_tokens": 76284702.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 690 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26803723816912334, "grad_norm": 0.19576935470104218, "kl": 0.05419921875, "learning_rate": 4.366767674497633e-06, "loss": 0.0005, "num_tokens": 76668958.0, "reward": 0.10499999672174454, "reward_std": 0.03609367832541466, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.2916127145290375, "step": 691 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2684251357641583, "grad_norm": 0.005398981738835573, "kl": 0.022430419921875, "learning_rate": 4.364800027550717e-06, "loss": 0.0002, "num_tokens": 77037950.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 692 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26881303335919315, "grad_norm": 0.09330437332391739, "kl": 0.03936767578125, "learning_rate": 4.362829830614143e-06, "loss": 0.0004, "num_tokens": 77410502.0, "reward": 0.012500000186264515, "reward_std": 0.0353553406894207, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 693 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2692009309542281, "grad_norm": 0.11179521679878235, "kl": 0.0330810546875, "learning_rate": 4.36085708679912e-06, "loss": 0.0003, "num_tokens": 77774902.0, "reward": 0.06000000238418579, "reward_std": 0.08675967901945114, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.21908903121948242, "step": 694 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.26958882854926297, "grad_norm": 0.006108632776886225, "kl": 0.023681640625, "learning_rate": 4.358881799220882e-06, "loss": 0.0002, "num_tokens": 78153854.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 695 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2699767261442979, "grad_norm": 2.595987558364868, "kl": 0.060333251953125, "learning_rate": 4.356903970998677e-06, "loss": 0.0006, "num_tokens": 78530614.0, "reward": 0.005000000353902578, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472136124968529, "step": 696 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27036462373933284, "grad_norm": 0.005465980153530836, "kl": 0.0225830078125, "learning_rate": 4.354923605255764e-06, "loss": 0.0002, "num_tokens": 78900238.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 697 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2707525213343677, "grad_norm": 0.005992440972477198, "kl": 0.02386474609375, "learning_rate": 4.352940705119413e-06, "loss": 0.0002, "num_tokens": 79272550.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 698 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27114041892940266, "grad_norm": 0.00610182574018836, "kl": 0.02337646484375, "learning_rate": 4.350955273720893e-06, "loss": 0.0002, "num_tokens": 79641558.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 699 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27152831652443754, "grad_norm": 0.00662188557907939, "kl": 0.02313232421875, "learning_rate": 4.348967314195471e-06, "loss": 0.0002, "num_tokens": 80010614.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 700 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2719162141194725, "grad_norm": 0.0046420618891716, "kl": 0.02203369140625, "learning_rate": 4.346976829682406e-06, "loss": 0.0002, "num_tokens": 80387646.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 701 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27230411171450736, "grad_norm": 0.006395313888788223, "kl": 0.023681640625, "learning_rate": 4.344983823324945e-06, "loss": 0.0002, "num_tokens": 80769662.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 702 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2726920093095423, "grad_norm": 0.004501127637922764, "kl": 0.021270751953125, "learning_rate": 4.342988298270319e-06, "loss": 0.0002, "num_tokens": 81151766.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 703 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2730799069045772, "grad_norm": 0.005169012118130922, "kl": 0.02227783203125, "learning_rate": 4.340990257669732e-06, "loss": 0.0002, "num_tokens": 81535334.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 704 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2734678044996121, "grad_norm": 0.036882154643535614, "kl": 0.02191162109375, "learning_rate": 4.338989704678364e-06, "loss": 0.0002, "num_tokens": 81898974.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360680624842644, "step": 705 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.273855702094647, "grad_norm": 0.0047412752173841, "kl": 0.0225830078125, "learning_rate": 4.3369866424553604e-06, "loss": 0.0002, "num_tokens": 82269582.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 706 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2742435996896819, "grad_norm": 0.0052949972450733185, "kl": 0.021728515625, "learning_rate": 4.334981074163832e-06, "loss": 0.0002, "num_tokens": 82645982.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 707 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2746314972847168, "grad_norm": 0.00499942060559988, "kl": 0.022186279296875, "learning_rate": 4.332973002970842e-06, "loss": 0.0002, "num_tokens": 83012622.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 708 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27501939487975174, "grad_norm": 0.005200771614909172, "kl": 0.024688720703125, "learning_rate": 4.330962432047412e-06, "loss": 0.0002, "num_tokens": 83379758.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 709 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2754072924747867, "grad_norm": 0.0040735118091106415, "kl": 0.02264404296875, "learning_rate": 4.328949364568504e-06, "loss": 0.0002, "num_tokens": 83750310.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 710 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27579519006982156, "grad_norm": 0.005266095511615276, "kl": 0.022735595703125, "learning_rate": 4.326933803713029e-06, "loss": 0.0002, "num_tokens": 84123790.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 711 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2761830876648565, "grad_norm": 0.005593872629106045, "kl": 0.022369384765625, "learning_rate": 4.324915752663833e-06, "loss": 0.0002, "num_tokens": 84501734.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 712 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2765709852598914, "grad_norm": 0.006202434189617634, "kl": 0.023284912109375, "learning_rate": 4.322895214607692e-06, "loss": 0.0002, "num_tokens": 84873430.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 713 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2769588828549263, "grad_norm": 0.006379022728651762, "kl": 0.027801513671875, "learning_rate": 4.320872192735314e-06, "loss": 0.0003, "num_tokens": 85246422.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 714 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2773467804499612, "grad_norm": 0.007177534978836775, "kl": 0.02288818359375, "learning_rate": 4.3188466902413246e-06, "loss": 0.0002, "num_tokens": 85630278.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 715 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27773467804499613, "grad_norm": 0.004926490131765604, "kl": 0.021331787109375, "learning_rate": 4.316818710324269e-06, "loss": 0.0002, "num_tokens": 86003030.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 716 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.278122575640031, "grad_norm": 0.005382516887038946, "kl": 0.02203369140625, "learning_rate": 4.314788256186606e-06, "loss": 0.0002, "num_tokens": 86377902.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 717 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27851047323506595, "grad_norm": 0.004839687142521143, "kl": 0.022674560546875, "learning_rate": 4.312755331034698e-06, "loss": 0.0002, "num_tokens": 86746878.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 718 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27889837083010083, "grad_norm": 0.005019054282456636, "kl": 0.02276611328125, "learning_rate": 4.3107199380788115e-06, "loss": 0.0002, "num_tokens": 87120782.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 719 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.27928626842513576, "grad_norm": 0.004998947028070688, "kl": 0.0225830078125, "learning_rate": 4.30868208053311e-06, "loss": 0.0002, "num_tokens": 87495566.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 720 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2796741660201707, "grad_norm": 305.5563049316406, "kl": 0.242828369140625, "learning_rate": 4.3066417616156496e-06, "loss": 0.0024, "num_tokens": 87872118.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 721 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2800620636152056, "grad_norm": 0.00457870913669467, "kl": 0.022491455078125, "learning_rate": 4.304598984548372e-06, "loss": 0.0002, "num_tokens": 88243358.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 722 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2804499612102405, "grad_norm": 0.016216536983847618, "kl": 0.02471923828125, "learning_rate": 4.3025537525570996e-06, "loss": 0.0002, "num_tokens": 88615638.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 723 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2808378588052754, "grad_norm": 0.03139244765043259, "kl": 0.02490234375, "learning_rate": 4.300506068871534e-06, "loss": 0.0002, "num_tokens": 88988774.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 724 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28122575640031033, "grad_norm": 0.00992588885128498, "kl": 0.025604248046875, "learning_rate": 4.298455936725248e-06, "loss": 0.0003, "num_tokens": 89364758.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 725 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2816136539953452, "grad_norm": 0.009371595457196236, "kl": 0.03082275390625, "learning_rate": 4.2964033593556785e-06, "loss": 0.0003, "num_tokens": 89733606.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 726 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28200155159038015, "grad_norm": 0.00547827547416091, "kl": 0.023040771484375, "learning_rate": 4.294348340004127e-06, "loss": 0.0002, "num_tokens": 90115918.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 727 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28238944918541503, "grad_norm": 0.056359633803367615, "kl": 0.029632568359375, "learning_rate": 4.292290881915748e-06, "loss": 0.0003, "num_tokens": 90497358.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 728 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28277734678044997, "grad_norm": 0.006745846476405859, "kl": 0.02484130859375, "learning_rate": 4.29023098833955e-06, "loss": 0.0002, "num_tokens": 90870742.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 729 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28316524437548485, "grad_norm": 0.005658056121319532, "kl": 0.025299072265625, "learning_rate": 4.288168662528385e-06, "loss": 0.0003, "num_tokens": 91248742.0, "reward": 0.020000001415610313, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.0603785514831543, "step": 730 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2835531419705198, "grad_norm": 0.20545116066932678, "kl": 0.025115966796875, "learning_rate": 4.2861039077389475e-06, "loss": 0.0003, "num_tokens": 91626446.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 731 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2839410395655547, "grad_norm": 0.005221615545451641, "kl": 0.0245361328125, "learning_rate": 4.284036727231767e-06, "loss": 0.0002, "num_tokens": 91997566.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 732 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2843289371605896, "grad_norm": 0.01572037860751152, "kl": 0.02496337890625, "learning_rate": 4.2819671242712034e-06, "loss": 0.0002, "num_tokens": 92371590.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 733 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28471683475562454, "grad_norm": 0.0046168724074959755, "kl": 0.024322509765625, "learning_rate": 4.279895102125444e-06, "loss": 0.0002, "num_tokens": 92746998.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 734 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2851047323506594, "grad_norm": 0.006837782915681601, "kl": 0.02508544921875, "learning_rate": 4.277820664066492e-06, "loss": 0.0003, "num_tokens": 93125990.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 735 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28549262994569435, "grad_norm": 0.006183963268995285, "kl": 0.023590087890625, "learning_rate": 4.2757438133701695e-06, "loss": 0.0002, "num_tokens": 93497870.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 736 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28588052754072923, "grad_norm": 0.005340836942195892, "kl": 0.023895263671875, "learning_rate": 4.2736645533161075e-06, "loss": 0.0002, "num_tokens": 93881438.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 737 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28626842513576417, "grad_norm": 0.005127822980284691, "kl": 0.02581787109375, "learning_rate": 4.271582887187742e-06, "loss": 0.0003, "num_tokens": 94267038.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 738 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28665632273079905, "grad_norm": 0.004898578394204378, "kl": 0.02392578125, "learning_rate": 4.269498818272305e-06, "loss": 0.0002, "num_tokens": 94644670.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 739 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.287044220325834, "grad_norm": 0.0052308873273432255, "kl": 0.024444580078125, "learning_rate": 4.267412349860829e-06, "loss": 0.0002, "num_tokens": 95029126.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 740 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28743211792086887, "grad_norm": 0.005041634663939476, "kl": 0.02337646484375, "learning_rate": 4.2653234852481315e-06, "loss": 0.0002, "num_tokens": 95403638.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 741 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2878200155159038, "grad_norm": 0.006597405299544334, "kl": 0.023101806640625, "learning_rate": 4.263232227732815e-06, "loss": 0.0002, "num_tokens": 95776646.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 742 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2882079131109387, "grad_norm": 0.0050996048375964165, "kl": 0.02447509765625, "learning_rate": 4.2611385806172595e-06, "loss": 0.0002, "num_tokens": 96162590.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.3018927276134491, "step": 743 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2885958107059736, "grad_norm": 0.006620405241847038, "kl": 0.024871826171875, "learning_rate": 4.25904254720762e-06, "loss": 0.0002, "num_tokens": 96540542.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 744 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28898370830100856, "grad_norm": 0.009507874958217144, "kl": 0.026214599609375, "learning_rate": 4.25694413081382e-06, "loss": 0.0003, "num_tokens": 96918798.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 745 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.28937160589604344, "grad_norm": 0.005583560094237328, "kl": 0.02435302734375, "learning_rate": 4.2548433347495456e-06, "loss": 0.0002, "num_tokens": 97295734.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 746 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2897595034910784, "grad_norm": 0.005060376599431038, "kl": 0.0238037109375, "learning_rate": 4.252740162332239e-06, "loss": 0.0002, "num_tokens": 97681870.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 747 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.29014740108611325, "grad_norm": 0.004713468719273806, "kl": 0.025848388671875, "learning_rate": 4.250634616883099e-06, "loss": 0.0003, "num_tokens": 98060990.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 748 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2905352986811482, "grad_norm": 0.004439684096723795, "kl": 0.023223876953125, "learning_rate": 4.248526701727066e-06, "loss": 0.0002, "num_tokens": 98429934.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 749 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2976.0, "completions/mean_length": 4082.0, "completions/mean_terminated_length": 2976.0, "completions/min_length": 2976.0, "completions/min_terminated_length": 2976.0, "epoch": 0.29092319627618307, "grad_norm": 0.04380488023161888, "kl": 0.027862548828125, "learning_rate": 4.246416420192829e-06, "loss": 0.0003, "num_tokens": 98811926.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 750 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.291311093871218, "grad_norm": 0.007455464918166399, "kl": 0.02685546875, "learning_rate": 4.244303775612809e-06, "loss": 0.0003, "num_tokens": 99192270.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 751 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2916989914662529, "grad_norm": 0.0388723649084568, "kl": 0.032958984375, "learning_rate": 4.242188771323159e-06, "loss": 0.0003, "num_tokens": 99572646.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 752 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2920868890612878, "grad_norm": 2.7323622703552246, "kl": 0.04351806640625, "learning_rate": 4.240071410663762e-06, "loss": 0.0004, "num_tokens": 99958470.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 753 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2924747866563227, "grad_norm": 0.007685428485274315, "kl": 0.028228759765625, "learning_rate": 4.237951696978217e-06, "loss": 0.0003, "num_tokens": 100338566.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 754 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.29286268425135764, "grad_norm": 0.00806245394051075, "kl": 0.02618408203125, "learning_rate": 4.2358296336138435e-06, "loss": 0.0003, "num_tokens": 100713814.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 755 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2932505818463926, "grad_norm": 0.005284970626235008, "kl": 0.025360107421875, "learning_rate": 4.233705223921667e-06, "loss": 0.0003, "num_tokens": 101089238.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 756 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.29363847944142746, "grad_norm": 0.023294668644666672, "kl": 0.02801513671875, "learning_rate": 4.231578471256422e-06, "loss": 0.0003, "num_tokens": 101470886.0, "reward": 0.02500000037252903, "reward_std": 0.04629100486636162, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.15710999071598053, "step": 757 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2940263770364624, "grad_norm": 0.0066102417185902596, "kl": 0.02801513671875, "learning_rate": 4.2294493789765425e-06, "loss": 0.0003, "num_tokens": 101844918.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 758 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2944142746314973, "grad_norm": 0.006387398112565279, "kl": 0.026153564453125, "learning_rate": 4.227317950444152e-06, "loss": 0.0003, "num_tokens": 102221798.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 759 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2948021722265322, "grad_norm": 0.02980141155421734, "kl": 0.03009033203125, "learning_rate": 4.22518418902507e-06, "loss": 0.0003, "num_tokens": 102598478.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 760 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2951900698215671, "grad_norm": 0.007296744268387556, "kl": 0.026641845703125, "learning_rate": 4.223048098088797e-06, "loss": 0.0003, "num_tokens": 102979422.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 761 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.295577967416602, "grad_norm": 0.03271764889359474, "kl": 0.026947021484375, "learning_rate": 4.22090968100851e-06, "loss": 0.0003, "num_tokens": 103354382.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 762 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2959658650116369, "grad_norm": 0.011606186628341675, "kl": 0.02923583984375, "learning_rate": 4.218768941161064e-06, "loss": 0.0003, "num_tokens": 103745886.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 763 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.29635376260667184, "grad_norm": 0.010566490702331066, "kl": 0.030364990234375, "learning_rate": 4.216625881926978e-06, "loss": 0.0003, "num_tokens": 104120918.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 764 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2967416602017067, "grad_norm": 0.006433264818042517, "kl": 0.02679443359375, "learning_rate": 4.214480506690435e-06, "loss": 0.0003, "num_tokens": 104494614.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 765 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.29712955779674166, "grad_norm": 0.04954459145665169, "kl": 0.038360595703125, "learning_rate": 4.2123328188392755e-06, "loss": 0.0004, "num_tokens": 104868926.0, "reward": 0.003750000149011612, "reward_std": 0.008164633996784687, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 766 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 395.0, "completions/mean_length": 4049.737548828125, "completions/mean_terminated_length": 395.0, "completions/min_length": 395.0, "completions/min_terminated_length": 395.0, "epoch": 0.2975174553917766, "grad_norm": 0.028313281014561653, "kl": 0.028900146484375, "learning_rate": 4.210182821764992e-06, "loss": 0.0286, "num_tokens": 105234609.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 767 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2979053529868115, "grad_norm": 0.009696447290480137, "kl": 0.029937744140625, "learning_rate": 4.2080305188627235e-06, "loss": 0.0003, "num_tokens": 105601769.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 768 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2982932505818464, "grad_norm": 0.027699870988726616, "kl": 0.027862548828125, "learning_rate": 4.205875913531251e-06, "loss": 0.0003, "num_tokens": 105984857.0, "reward": 0.002500000176951289, "reward_std": 0.004629100672900677, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 769 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2986811481768813, "grad_norm": 0.02361246943473816, "kl": 0.029754638671875, "learning_rate": 4.203719009172988e-06, "loss": 0.0003, "num_tokens": 106359457.0, "reward": 0.005000000353902578, "reward_std": 0.005345224868506193, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 770 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 395.0, "completions/mean_length": 4002.400146484375, "completions/mean_terminated_length": 352.0, "completions/min_length": 309.0, "completions/min_terminated_length": 309.0, "epoch": 0.29906904577191623, "grad_norm": 0.03961588814854622, "kl": 0.03289794921875, "learning_rate": 4.2015598091939815e-06, "loss": 0.0292, "num_tokens": 106730185.0, "reward": 0.0062500000931322575, "reward_std": 0.00980459339916706, "rewards/format_reward/mean": 0.0625, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 771 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.2994569433669511, "grad_norm": 0.01738235540688038, "kl": 0.029571533203125, "learning_rate": 4.1993983170039045e-06, "loss": 0.0003, "num_tokens": 107101801.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 772 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.29984484096198605, "grad_norm": 0.03652869537472725, "kl": 0.030914306640625, "learning_rate": 4.197234536016046e-06, "loss": 0.0003, "num_tokens": 107478409.0, "reward": 0.0062500000931322575, "reward_std": 0.0176776684820652, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.044721364974975586, "step": 773 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.30023273855702093, "grad_norm": 0.006859822664409876, "kl": 0.029693603515625, "learning_rate": 4.195068469647315e-06, "loss": 0.0003, "num_tokens": 107858265.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 774 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.30062063615205586, "grad_norm": 0.036865297704935074, "kl": 0.030059814453125, "learning_rate": 4.192900121318223e-06, "loss": 0.0003, "num_tokens": 108233073.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 775 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 416.0, "completions/mean_length": 4050.0, "completions/mean_terminated_length": 416.0, "completions/min_length": 416.0, "completions/min_terminated_length": 416.0, "epoch": 0.30100853374709075, "grad_norm": 0.0351514108479023, "kl": 0.030426025390625, "learning_rate": 4.190729494452889e-06, "loss": 0.0297, "num_tokens": 108605985.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 776 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 336.0, "completions/mean_length": 4049.0, "completions/mean_terminated_length": 336.0, "completions/min_length": 336.0, "completions/min_terminated_length": 336.0, "epoch": 0.3013964313421257, "grad_norm": 0.029969869181513786, "kl": 0.03094482421875, "learning_rate": 4.1885565924790286e-06, "loss": 0.0184, "num_tokens": 108971169.0, "reward": 0.002500000176951289, "reward_std": 0.004629100672900677, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 777 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 249.0, "completions/mean_length": 4047.91259765625, "completions/mean_terminated_length": 249.0, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "epoch": 0.30178432893716056, "grad_norm": 0.04163690283894539, "kl": 0.030731201171875, "learning_rate": 4.186381418827951e-06, "loss": -0.0009, "num_tokens": 109345874.0, "reward": 0.013750000856816769, "reward_std": 0.03502550348639488, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 778 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 494.0, "completions/mean_length": 4004.074951171875, "completions/mean_terminated_length": 419.0, "completions/min_length": 344.0, "completions/min_terminated_length": 344.0, "epoch": 0.3021722265321955, "grad_norm": 0.03637642785906792, "kl": 0.03106689453125, "learning_rate": 4.184203976934552e-06, "loss": 0.0393, "num_tokens": 109714800.0, "reward": 0.002500000176951289, "reward_std": 0.004629100672900677, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 779 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 308.0, "completions/mean_length": 4001.0625, "completions/mean_terminated_length": 298.5, "completions/min_length": 289.0, "completions/min_terminated_length": 289.0, "epoch": 0.30256012412723043, "grad_norm": 0.05782780423760414, "kl": 0.030517578125, "learning_rate": 4.18202427023731e-06, "loss": 0.0263, "num_tokens": 110078549.0, "reward": 0.005000000353902578, "reward_std": 0.008711026050150394, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 780 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 404.0, "completions/mean_length": 3954.712646484375, "completions/mean_terminated_length": 328.3333435058594, "completions/min_length": 254.0, "completions/min_terminated_length": 254.0, "epoch": 0.3029480217222653, "grad_norm": 0.10995858907699585, "kl": 0.03448486328125, "learning_rate": 4.17984230217828e-06, "loss": 0.0631, "num_tokens": 110437686.0, "reward": 0.005000000353902578, "reward_std": 0.011700167320668697, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 781 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.30333591931730025, "grad_norm": 0.016963472589850426, "kl": 0.035064697265625, "learning_rate": 4.177658076203085e-06, "loss": 0.0004, "num_tokens": 110815118.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 782 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.30372381691233513, "grad_norm": 0.010335095226764679, "kl": 0.0428466796875, "learning_rate": 4.175471595760919e-06, "loss": 0.0004, "num_tokens": 111194630.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 783 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 498.0, "completions/mean_length": 4005.199951171875, "completions/mean_terminated_length": 464.0, "completions/min_length": 430.0, "completions/min_terminated_length": 430.0, "epoch": 0.30411171450737007, "grad_norm": 0.05281098559498787, "kl": 0.03302001953125, "learning_rate": 4.17328286430453e-06, "loss": 0.0556, "num_tokens": 111572238.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 784 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.30449961210240495, "grad_norm": 0.05063151568174362, "kl": 0.032684326171875, "learning_rate": 4.171091885290226e-06, "loss": 0.0003, "num_tokens": 111947342.0, "reward": 0.003750000149011612, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360680624842644, "step": 785 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3048875096974399, "grad_norm": 0.015285806730389595, "kl": 0.03387451171875, "learning_rate": 4.168898662177862e-06, "loss": 0.0003, "num_tokens": 112327078.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 786 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 554.0, "completions/mean_length": 3959.425048828125, "completions/mean_terminated_length": 454.0, "completions/min_length": 307.0, "completions/min_terminated_length": 307.0, "epoch": 0.30527540729247477, "grad_norm": 0.06301887333393097, "kl": 0.0364990234375, "learning_rate": 4.166703198430835e-06, "loss": 0.027, "num_tokens": 112689944.0, "reward": 0.013750001788139343, "reward_std": 0.024939196184277534, "rewards/format_reward/mean": 0.0625, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 787 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 394.0, "completions/mean_length": 3956.10009765625, "completions/mean_terminated_length": 365.3333435058594, "completions/min_length": 344.0, "completions/min_terminated_length": 344.0, "epoch": 0.3056633048875097, "grad_norm": 0.05109215900301933, "kl": 0.03546142578125, "learning_rate": 4.164505497516084e-06, "loss": 0.0594, "num_tokens": 113054576.0, "reward": 0.005000000353902578, "reward_std": 0.008711026050150394, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 788 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 387.0, "completions/mean_length": 4049.637451171875, "completions/mean_terminated_length": 387.0, "completions/min_length": 387.0, "completions/min_terminated_length": 387.0, "epoch": 0.3060512024825446, "grad_norm": 0.07829269766807556, "kl": 0.05194091796875, "learning_rate": 4.162305562904077e-06, "loss": 0.0289, "num_tokens": 113428891.0, "reward": 0.003750000149011612, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 789 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 490.0, "completions/mean_length": 4003.925048828125, "completions/mean_terminated_length": 413.0, "completions/min_length": 336.0, "completions/min_terminated_length": 336.0, "epoch": 0.3064391000775795, "grad_norm": 0.05220162495970726, "kl": 0.0372314453125, "learning_rate": 4.160103398068811e-06, "loss": 0.0579, "num_tokens": 113803045.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 790 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9125, "completions/max_length": 4096.0, "completions/max_terminated_length": 369.0, "completions/mean_length": 3765.112548828125, "completions/mean_terminated_length": 314.4285888671875, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.30682699767261445, "grad_norm": 0.07437063008546829, "kl": 0.0458984375, "learning_rate": 4.157899006487806e-06, "loss": 0.0903, "num_tokens": 114147950.0, "reward": 0.012500000186264515, "reward_std": 0.013340127654373646, "rewards/format_reward/mean": 0.125, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 791 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 349.0, "completions/mean_length": 3809.637451171875, "completions/mean_terminated_length": 277.8333435058594, "completions/min_length": 217.0, "completions/min_terminated_length": 217.0, "epoch": 0.30721489526764933, "grad_norm": 0.07203198224306107, "kl": 0.0400390625, "learning_rate": 4.155692391642096e-06, "loss": 0.0827, "num_tokens": 114493377.0, "reward": 0.013750000856816769, "reward_std": 0.022043664008378983, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 792 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 437.0, "completions/mean_length": 3815.625, "completions/mean_terminated_length": 357.66668701171875, "completions/min_length": 238.0, "completions/min_terminated_length": 238.0, "epoch": 0.30760279286268427, "grad_norm": 0.07926422357559204, "kl": 0.047119140625, "learning_rate": 4.153483557016228e-06, "loss": 0.1012, "num_tokens": 114849555.0, "reward": 0.0162500012665987, "reward_std": 0.031124252825975418, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472135752439499, "step": 793 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.30799069045771915, "grad_norm": 0.04874833673238754, "kl": 0.03759765625, "learning_rate": 4.151272506098254e-06, "loss": 0.0004, "num_tokens": 115229499.0, "reward": 0.007500000298023224, "reward_std": 0.02121320366859436, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 794 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 363.0, "completions/mean_length": 4001.612548828125, "completions/mean_terminated_length": 320.5, "completions/min_length": 278.0, "completions/min_terminated_length": 278.0, "epoch": 0.3083785880527541, "grad_norm": 0.041505344212055206, "kl": 0.03912353515625, "learning_rate": 4.149059242379722e-06, "loss": 0.0039, "num_tokens": 115592276.0, "reward": 0.023750001564621925, "reward_std": 0.05572747066617012, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.12965388596057892, "step": 795 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 463.0, "completions/mean_length": 3859.824951171875, "completions/mean_terminated_length": 317.20001220703125, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.30876648564778897, "grad_norm": 0.07149551808834076, "kl": 0.043212890625, "learning_rate": 4.146843769355682e-06, "loss": 0.0443, "num_tokens": 115954694.0, "reward": 0.021250000223517418, "reward_std": 0.04106973484158516, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.09192044287919998, "step": 796 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 299.0, "completions/mean_length": 3856.53759765625, "completions/mean_terminated_length": 264.6000061035156, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "epoch": 0.3091543832428239, "grad_norm": 0.06827827543020248, "kl": 0.0401611328125, "learning_rate": 4.144626090524665e-06, "loss": -0.0009, "num_tokens": 116308961.0, "reward": 0.017500001937150955, "reward_std": 0.030141903087496758, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531530380249, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.06284399330615997, "step": 797 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 369.0, "completions/mean_length": 3955.987548828125, "completions/mean_terminated_length": 362.3333435058594, "completions/min_length": 355.0, "completions/min_terminated_length": 355.0, "epoch": 0.3095422808378588, "grad_norm": 0.1126488670706749, "kl": 0.038818359375, "learning_rate": 4.142406209388691e-06, "loss": 0.0264, "num_tokens": 116669408.0, "reward": 0.020000001415610313, "reward_std": 0.04633110761642456, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.09426598995923996, "step": 798 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 504.0, "completions/mean_length": 3955.91259765625, "completions/mean_terminated_length": 360.3333435058594, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.3099301784328937, "grad_norm": 0.1108052209019661, "kl": 0.044677734375, "learning_rate": 4.140184129453253e-06, "loss": 0.0437, "num_tokens": 117026793.0, "reward": 0.013750000856816769, "reward_std": 0.019864803180098534, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 799 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 430.0, "completions/mean_length": 3957.1875, "completions/mean_terminated_length": 394.3333435058594, "completions/min_length": 372.0, "completions/min_terminated_length": 372.0, "epoch": 0.3103180760279286, "grad_norm": 0.059603501111269, "kl": 0.04315185546875, "learning_rate": 4.137959854227319e-06, "loss": 0.0384, "num_tokens": 117392760.0, "reward": 0.023750001564621925, "reward_std": 0.042208295315504074, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.12568798661231995, "step": 800 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 465.0, "completions/mean_length": 3677.66259765625, "completions/mean_terminated_length": 377.4444580078125, "completions/min_length": 280.0, "completions/min_terminated_length": 280.0, "epoch": 0.31070597362296354, "grad_norm": 0.0876493975520134, "kl": 0.04962158203125, "learning_rate": 4.135733387223325e-06, "loss": 0.0604, "num_tokens": 117725541.0, "reward": 0.057500001043081284, "reward_std": 0.08996322005987167, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.16866940259933472, "step": 801 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 447.0, "completions/mean_length": 3721.449951171875, "completions/mean_terminated_length": 350.5, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.3110938712179985, "grad_norm": 1.46126389503479, "kl": 0.047119140625, "learning_rate": 4.133504731957163e-06, "loss": 0.0379, "num_tokens": 118063457.0, "reward": 0.125, "reward_std": 0.19580483436584473, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.26624438166618347, "step": 802 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 546.0, "completions/mean_length": 3911.78759765625, "completions/mean_terminated_length": 411.75, "completions/min_length": 299.0, "completions/min_terminated_length": 299.0, "epoch": 0.31148176881303336, "grad_norm": 0.10018559545278549, "kl": 0.05145263671875, "learning_rate": 4.131273891948187e-06, "loss": 0.0466, "num_tokens": 118425568.0, "reward": 0.055000003427267075, "reward_std": 0.09071048349142075, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.04749999940395355, "rewards/unicoder_reward_fn/std": 0.17787422239780426, "step": 803 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 333.0, "completions/mean_length": 3953.762451171875, "completions/mean_terminated_length": 303.0, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.3118696664080683, "grad_norm": 0.09739907830953598, "kl": 0.044921875, "learning_rate": 4.129040870719198e-06, "loss": 0.0607, "num_tokens": 118798533.0, "reward": 0.0625, "reward_std": 0.0781475305557251, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.14917494356632233, "step": 804 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 519.0, "completions/mean_length": 3350.16259765625, "completions/mean_terminated_length": 366.8125, "completions/min_length": 268.0, "completions/min_terminated_length": 268.0, "epoch": 0.31225756400310317, "grad_norm": 0.11334013193845749, "kl": 0.05609130859375, "learning_rate": 4.126805671796438e-06, "loss": 0.0601, "num_tokens": 119120850.0, "reward": 0.08125000447034836, "reward_std": 0.1048998013138771, "rewards/format_reward/mean": 0.26249998807907104, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.19086743891239166, "step": 805 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 619.0, "completions/mean_length": 3589.337646484375, "completions/mean_terminated_length": 411.18182373046875, "completions/min_length": 264.0, "completions/min_terminated_length": 264.0, "epoch": 0.3126454615981381, "grad_norm": 0.12469179928302765, "kl": 0.05255126953125, "learning_rate": 4.1245682987095945e-06, "loss": 0.1055, "num_tokens": 119459477.0, "reward": 0.03875000402331352, "reward_std": 0.053834009915590286, "rewards/format_reward/mean": 0.21250000596046448, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.09648899734020233, "step": 806 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 565.0, "completions/mean_length": 3493.112548828125, "completions/mean_terminated_length": 385.923095703125, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "epoch": 0.313033359193173, "grad_norm": 0.13700808584690094, "kl": 0.0567626953125, "learning_rate": 4.1223287549917835e-06, "loss": 0.1551, "num_tokens": 119784614.0, "reward": 0.07000000774860382, "reward_std": 0.05445485934615135, "rewards/format_reward/mean": 0.22499999403953552, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.04749999940395355, "rewards/unicoder_reward_fn/std": 0.11580756306648254, "step": 807 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.85, "completions/max_length": 4096.0, "completions/max_terminated_length": 490.0, "completions/mean_length": 3532.0, "completions/mean_terminated_length": 336.0, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.3134212567882079, "grad_norm": 0.13797122240066528, "kl": 0.05322265625, "learning_rate": 4.1200870441795496e-06, "loss": 0.0514, "num_tokens": 120108814.0, "reward": 0.11500000208616257, "reward_std": 0.11095046997070312, "rewards/format_reward/mean": 0.17499999701976776, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.23812292516231537, "step": 808 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7625, "completions/max_length": 4096.0, "completions/max_terminated_length": 462.0, "completions/mean_length": 3215.587646484375, "completions/mean_terminated_length": 389.0, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "epoch": 0.3138091543832428, "grad_norm": 0.7634460926055908, "kl": 0.07177734375, "learning_rate": 4.117843169812863e-06, "loss": 0.1219, "num_tokens": 120412485.0, "reward": 0.12000000476837158, "reward_std": 0.09073217213153839, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.18540069460868835, "step": 809 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 667.0, "completions/mean_length": 2979.737548828125, "completions/mean_terminated_length": 375.125, "completions/min_length": 245.0, "completions/min_terminated_length": 245.0, "epoch": 0.31419705197827774, "grad_norm": 0.16730360686779022, "kl": 0.06512451171875, "learning_rate": 4.115597135435103e-06, "loss": 0.0576, "num_tokens": 120705512.0, "reward": 0.07999999821186066, "reward_std": 0.08507223427295685, "rewards/format_reward/mean": 0.3499999940395355, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.15903188288211823, "step": 810 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 641.0, "completions/mean_length": 3027.862548828125, "completions/mean_terminated_length": 380.7391357421875, "completions/min_length": 217.0, "completions/min_terminated_length": 217.0, "epoch": 0.3145849495733126, "grad_norm": 0.4137221872806549, "kl": 0.09075927734375, "learning_rate": 4.113348944593069e-06, "loss": 0.1848, "num_tokens": 120995741.0, "reward": 0.14374999701976776, "reward_std": 0.07293296605348587, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.19051894545555115, "step": 811 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 447.0, "completions/mean_length": 3817.125, "completions/mean_terminated_length": 377.66668701171875, "completions/min_length": 313.0, "completions/min_terminated_length": 313.0, "epoch": 0.31497284716834756, "grad_norm": 0.26148542761802673, "kl": 0.064453125, "learning_rate": 4.1110986008369595e-06, "loss": -0.07, "num_tokens": 121345303.0, "reward": 0.15250001847743988, "reward_std": 0.12633608281612396, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.25452861189842224, "step": 812 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 464.0, "completions/mean_length": 3489.9375, "completions/mean_terminated_length": 366.3846435546875, "completions/min_length": 260.0, "completions/min_terminated_length": 260.0, "epoch": 0.3153607447633825, "grad_norm": 0.2976844608783722, "kl": 0.072021484375, "learning_rate": 4.108846107720373e-06, "loss": -0.0692, "num_tokens": 121674066.0, "reward": 0.1599999964237213, "reward_std": 0.10657472908496857, "rewards/format_reward/mean": 0.17499999701976776, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.29025307297706604, "step": 813 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 442.0, "completions/mean_length": 3722.087646484375, "completions/mean_terminated_length": 356.875, "completions/min_length": 267.0, "completions/min_terminated_length": 267.0, "epoch": 0.3157486423584174, "grad_norm": 0.10205190628767014, "kl": 0.06390380859375, "learning_rate": 4.106591468800306e-06, "loss": 0.0485, "num_tokens": 122021297.0, "reward": 0.0962500050663948, "reward_std": 0.015235701575875282, "rewards/format_reward/mean": 0.13750000298023224, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.18471670150756836, "step": 814 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 511.0, "completions/mean_length": 2973.962646484375, "completions/mean_terminated_length": 355.875, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.3161365399534523, "grad_norm": 0.3915945291519165, "kl": 0.092041015625, "learning_rate": 4.104334687637139e-06, "loss": 0.1201, "num_tokens": 122301806.0, "reward": 0.05750000476837158, "reward_std": 0.033374663442373276, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.02500000223517418, "rewards/unicoder_reward_fn/std": 0.06656109541654587, "step": 815 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7625, "completions/max_length": 4096.0, "completions/max_terminated_length": 664.0, "completions/mean_length": 3228.612548828125, "completions/mean_terminated_length": 443.84210205078125, "completions/min_length": 345.0, "completions/min_terminated_length": 345.0, "epoch": 0.3165244375484872, "grad_norm": 0.30821049213409424, "kl": 0.07855224609375, "learning_rate": 4.102075767794639e-06, "loss": 0.0732, "num_tokens": 122617231.0, "reward": 0.29875001311302185, "reward_std": 0.04618903994560242, "rewards/format_reward/mean": 0.26249998807907104, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.27250000834465027, "rewards/unicoder_reward_fn/std": 0.3673287034034729, "step": 816 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 604.0, "completions/mean_length": 1985.9124755859375, "completions/mean_terminated_length": 344.73333740234375, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.31691233514352213, "grad_norm": 5.275448322296143, "kl": 0.1322021484375, "learning_rate": 4.0998147128399485e-06, "loss": 0.1846, "num_tokens": 122823416.0, "reward": 0.10750000923871994, "reward_std": 0.049242641776800156, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.04749999940395355, "rewards/unicoder_reward_fn/std": 0.11580756306648254, "step": 817 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 953.0, "completions/mean_length": 3365.962646484375, "completions/mean_terminated_length": 445.8125, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.317300232738557, "grad_norm": 0.1263582408428192, "kl": 0.06109619140625, "learning_rate": 4.097551526343581e-06, "loss": -0.0225, "num_tokens": 123142509.0, "reward": 0.24124999344348907, "reward_std": 0.0679134875535965, "rewards/format_reward/mean": 0.21250000596046448, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.2199999988079071, "rewards/unicoder_reward_fn/std": 0.24772383272647858, "step": 818 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 604.0, "completions/mean_length": 3717.25, "completions/mean_terminated_length": 308.5, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.31768813033359194, "grad_norm": 0.1769849956035614, "kl": 0.05865478515625, "learning_rate": 4.095286211879418e-06, "loss": 0.0687, "num_tokens": 123483841.0, "reward": 0.33500000834465027, "reward_std": 0.059361666440963745, "rewards/format_reward/mean": 0.125, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.32249999046325684, "rewards/unicoder_reward_fn/std": 0.3758450448513031, "step": 819 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1652.0, "completions/mean_length": 2394.175048828125, "completions/mean_terminated_length": 416.3783874511719, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.3180760279286268, "grad_norm": 0.5682273507118225, "kl": 0.1513671875, "learning_rate": 4.093018773024699e-06, "loss": 0.0698, "num_tokens": 123722975.0, "reward": 0.07875000685453415, "reward_std": 0.07092573493719101, "rewards/format_reward/mean": 0.512499988079071, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.13404609262943268, "step": 820 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 2551.0, "completions/mean_length": 3023.762451171875, "completions/mean_terminated_length": 521.875, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "epoch": 0.31846392552366176, "grad_norm": 0.44955238699913025, "kl": 0.1201171875, "learning_rate": 4.0907492133600205e-06, "loss": -0.1745, "num_tokens": 124007668.0, "reward": 0.20250001549720764, "reward_std": 0.13045771420001984, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.1599999964237213, "rewards/unicoder_reward_fn/std": 0.22591586410999298, "step": 821 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 607.0, "completions/mean_length": 2859.8125, "completions/mean_terminated_length": 433.22222900390625, "completions/min_length": 288.0, "completions/min_terminated_length": 288.0, "epoch": 0.31885182311869664, "grad_norm": 0.38860437273979187, "kl": 0.095947265625, "learning_rate": 4.088477536469327e-06, "loss": 0.3365, "num_tokens": 124285261.0, "reward": 0.14874999225139618, "reward_std": 0.0703776478767395, "rewards/format_reward/mean": 0.36250001192092896, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.22240059077739716, "step": 822 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.65, "completions/max_length": 4096.0, "completions/max_terminated_length": 2330.0, "completions/mean_length": 2833.337646484375, "completions/mean_terminated_length": 488.39288330078125, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.3192397207137316, "grad_norm": 0.7526530623435974, "kl": 0.22021484375, "learning_rate": 4.086203745939906e-06, "loss": 0.1783, "num_tokens": 124559288.0, "reward": 0.1550000011920929, "reward_std": 0.08903271704912186, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.20584805309772491, "step": 823 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3657.0, "completions/mean_length": 2360.925048828125, "completions/mean_terminated_length": 710.48779296875, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.31962761830876646, "grad_norm": 1.9124289751052856, "kl": 0.26025390625, "learning_rate": 4.083927845362384e-06, "loss": 0.0995, "num_tokens": 124812122.0, "reward": 0.10499999672174454, "reward_std": 0.1015043780207634, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.1550622582435608, "step": 824 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3796.0, "completions/mean_length": 831.3624877929688, "completions/mean_terminated_length": 566.6621704101562, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.3200155159038014, "grad_norm": 1.9814801216125488, "kl": 0.3447265625, "learning_rate": 4.081649838330717e-06, "loss": 0.0886, "num_tokens": 124917199.0, "reward": 0.12250000238418579, "reward_std": 0.07339345663785934, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.13416408002376556, "step": 825 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 4030.0, "completions/mean_length": 944.8125, "completions/mean_terminated_length": 904.924072265625, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "epoch": 0.32040341349883633, "grad_norm": 1.500654697418213, "kl": 0.37841796875, "learning_rate": 4.07936972844219e-06, "loss": 0.1411, "num_tokens": 125041864.0, "reward": 0.10750000923871994, "reward_std": 0.05254964902997017, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.0725223645567894, "step": 826 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3824.0, "completions/max_terminated_length": 3824.0, "completions/mean_length": 622.75, "completions/mean_terminated_length": 622.75, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.3207913110938712, "grad_norm": 1.244160771369934, "kl": 0.4013671875, "learning_rate": 4.077087519297407e-06, "loss": -0.019, "num_tokens": 125141916.0, "reward": 0.16750001907348633, "reward_std": 0.12639085948467255, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.07249999791383743, "rewards/unicoder_reward_fn/std": 0.20436689257621765, "step": 827 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3710.0, "completions/mean_length": 1533.125, "completions/mean_terminated_length": 1287.369873046875, "completions/min_length": 224.0, "completions/min_terminated_length": 224.0, "epoch": 0.32117920868890615, "grad_norm": 1.033677339553833, "kl": 0.4765625, "learning_rate": 4.074803214500289e-06, "loss": 0.0168, "num_tokens": 125306782.0, "reward": 0.21625001728534698, "reward_std": 0.144914910197258, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.24699458479881287, "step": 828 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3307.0, "completions/max_terminated_length": 3307.0, "completions/mean_length": 609.4625244140625, "completions/mean_terminated_length": 609.4625244140625, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.32156710628394103, "grad_norm": 0.5882471799850464, "kl": 0.3818359375, "learning_rate": 4.072516817658065e-06, "loss": -0.0276, "num_tokens": 125396931.0, "reward": 0.18000000715255737, "reward_std": 0.12475258111953735, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.20547877252101898, "step": 829 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2288.0, "completions/max_terminated_length": 2288.0, "completions/mean_length": 644.1000366210938, "completions/mean_terminated_length": 644.1000366210938, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "epoch": 0.32195500387897596, "grad_norm": 0.6730323433876038, "kl": 0.337890625, "learning_rate": 4.070228332381267e-06, "loss": 0.0246, "num_tokens": 125498299.0, "reward": 0.2912500202655792, "reward_std": 0.1379171460866928, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.2745652496814728, "step": 830 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2738.0, "completions/max_terminated_length": 2738.0, "completions/mean_length": 748.0125122070312, "completions/mean_terminated_length": 748.0125122070312, "completions/min_length": 257.0, "completions/min_terminated_length": 257.0, "epoch": 0.32234290147401085, "grad_norm": 0.8663316369056702, "kl": 0.33740234375, "learning_rate": 4.0679377622837256e-06, "loss": 0.1367, "num_tokens": 125602252.0, "reward": 0.3575000464916229, "reward_std": 0.1490996927022934, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.2707233726978302, "step": 831 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1627.0, "completions/max_terminated_length": 1627.0, "completions/mean_length": 567.0875244140625, "completions/mean_terminated_length": 567.0875244140625, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "epoch": 0.3227307990690458, "grad_norm": 0.7563814520835876, "kl": 0.341796875, "learning_rate": 4.065645110982563e-06, "loss": 0.0608, "num_tokens": 125691931.0, "reward": 0.4475000500679016, "reward_std": 0.1988832801580429, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.3475000262260437, "rewards/unicoder_reward_fn/std": 0.38417261838912964, "step": 832 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1593.0, "completions/max_terminated_length": 1593.0, "completions/mean_length": 522.7000122070312, "completions/mean_terminated_length": 522.7000122070312, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.32311869666408066, "grad_norm": 0.6937964558601379, "kl": 0.3671875, "learning_rate": 4.06335038209819e-06, "loss": 0.0536, "num_tokens": 125782939.0, "reward": 0.2900000214576721, "reward_std": 0.14512352645397186, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1900000125169754, "rewards/unicoder_reward_fn/std": 0.27356067299842834, "step": 833 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1543.0, "completions/max_terminated_length": 1543.0, "completions/mean_length": 511.8500061035156, "completions/mean_terminated_length": 511.8500061035156, "completions/min_length": 196.0, "completions/min_terminated_length": 196.0, "epoch": 0.3235065942591156, "grad_norm": 2.1209213733673096, "kl": 0.4345703125, "learning_rate": 4.0610535792542965e-06, "loss": 0.0548, "num_tokens": 125876663.0, "reward": 0.2950000464916229, "reward_std": 0.13952934741973877, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.19500000774860382, "rewards/unicoder_reward_fn/std": 0.27184322476387024, "step": 834 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2905.0, "completions/max_terminated_length": 2905.0, "completions/mean_length": 553.4874877929688, "completions/mean_terminated_length": 553.4874877929688, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.3238944918541505, "grad_norm": 0.7365606427192688, "kl": 0.400390625, "learning_rate": 4.058754706077847e-06, "loss": 0.0276, "num_tokens": 125972006.0, "reward": 0.22875002026557922, "reward_std": 0.06377404183149338, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.23621487617492676, "step": 835 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1523.0, "completions/mean_length": 447.9125061035156, "completions/mean_terminated_length": 401.73419189453125, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.3242823894491854, "grad_norm": 0.9067931175231934, "kl": 0.357421875, "learning_rate": 4.056453766199077e-06, "loss": 0.0227, "num_tokens": 126061695.0, "reward": 0.42750000953674316, "reward_std": 0.16011038422584534, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.32750001549720764, "rewards/unicoder_reward_fn/std": 0.2609500586986542, "step": 836 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1086.0, "completions/max_terminated_length": 1086.0, "completions/mean_length": 410.0, "completions/mean_terminated_length": 410.0, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.32467028704422035, "grad_norm": 1.023411750793457, "kl": 0.359375, "learning_rate": 4.054150763251483e-06, "loss": 0.021, "num_tokens": 126139919.0, "reward": 0.3175000250339508, "reward_std": 0.15799902379512787, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.2199999988079071, "rewards/unicoder_reward_fn/std": 0.20769988000392914, "step": 837 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2344.0, "completions/mean_length": 527.25, "completions/mean_terminated_length": 435.74359130859375, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.32505818463925523, "grad_norm": 3.1971120834350586, "kl": 0.48974609375, "learning_rate": 4.051845700871822e-06, "loss": 0.1049, "num_tokens": 126231851.0, "reward": 0.2775000035762787, "reward_std": 0.1358516961336136, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.18718433380126953, "step": 838 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3747.0, "completions/mean_length": 819.9125366210938, "completions/mean_terminated_length": 505.7671203613281, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.32544608223429017, "grad_norm": 5.6442790031433105, "kl": 0.796875, "learning_rate": 4.049538582700102e-06, "loss": 0.1694, "num_tokens": 126352524.0, "reward": 0.32874998450279236, "reward_std": 0.1861737072467804, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.24000000953674316, "rewards/unicoder_reward_fn/std": 0.23254236578941345, "step": 839 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 978.0, "completions/mean_length": 1521.362548828125, "completions/mean_terminated_length": 482.47369384765625, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "epoch": 0.32583397982932505, "grad_norm": 2.536405563354492, "kl": 2.15625, "learning_rate": 4.047229412379578e-06, "loss": 0.04, "num_tokens": 126524017.0, "reward": 0.26125001907348633, "reward_std": 0.12388323992490768, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.17249999940395355, "rewards/unicoder_reward_fn/std": 0.2859472334384918, "step": 840 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1084.0, "completions/mean_length": 2416.987548828125, "completions/mean_terminated_length": 465.7027282714844, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.32622187742436, "grad_norm": 3.842099666595459, "kl": 6.640625, "learning_rate": 4.044918193556746e-06, "loss": 0.189, "num_tokens": 126763224.0, "reward": 0.24500000476837158, "reward_std": 0.14310400187969208, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.29757457971572876, "step": 841 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 684.0, "completions/mean_length": 2611.91259765625, "completions/mean_terminated_length": 385.78125, "completions/min_length": 196.0, "completions/min_terminated_length": 196.0, "epoch": 0.32660977501939487, "grad_norm": 18.47635841369629, "kl": 6.71875, "learning_rate": 4.042604929881333e-06, "loss": 0.1806, "num_tokens": 127028321.0, "reward": 0.21875, "reward_std": 0.1319543421268463, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.442768931388855, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.23379099369049072, "step": 842 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.65, "completions/max_length": 4096.0, "completions/max_terminated_length": 673.0, "completions/mean_length": 2803.97509765625, "completions/mean_terminated_length": 404.5000305175781, "completions/min_length": 243.0, "completions/min_terminated_length": 243.0, "epoch": 0.3269976726144298, "grad_norm": 28.211965560913086, "kl": 10.484375, "learning_rate": 4.0402896250063e-06, "loss": 0.4319, "num_tokens": 127302503.0, "reward": 0.1862500160932541, "reward_std": 0.17855869233608246, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.25774091482162476, "step": 843 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7625, "completions/max_length": 4096.0, "completions/max_terminated_length": 439.0, "completions/mean_length": 3205.375, "completions/mean_terminated_length": 346.0, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.3273855702094647, "grad_norm": 9.248655319213867, "kl": 14.28125, "learning_rate": 4.03797228258783e-06, "loss": 0.2986, "num_tokens": 127614453.0, "reward": 0.1550000011920929, "reward_std": 0.15684005618095398, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.09000001102685928, "rewards/unicoder_reward_fn/std": 0.20842993259429932, "step": 844 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.75, "completions/max_length": 4096.0, "completions/max_terminated_length": 798.0, "completions/mean_length": 3173.75, "completions/mean_terminated_length": 407.0, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.3277734678044996, "grad_norm": 21.835052490234375, "kl": 19.4375, "learning_rate": 4.0356529062853224e-06, "loss": 0.4055, "num_tokens": 127915905.0, "reward": 0.125, "reward_std": 0.13815723359584808, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.18029510974884033, "step": 845 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 720.0, "completions/mean_length": 3635.824951171875, "completions/mean_terminated_length": 414.6000061035156, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "epoch": 0.3281613653995345, "grad_norm": 28.67731475830078, "kl": 20.25, "learning_rate": 4.033331499761388e-06, "loss": 0.3071, "num_tokens": 128247723.0, "reward": 0.0650000050663948, "reward_std": 0.08037122339010239, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.0325000025331974, "rewards/unicoder_reward_fn/std": 0.11667872965335846, "step": 846 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 603.0, "completions/mean_length": 3630.85009765625, "completions/mean_terminated_length": 374.8000183105469, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.32854926299456944, "grad_norm": 16.95815086364746, "kl": 16.5625, "learning_rate": 4.031008066681846e-06, "loss": 0.3058, "num_tokens": 128587135.0, "reward": 0.058750003576278687, "reward_std": 0.09988243877887726, "rewards/format_reward/mean": 0.26249998807907104, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.0325000025331974, "rewards/unicoder_reward_fn/std": 0.11667872220277786, "step": 847 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 376.0, "completions/mean_length": 4001.72509765625, "completions/mean_terminated_length": 325.0, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "epoch": 0.32893716058960437, "grad_norm": 15.207107543945312, "kl": 5.8046875, "learning_rate": 4.028682610715713e-06, "loss": 0.0756, "num_tokens": 128960409.0, "reward": 0.0062500000931322575, "reward_std": 0.012793734669685364, "rewards/format_reward/mean": 0.0625, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 848 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 952.0, "completions/mean_length": 4056.699951171875, "completions/mean_terminated_length": 952.0, "completions/min_length": 952.0, "completions/min_terminated_length": 952.0, "epoch": 0.32932505818463925, "grad_norm": 5.029456615447998, "kl": 0.7578125, "learning_rate": 4.0263551355352025e-06, "loss": 0.0196, "num_tokens": 129331361.0, "reward": 0.07250000536441803, "reward_std": 0.11090842634439468, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.21839460730552673, "step": 849 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3297129557796742, "grad_norm": 0.3946073055267334, "kl": 0.3212890625, "learning_rate": 4.024025644815715e-06, "loss": 0.0032, "num_tokens": 129706921.0, "reward": 0.005000000353902578, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.044721364974975586, "step": 850 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33010085337470907, "grad_norm": 0.7009381651878357, "kl": 0.1962890625, "learning_rate": 4.021694142235834e-06, "loss": 0.002, "num_tokens": 130086593.0, "reward": 0.04000000283122063, "reward_std": 0.05656854063272476, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.12075711041688919, "step": 851 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.330488750969744, "grad_norm": 0.17626994848251343, "kl": 0.10302734375, "learning_rate": 4.0193606314773226e-06, "loss": 0.001, "num_tokens": 130458721.0, "reward": 0.017500001937150955, "reward_std": 0.04949747398495674, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.1111220121383667, "step": 852 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 135.0, "completions/mean_length": 4046.487548828125, "completions/mean_terminated_length": 135.0, "completions/min_length": 135.0, "completions/min_terminated_length": 135.0, "epoch": 0.3308766485647789, "grad_norm": 0.10242162644863129, "kl": 0.0816650390625, "learning_rate": 4.01702511622511e-06, "loss": 0.0217, "num_tokens": 130833936.0, "reward": 0.011249999515712261, "reward_std": 0.021001702174544334, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.06284399330615997, "step": 853 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 105.0, "completions/mean_length": 4046.112548828125, "completions/mean_terminated_length": 105.0, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.3312645461598138, "grad_norm": 0.06257182359695435, "kl": 0.05572509765625, "learning_rate": 4.014687600167294e-06, "loss": 0.0319, "num_tokens": 131197569.0, "reward": 0.011250000447034836, "reward_std": 0.03181980922818184, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.07044164836406708, "step": 854 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3316524437548487, "grad_norm": 0.020299937576055527, "kl": 0.05059814453125, "learning_rate": 4.012348086995132e-06, "loss": 0.0005, "num_tokens": 131569265.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 855 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33204034134988364, "grad_norm": 0.03129427507519722, "kl": 0.04803466796875, "learning_rate": 4.0100065804030355e-06, "loss": 0.0005, "num_tokens": 131948561.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 856 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3324282389449185, "grad_norm": 0.019153133034706116, "kl": 0.04156494140625, "learning_rate": 4.0076630840885625e-06, "loss": 0.0004, "num_tokens": 132328985.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 857 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33281613653995346, "grad_norm": 0.02189209684729576, "kl": 0.0357666015625, "learning_rate": 4.005317601752413e-06, "loss": 0.0004, "num_tokens": 132707881.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 858 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33320403413498834, "grad_norm": 0.015887359157204628, "kl": 0.032684326171875, "learning_rate": 4.0029701370984236e-06, "loss": 0.0003, "num_tokens": 133086897.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 859 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3335919317300233, "grad_norm": 0.012759176082909107, "kl": 0.030029296875, "learning_rate": 4.000620693833562e-06, "loss": 0.0003, "num_tokens": 133462801.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 860 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3339798293250582, "grad_norm": 0.037155475467443466, "kl": 0.04156494140625, "learning_rate": 3.998269275667919e-06, "loss": 0.0004, "num_tokens": 133852169.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 861 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 211.0, "completions/mean_length": 4047.4375, "completions/mean_terminated_length": 211.0, "completions/min_length": 211.0, "completions/min_terminated_length": 211.0, "epoch": 0.3343677269200931, "grad_norm": 0.07289633899927139, "kl": 0.0386962890625, "learning_rate": 3.995915886314707e-06, "loss": 0.0306, "num_tokens": 134219180.0, "reward": 0.008750000968575478, "reward_std": 0.02474874071776867, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 862 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.334755624515128, "grad_norm": 0.01689443737268448, "kl": 0.035125732421875, "learning_rate": 3.993560529490246e-06, "loss": 0.0004, "num_tokens": 134588964.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 863 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3351435221101629, "grad_norm": 0.02781241573393345, "kl": 0.03338623046875, "learning_rate": 3.991203208913969e-06, "loss": 0.0003, "num_tokens": 134967380.0, "reward": 0.005000000353902578, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.044721364974975586, "step": 864 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33553141970519784, "grad_norm": 0.01512977760285139, "kl": 0.0322265625, "learning_rate": 3.988843928308404e-06, "loss": 0.0003, "num_tokens": 135338108.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 865 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3359193173002327, "grad_norm": 0.04068583622574806, "kl": 0.03118896484375, "learning_rate": 3.9864826913991785e-06, "loss": 0.0003, "num_tokens": 135707244.0, "reward": 0.0162500012665987, "reward_std": 0.04596194252371788, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.11373966187238693, "step": 866 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33630721489526766, "grad_norm": 0.027696726843714714, "kl": 0.03131103515625, "learning_rate": 3.984119501915008e-06, "loss": 0.0003, "num_tokens": 136091012.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 867 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33669511249030254, "grad_norm": 0.011619891971349716, "kl": 0.029052734375, "learning_rate": 3.981754363587693e-06, "loss": 0.0003, "num_tokens": 136465868.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 868 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3370830100853375, "grad_norm": 0.01235915906727314, "kl": 0.0291748046875, "learning_rate": 3.979387280152106e-06, "loss": 0.0003, "num_tokens": 136840412.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 869 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33747090768037236, "grad_norm": 0.04171515256166458, "kl": 0.030120849609375, "learning_rate": 3.977018255346197e-06, "loss": 0.0003, "num_tokens": 137208836.0, "reward": 0.010000000707805157, "reward_std": 0.02828427031636238, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.054191410541534424, "step": 870 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3378588052754073, "grad_norm": 0.04117414727807045, "kl": 0.03753662109375, "learning_rate": 3.974647292910979e-06, "loss": 0.0004, "num_tokens": 137599364.0, "reward": 0.003750000149011612, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 871 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33824670287044223, "grad_norm": 0.04099590331315994, "kl": 0.029266357421875, "learning_rate": 3.972274396590527e-06, "loss": 0.0003, "num_tokens": 137973404.0, "reward": 0.010000000707805157, "reward_std": 0.02828427031636238, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.06284399330615997, "step": 872 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3386346004654771, "grad_norm": 0.02120397798717022, "kl": 0.029144287109375, "learning_rate": 3.969899570131965e-06, "loss": 0.0003, "num_tokens": 138342956.0, "reward": 0.005000000353902578, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472136124968529, "step": 873 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33902249806051205, "grad_norm": 0.021987635642290115, "kl": 0.033416748046875, "learning_rate": 3.967522817285471e-06, "loss": 0.0003, "num_tokens": 138719212.0, "reward": 0.010000000707805157, "reward_std": 0.02828427031636238, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944271504878998, "step": 874 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3394103956555469, "grad_norm": 0.027832189574837685, "kl": 0.0321044921875, "learning_rate": 3.965144141804261e-06, "loss": 0.0003, "num_tokens": 139098420.0, "reward": 0.015000000596046448, "reward_std": 0.03265853598713875, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.07647287845611572, "step": 875 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.33979829325058186, "grad_norm": 0.011107187718153, "kl": 0.03009033203125, "learning_rate": 3.962763547444588e-06, "loss": 0.0003, "num_tokens": 139472692.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 876 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 198.0, "completions/mean_length": 4047.275146484375, "completions/mean_terminated_length": 198.0, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.34018619084561674, "grad_norm": 0.033004503697156906, "kl": 0.0340576171875, "learning_rate": 3.960381037965735e-06, "loss": -0.0021, "num_tokens": 139842266.0, "reward": 0.026250002905726433, "reward_std": 0.0530330054461956, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.11194483935832977, "step": 877 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3405740884406517, "grad_norm": 0.03234517201781273, "kl": 0.033355712890625, "learning_rate": 3.957996617130012e-06, "loss": 0.0003, "num_tokens": 140218314.0, "reward": 0.007500000298023224, "reward_std": 0.02121320366859436, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 878 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34096198603568656, "grad_norm": 0.020171646028757095, "kl": 0.032470703125, "learning_rate": 3.9556102887027425e-06, "loss": 0.0003, "num_tokens": 140604434.0, "reward": 0.005000000353902578, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472136124968529, "step": 879 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3413498836307215, "grad_norm": 0.06355626881122589, "kl": 0.034088134765625, "learning_rate": 3.953222056452266e-06, "loss": 0.0003, "num_tokens": 140970322.0, "reward": 0.032499998807907104, "reward_std": 0.07238814234733582, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.032499998807907104, "rewards/unicoder_reward_fn/std": 0.13666461408138275, "step": 880 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3417377812257564, "grad_norm": 0.03964957222342491, "kl": 0.0333251953125, "learning_rate": 3.950831924149927e-06, "loss": 0.0003, "num_tokens": 141347450.0, "reward": 0.03875000402331352, "reward_std": 0.0530330054461956, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.14957746863365173, "step": 881 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3421256788207913, "grad_norm": 0.03943438455462456, "kl": 0.0372314453125, "learning_rate": 3.948439895570071e-06, "loss": 0.0004, "num_tokens": 141731394.0, "reward": 0.027500001713633537, "reward_std": 0.07778175175189972, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.1377716064453125, "step": 882 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34251357641582625, "grad_norm": 0.027013929560780525, "kl": 0.03753662109375, "learning_rate": 3.946045974490036e-06, "loss": 0.0004, "num_tokens": 142110458.0, "reward": 0.011250000447034836, "reward_std": 0.03181980177760124, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944272994995117, "step": 883 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34290147401086113, "grad_norm": 0.025192009285092354, "kl": 0.03753662109375, "learning_rate": 3.943650164690152e-06, "loss": 0.0004, "num_tokens": 142485170.0, "reward": 0.005000000353902578, "reward_std": 0.009258201345801353, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.03142199665307999, "step": 884 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34328937160589607, "grad_norm": 0.05216633528470993, "kl": 0.0379638671875, "learning_rate": 3.941252469953729e-06, "loss": 0.0004, "num_tokens": 142866082.0, "reward": 0.022499999031424522, "reward_std": 0.05387173965573311, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.022499999031424522, "rewards/unicoder_reward_fn/std": 0.0899718627333641, "step": 885 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34367726920093095, "grad_norm": 0.031144343316555023, "kl": 0.03582763671875, "learning_rate": 3.938852894067054e-06, "loss": 0.0004, "num_tokens": 143227586.0, "reward": 0.02500000037252903, "reward_std": 0.051174938678741455, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.13264591991901398, "step": 886 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3440651667959659, "grad_norm": 0.07033687829971313, "kl": 0.03851318359375, "learning_rate": 3.936451440819386e-06, "loss": 0.0004, "num_tokens": 143607242.0, "reward": 0.044999998062849045, "reward_std": 0.10664913803339005, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.15253137052059174, "step": 887 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34445306439100076, "grad_norm": 0.07134383171796799, "kl": 0.03509521484375, "learning_rate": 3.934048114002946e-06, "loss": 0.0004, "num_tokens": 143974482.0, "reward": 0.052500005811452866, "reward_std": 0.1191888079047203, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.05250000208616257, "rewards/unicoder_reward_fn/std": 0.15829390287399292, "step": 888 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3448409619860357, "grad_norm": 0.09977294504642487, "kl": 0.040283203125, "learning_rate": 3.931642917412916e-06, "loss": 0.0004, "num_tokens": 144342506.0, "reward": 0.05625000596046448, "reward_std": 0.11418004333972931, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.19866643846035004, "step": 889 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3452288595810706, "grad_norm": 0.061664268374443054, "kl": 0.03887939453125, "learning_rate": 3.929235854847429e-06, "loss": 0.0004, "num_tokens": 144713274.0, "reward": 0.03750000149011612, "reward_std": 0.08324871212244034, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.11946611851453781, "step": 890 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3456167571761055, "grad_norm": 0.05339396372437477, "kl": 0.0423583984375, "learning_rate": 3.926826930107566e-06, "loss": 0.0004, "num_tokens": 145096242.0, "reward": 0.06625001132488251, "reward_std": 0.0684824138879776, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.2027500867843628, "step": 891 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3460046547711404, "grad_norm": 0.07034501433372498, "kl": 0.03955078125, "learning_rate": 3.924416146997348e-06, "loss": 0.0004, "num_tokens": 145466602.0, "reward": 0.044999998062849045, "reward_std": 0.10176519304513931, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.11896175146102905, "step": 892 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34639255236617533, "grad_norm": 0.10098449885845184, "kl": 0.04412841796875, "learning_rate": 3.92200350932373e-06, "loss": 0.0004, "num_tokens": 145857802.0, "reward": 0.08749999850988388, "reward_std": 0.15191610157489777, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.19576206803321838, "step": 893 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3467804499612102, "grad_norm": 0.09960783272981644, "kl": 0.04156494140625, "learning_rate": 3.919589020896596e-06, "loss": 0.0004, "num_tokens": 146230306.0, "reward": 0.09750000387430191, "reward_std": 0.1348629593849182, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.20374655723571777, "step": 894 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34716834755624515, "grad_norm": 0.06801395118236542, "kl": 0.045166015625, "learning_rate": 3.917172685528755e-06, "loss": 0.0005, "num_tokens": 146601706.0, "reward": 0.09750000387430191, "reward_std": 0.14677582681179047, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.2700093686580658, "step": 895 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3475562451512801, "grad_norm": 0.09643805772066116, "kl": 0.04071044921875, "learning_rate": 3.914754507035927e-06, "loss": 0.0004, "num_tokens": 146970218.0, "reward": 0.12999999523162842, "reward_std": 0.1196623370051384, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.27207595109939575, "step": 896 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34794414274631497, "grad_norm": 0.08417775481939316, "kl": 0.046142578125, "learning_rate": 3.912334489236749e-06, "loss": 0.0005, "num_tokens": 147333018.0, "reward": 0.14250001311302185, "reward_std": 0.15597233176231384, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.14250001311302185, "rewards/unicoder_reward_fn/std": 0.27038416266441345, "step": 897 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 94.0, "completions/mean_length": 4045.97509765625, "completions/mean_terminated_length": 94.0, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.3483320403413499, "grad_norm": 0.08948733657598495, "kl": 0.04730224609375, "learning_rate": 3.909912635952757e-06, "loss": 0.0005, "num_tokens": 147708424.0, "reward": 0.08250000327825546, "reward_std": 0.10117393732070923, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.15813389420509338, "step": 898 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3487199379363848, "grad_norm": 0.127934530377388, "kl": 0.0462646484375, "learning_rate": 3.9074889510083896e-06, "loss": 0.0005, "num_tokens": 148081888.0, "reward": 0.1587500125169754, "reward_std": 0.16628189384937286, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.23318108916282654, "step": 899 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3491078355314197, "grad_norm": 0.0996997132897377, "kl": 0.04705810546875, "learning_rate": 3.9050634382309755e-06, "loss": 0.0005, "num_tokens": 148452592.0, "reward": 0.17500001192092896, "reward_std": 0.16493408381938934, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.17499998211860657, "rewards/unicoder_reward_fn/std": 0.32666322588920593, "step": 900 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 209.0, "completions/mean_length": 4047.41259765625, "completions/mean_terminated_length": 209.0, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.3494957331264546, "grad_norm": 0.14279212057590485, "kl": 0.04632568359375, "learning_rate": 3.902636101450731e-06, "loss": -0.0014, "num_tokens": 148820081.0, "reward": 0.15000000596046448, "reward_std": 0.12438137829303741, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.25406816601753235, "step": 901 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.34988363072148954, "grad_norm": 0.0950399711728096, "kl": 0.0572509765625, "learning_rate": 3.900206944500751e-06, "loss": 0.0006, "num_tokens": 149191921.0, "reward": 0.07750000059604645, "reward_std": 0.07450627535581589, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.1698659211397171, "step": 902 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3502715283165244, "grad_norm": 0.12426122277975082, "kl": 0.05157470703125, "learning_rate": 3.897775971217006e-06, "loss": 0.0005, "num_tokens": 149570913.0, "reward": 0.20499999821186066, "reward_std": 0.15763233602046967, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.20499999821186066, "rewards/unicoder_reward_fn/std": 0.30352362990379333, "step": 903 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.35065942591155935, "grad_norm": 0.16581794619560242, "kl": 0.05419921875, "learning_rate": 3.895343185438335e-06, "loss": 0.0005, "num_tokens": 149952105.0, "reward": 0.13250000774860382, "reward_std": 0.09304298460483551, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.22713126242160797, "step": 904 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.35104732350659423, "grad_norm": 0.20443221926689148, "kl": 0.05096435546875, "learning_rate": 3.89290859100644e-06, "loss": 0.0005, "num_tokens": 150330129.0, "reward": 0.2874999940395355, "reward_std": 0.203382208943367, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.33199626207351685, "step": 905 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 482.0, "completions/mean_length": 4050.824951171875, "completions/mean_terminated_length": 482.0, "completions/min_length": 482.0, "completions/min_terminated_length": 482.0, "epoch": 0.35143522110162917, "grad_norm": 0.2127758413553238, "kl": 0.06304931640625, "learning_rate": 3.890472191765878e-06, "loss": 0.0306, "num_tokens": 150711251.0, "reward": 0.09375, "reward_std": 0.1257413625717163, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.09250000864267349, "rewards/unicoder_reward_fn/std": 0.17413438856601715, "step": 906 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3518231186966641, "grad_norm": 0.23113968968391418, "kl": 0.064697265625, "learning_rate": 3.888033991564053e-06, "loss": 0.0006, "num_tokens": 151088915.0, "reward": 0.20124998688697815, "reward_std": 0.15547382831573486, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.28104689717292786, "step": 907 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.352211016291699, "grad_norm": 0.4590258300304413, "kl": 0.06890869140625, "learning_rate": 3.885593994251219e-06, "loss": 0.0007, "num_tokens": 151460491.0, "reward": 0.23749999701976776, "reward_std": 0.15293222665786743, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.23749999701976776, "rewards/unicoder_reward_fn/std": 0.2587090730667114, "step": 908 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3525989138867339, "grad_norm": 0.4544951617717743, "kl": 0.0977783203125, "learning_rate": 3.883152203680463e-06, "loss": 0.001, "num_tokens": 151839315.0, "reward": 0.0950000062584877, "reward_std": 0.1281019151210785, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.15903188288211823, "step": 909 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 368.0, "completions/mean_length": 4049.400146484375, "completions/mean_terminated_length": 368.0, "completions/min_length": 368.0, "completions/min_terminated_length": 368.0, "epoch": 0.3529868114817688, "grad_norm": 0.20940515398979187, "kl": 0.0784912109375, "learning_rate": 3.880708623707706e-06, "loss": -0.0175, "num_tokens": 152207547.0, "reward": 0.2912500202655792, "reward_std": 0.13864675164222717, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.2900000214576721, "rewards/unicoder_reward_fn/std": 0.2844493091106415, "step": 910 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.35337470907680374, "grad_norm": 0.8328195214271545, "kl": 0.103515625, "learning_rate": 3.878263258191695e-06, "loss": 0.001, "num_tokens": 152579139.0, "reward": 0.32249999046325684, "reward_std": 0.1780761480331421, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.32249999046325684, "rewards/unicoder_reward_fn/std": 0.3190075755119324, "step": 911 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 298.0, "completions/mean_length": 3951.400146484375, "completions/mean_terminated_length": 240.0, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.3537626066718386, "grad_norm": 1.0281991958618164, "kl": 0.1314697265625, "learning_rate": 3.8758161109939944e-06, "loss": -0.0312, "num_tokens": 152940251.0, "reward": 0.3487500250339508, "reward_std": 0.20431159436702728, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.3450000286102295, "rewards/unicoder_reward_fn/std": 0.34269964694976807, "step": 912 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 345.0, "completions/mean_length": 3904.47509765625, "completions/mean_terminated_length": 265.5, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.35415050426687356, "grad_norm": 2.132458448410034, "kl": 0.287109375, "learning_rate": 3.873367185978985e-06, "loss": -0.0275, "num_tokens": 153309633.0, "reward": 0.1875, "reward_std": 0.1527518779039383, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.23209284245967865, "step": 913 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 578.0, "completions/mean_length": 3580.387451171875, "completions/mean_terminated_length": 346.0909118652344, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.35453840186190844, "grad_norm": 7.502580642700195, "kl": 1.0361328125, "learning_rate": 3.870916487013855e-06, "loss": 0.0497, "num_tokens": 153645608.0, "reward": 0.21124999225139618, "reward_std": 0.1589161604642868, "rewards/format_reward/mean": 0.1875, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.1925000250339508, "rewards/unicoder_reward_fn/std": 0.28718405961990356, "step": 914 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 518.0, "completions/mean_length": 2289.375, "completions/mean_terminated_length": 292.5789489746094, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.3549262994569434, "grad_norm": 12.065217018127441, "kl": 3.96484375, "learning_rate": 3.86846401796859e-06, "loss": 0.0659, "num_tokens": 153880846.0, "reward": 0.11125000566244125, "reward_std": 0.08768069744110107, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.14649923145771027, "step": 915 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 816.0, "completions/mean_length": 1915.7874755859375, "completions/mean_terminated_length": 304.3260803222656, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.35531419705197825, "grad_norm": 16.187471389770508, "kl": 11.703125, "learning_rate": 3.866009782715974e-06, "loss": 0.2034, "num_tokens": 154078565.0, "reward": 0.13875000178813934, "reward_std": 0.09831933677196503, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.15342091023921967, "step": 916 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 807.0, "completions/mean_length": 1814.362548828125, "completions/mean_terminated_length": 293.2708435058594, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.3557020946470132, "grad_norm": 9.715827941894531, "kl": 15.78125, "learning_rate": 3.86355378513158e-06, "loss": 0.3471, "num_tokens": 154269626.0, "reward": 0.09250000864267349, "reward_std": 0.07986476272344589, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.10838282108306885, "step": 917 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 784.0, "completions/mean_length": 2439.762451171875, "completions/mean_terminated_length": 310.3143005371094, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "epoch": 0.3560899922420481, "grad_norm": 11.814654350280762, "kl": 20.46875, "learning_rate": 3.861096029093763e-06, "loss": 0.5459, "num_tokens": 154517879.0, "reward": 0.08875000476837158, "reward_std": 0.07585344463586807, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.017500000074505806, "rewards/unicoder_reward_fn/std": 0.09109044820070267, "step": 918 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 443.0, "completions/mean_length": 3322.25, "completions/mean_terminated_length": 227.25, "completions/min_length": 131.0, "completions/min_terminated_length": 131.0, "epoch": 0.356477889837083, "grad_norm": 45.79011917114258, "kl": 33.75, "learning_rate": 3.858636518483656e-06, "loss": 0.5015, "num_tokens": 154831995.0, "reward": 0.048750001937150955, "reward_std": 0.06528272479772568, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.04974619299173355, "step": 919 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 344.0, "completions/mean_length": 3809.3125, "completions/mean_terminated_length": 273.5, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.35686578743211794, "grad_norm": 45.81547927856445, "kl": 32.84375, "learning_rate": 3.856175257185162e-06, "loss": 0.4637, "num_tokens": 155180332.0, "reward": 0.03375000134110451, "reward_std": 0.05539288744330406, "rewards/format_reward/mean": 0.2874999940395355, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472136124968529, "step": 920 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 296.0, "completions/mean_length": 3611.650146484375, "completions/mean_terminated_length": 221.1999969482422, "completions/min_length": 123.0, "completions/min_terminated_length": 123.0, "epoch": 0.3572536850271528, "grad_norm": 31.66326332092285, "kl": 17.1875, "learning_rate": 3.853712249084947e-06, "loss": 0.2467, "num_tokens": 155520600.0, "reward": 0.04125000163912773, "reward_std": 0.04953137785196304, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 921 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 246.0, "completions/mean_length": 3851.150146484375, "completions/mean_terminated_length": 178.40000915527344, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.35764158262218776, "grad_norm": 13.923694610595703, "kl": 8.5625, "learning_rate": 3.851247498072439e-06, "loss": 0.1597, "num_tokens": 155877444.0, "reward": 0.008750000968575478, "reward_std": 0.017422834411263466, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 922 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 180.0, "completions/mean_length": 4047.050048828125, "completions/mean_terminated_length": 180.0, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.35802948021722264, "grad_norm": 10.318367004394531, "kl": 3.9375, "learning_rate": 3.848781008039814e-06, "loss": 0.0393, "num_tokens": 156254312.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 923 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3584173778122576, "grad_norm": 1.5403759479522705, "kl": 0.77734375, "learning_rate": 3.8463127828819975e-06, "loss": 0.0078, "num_tokens": 156628056.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 924 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.35880527540729246, "grad_norm": 1.548808217048645, "kl": 0.1611328125, "learning_rate": 3.843842826496656e-06, "loss": 0.0016, "num_tokens": 157000024.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 925 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 214.0, "completions/mean_length": 3450.1875, "completions/mean_terminated_length": 121.76923370361328, "completions/min_length": 47.0, "completions/min_terminated_length": 47.0, "epoch": 0.3591931730023274, "grad_norm": 0.48011860251426697, "kl": 0.0921630859375, "learning_rate": 3.841371142784185e-06, "loss": 0.0665, "num_tokens": 157336023.0, "reward": 0.005000000353902578, "reward_std": 0.011700167320668697, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 926 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 237.0, "completions/mean_length": 3006.237548828125, "completions/mean_terminated_length": 133.22727966308594, "completions/min_length": 43.0, "completions/min_terminated_length": 43.0, "epoch": 0.3595810705973623, "grad_norm": 3.370405912399292, "kl": 0.1302490234375, "learning_rate": 3.83889773564771e-06, "loss": 0.2827, "num_tokens": 157624002.0, "reward": 0.016249999403953552, "reward_std": 0.03564689680933952, "rewards/format_reward/mean": 0.16249999403953552, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 927 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 303.0, "completions/mean_length": 2768.712646484375, "completions/mean_terminated_length": 163.29629516601562, "completions/min_length": 62.0, "completions/min_terminated_length": 62.0, "epoch": 0.3599689681923972, "grad_norm": 48.48870849609375, "kl": 0.47119140625, "learning_rate": 3.836422608993081e-06, "loss": 0.4257, "num_tokens": 157881763.0, "reward": 0.026249999180436134, "reward_std": 0.04464869573712349, "rewards/format_reward/mean": 0.26249998807907104, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 928 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1908.0, "completions/mean_length": 3879.5, "completions/mean_terminated_length": 632.0, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.3603568657874321, "grad_norm": 1.6558754444122314, "kl": 0.6123046875, "learning_rate": 3.833945766728859e-06, "loss": 0.0532, "num_tokens": 158231339.0, "reward": 0.003750000149011612, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 929 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2563.0, "completions/mean_length": 3836.97509765625, "completions/mean_terminated_length": 1135.71435546875, "completions/min_length": 101.0, "completions/min_terminated_length": 101.0, "epoch": 0.360744763382467, "grad_norm": 2.0974254608154297, "kl": 0.935546875, "learning_rate": 3.831467212766318e-06, "loss": 0.0352, "num_tokens": 158579337.0, "reward": 0.0162500012665987, "reward_std": 0.030215788632631302, "rewards/format_reward/mean": 0.16249999403953552, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 930 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1089.0, "completions/mean_length": 3861.925048828125, "completions/mean_terminated_length": 350.8000183105469, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.36113266097750196, "grad_norm": 3.64693546295166, "kl": 1.0234375, "learning_rate": 3.828986951019431e-06, "loss": 0.042, "num_tokens": 158928915.0, "reward": 0.020000001415610313, "reward_std": 0.03800376132130623, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 931 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 3039.0, "completions/mean_length": 3778.712646484375, "completions/mean_terminated_length": 923.125, "completions/min_length": 64.0, "completions/min_terminated_length": 64.0, "epoch": 0.36152055857253684, "grad_norm": 13.341136932373047, "kl": 1.359375, "learning_rate": 3.82650498540487e-06, "loss": -0.0035, "num_tokens": 159281316.0, "reward": 0.032499998807907104, "reward_std": 0.04330029711127281, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 932 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1426.0, "completions/mean_length": 3701.574951171875, "completions/mean_terminated_length": 590.0, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.3619084561675718, "grad_norm": 3.1917710304260254, "kl": 0.8232421875, "learning_rate": 3.824021319841999e-06, "loss": 0.0368, "num_tokens": 159627154.0, "reward": 0.036250002682209015, "reward_std": 0.04864630475640297, "rewards/format_reward/mean": 0.36250001192092896, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 933 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 870.0, "completions/mean_length": 3858.962646484375, "completions/mean_terminated_length": 303.3999938964844, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.36229635376260666, "grad_norm": 91.2900161743164, "kl": 1.015625, "learning_rate": 3.821535958252863e-06, "loss": 0.0137, "num_tokens": 159972527.0, "reward": 0.04125000163912773, "reward_std": 0.048099126666784286, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 934 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3421.0, "completions/mean_length": 2759.41259765625, "completions/mean_terminated_length": 531.7667236328125, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.3626842513576416, "grad_norm": 2.6311230659484863, "kl": 0.8134765625, "learning_rate": 3.819048904562189e-06, "loss": -0.0195, "num_tokens": 160232160.0, "reward": 0.08624999970197678, "reward_std": 0.031564973294734955, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 935 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3240.0, "completions/mean_length": 798.0750122070312, "completions/mean_terminated_length": 380.0281677246094, "completions/min_length": 62.0, "completions/min_terminated_length": 62.0, "epoch": 0.3630721489526765, "grad_norm": 389.3387756347656, "kl": 14.755859375, "learning_rate": 3.816560162697373e-06, "loss": 0.2989, "num_tokens": 160340846.0, "reward": 0.09000000357627869, "reward_std": 0.025842303410172462, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 936 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.65, "completions/max_length": 4096.0, "completions/max_terminated_length": 673.0, "completions/mean_length": 2723.675048828125, "completions/mean_terminated_length": 175.07144165039062, "completions/min_length": 52.0, "completions/min_terminated_length": 52.0, "epoch": 0.3634600465477114, "grad_norm": 88.4507064819336, "kl": 3.6953125, "learning_rate": 3.8140697365884786e-06, "loss": 0.3754, "num_tokens": 160608804.0, "reward": 0.02875000238418579, "reward_std": 0.045535337179899216, "rewards/format_reward/mean": 0.2874999940395355, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 937 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 458.0, "completions/mean_length": 3324.400146484375, "completions/mean_terminated_length": 238.0, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.3638479441427463, "grad_norm": 916.7597045898438, "kl": 76.5, "learning_rate": 3.8115776301682284e-06, "loss": 0.9542, "num_tokens": 160920316.0, "reward": 0.06625000387430191, "reward_std": 0.10181555896997452, "rewards/format_reward/mean": 0.23749999701976776, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.042500000447034836, "rewards/unicoder_reward_fn/std": 0.1447608321905136, "step": 938 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 505.0, "completions/mean_length": 1946.9000244140625, "completions/mean_terminated_length": 275.3777770996094, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.36423584173778123, "grad_norm": 103.04866027832031, "kl": 1.412109375, "learning_rate": 3.809083847371999e-06, "loss": 0.4934, "num_tokens": 161119212.0, "reward": 0.29625001549720764, "reward_std": 0.1828012466430664, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.23500001430511475, "rewards/unicoder_reward_fn/std": 0.3394336402416229, "step": 939 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 537.0, "completions/mean_length": 1174.0875244140625, "completions/mean_terminated_length": 263.98358154296875, "completions/min_length": 144.0, "completions/min_terminated_length": 144.0, "epoch": 0.3646237393328161, "grad_norm": 44.3495979309082, "kl": 2.322265625, "learning_rate": 3.8065883921378126e-06, "loss": 0.4128, "num_tokens": 161266243.0, "reward": 0.24000000953674316, "reward_std": 0.17849121987819672, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.22319599986076355, "step": 940 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2905.0, "completions/mean_length": 1558.3875732421875, "completions/mean_terminated_length": 336.5740661621094, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.36501163692785105, "grad_norm": 78.12688446044922, "kl": 2.85546875, "learning_rate": 3.804091268406335e-06, "loss": 0.2685, "num_tokens": 161432858.0, "reward": 0.2175000011920929, "reward_std": 0.19223761558532715, "rewards/format_reward/mean": 0.4000000059604645, "rewards/format_reward/std": 0.49298882484436035, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.2700093984603882, "step": 941 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 3409.0, "completions/mean_length": 1658.8875732421875, "completions/mean_terminated_length": 346.5961608886719, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.365399534522886, "grad_norm": 127.52397918701172, "kl": 4.453125, "learning_rate": 3.801592480120866e-06, "loss": 0.2032, "num_tokens": 161608993.0, "reward": 0.1587500125169754, "reward_std": 0.1566767692565918, "rewards/format_reward/mean": 0.26249998807907104, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.2011202871799469, "step": 942 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 565.0, "completions/mean_length": 2383.6875, "completions/mean_terminated_length": 290.8611145019531, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.36578743211792086, "grad_norm": 300.06585693359375, "kl": 6.8828125, "learning_rate": 3.799092031227331e-06, "loss": 0.1503, "num_tokens": 161851296.0, "reward": 0.11750000715255737, "reward_std": 0.1347566694021225, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.08999999612569809, "rewards/unicoder_reward_fn/std": 0.16504698991775513, "step": 943 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 627.0, "completions/mean_length": 857.8875122070312, "completions/mean_terminated_length": 286.45587158203125, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.3661753297129558, "grad_norm": 78.94376373291016, "kl": 2.7421875, "learning_rate": 3.796589925674282e-06, "loss": 0.2233, "num_tokens": 161961487.0, "reward": 0.29625001549720764, "reward_std": 0.2537446618080139, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.22999998927116394, "rewards/unicoder_reward_fn/std": 0.28657737374305725, "step": 944 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 806.0, "completions/mean_length": 1548.4749755859375, "completions/mean_terminated_length": 321.8888854980469, "completions/min_length": 68.0, "completions/min_terminated_length": 68.0, "epoch": 0.3665632273079907, "grad_norm": 34.21145248413086, "kl": 6.61328125, "learning_rate": 3.7940861674128842e-06, "loss": 0.3366, "num_tokens": 162140493.0, "reward": 0.1287499964237213, "reward_std": 0.12271682173013687, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.1617741882801056, "step": 945 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1608.0, "completions/mean_length": 1140.9749755859375, "completions/mean_terminated_length": 343.5873107910156, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "epoch": 0.3669511249030256, "grad_norm": 44.27782440185547, "kl": 6.1875, "learning_rate": 3.791580760396914e-06, "loss": 0.3008, "num_tokens": 162274675.0, "reward": 0.17625001072883606, "reward_std": 0.1976073980331421, "rewards/format_reward/mean": 0.3375000059604645, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.28139570355415344, "step": 946 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 600.0, "completions/mean_length": 1328.050048828125, "completions/mean_terminated_length": 342.84747314453125, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.3673390224980605, "grad_norm": 42.64070129394531, "kl": 10.046875, "learning_rate": 3.7890737085827517e-06, "loss": 0.3183, "num_tokens": 162431871.0, "reward": 0.10499999672174454, "reward_std": 0.09090269356966019, "rewards/format_reward/mean": 0.30000001192092896, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.1603003442287445, "step": 947 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 824.0, "completions/mean_length": 1167.362548828125, "completions/mean_terminated_length": 317.1128845214844, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.36772692009309543, "grad_norm": 12.452713012695312, "kl": 6.390625, "learning_rate": 3.786565015929373e-06, "loss": 0.2368, "num_tokens": 162572004.0, "reward": 0.17625001072883606, "reward_std": 0.18433301150798798, "rewards/format_reward/mean": 0.3375000059604645, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.14250001311302185, "rewards/unicoder_reward_fn/std": 0.2685050070285797, "step": 948 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 704.0, "completions/mean_length": 1448.5374755859375, "completions/mean_terminated_length": 313.9107360839844, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.3681148176881303, "grad_norm": 19.919057846069336, "kl": 11.0078125, "learning_rate": 3.7840546863983486e-06, "loss": 0.1356, "num_tokens": 162731511.0, "reward": 0.12999999523162842, "reward_std": 0.17872880399227142, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.2110702097415924, "step": 949 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 645.0, "completions/mean_length": 1729.0125732421875, "completions/mean_terminated_length": 308.82000732421875, "completions/min_length": 33.0, "completions/min_terminated_length": 33.0, "epoch": 0.36850271528316525, "grad_norm": 30.69987678527832, "kl": 11.5078125, "learning_rate": 3.7815427239538283e-06, "loss": 0.273, "num_tokens": 162916960.0, "reward": 0.07875000685453415, "reward_std": 0.09578166902065277, "rewards/format_reward/mean": 0.36250001192092896, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.042500000447034836, "rewards/unicoder_reward_fn/std": 0.10881990939378738, "step": 950 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 630.0, "completions/mean_length": 1773.2625732421875, "completions/mean_terminated_length": 303.7755126953125, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.36889061287820013, "grad_norm": 25.459165573120117, "kl": 15.78125, "learning_rate": 3.7790291325625446e-06, "loss": 0.4734, "num_tokens": 163104493.0, "reward": 0.1912500113248825, "reward_std": 0.20621357858181, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2584765553474426, "step": 951 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1091.0, "completions/mean_length": 1895.5374755859375, "completions/mean_terminated_length": 350.5318908691406, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "epoch": 0.36927851047323507, "grad_norm": 44.885005950927734, "kl": 13.8671875, "learning_rate": 3.7765139161938027e-06, "loss": 0.4872, "num_tokens": 163306776.0, "reward": 0.1875000149011612, "reward_std": 0.19785454869270325, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.26428794860839844, "step": 952 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 560.0, "completions/mean_length": 2291.612548828125, "completions/mean_terminated_length": 297.28948974609375, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.36966640806827, "grad_norm": 30.171201705932617, "kl": 20.625, "learning_rate": 3.7739970788194713e-06, "loss": 0.5366, "num_tokens": 163547873.0, "reward": 0.17249999940395355, "reward_std": 0.1965734213590622, "rewards/format_reward/mean": 0.3499999940395355, "rewards/format_reward/std": 0.47997885942459106, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.26355651021003723, "step": 953 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 996.0, "completions/mean_length": 1515.6500244140625, "completions/mean_terminated_length": 273.2592468261719, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.3700543056633049, "grad_norm": 19.300580978393555, "kl": 12.0625, "learning_rate": 3.7714786244139808e-06, "loss": 0.3978, "num_tokens": 163718829.0, "reward": 0.16250000894069672, "reward_std": 0.1860075294971466, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.19858679175376892, "step": 954 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2402.0, "completions/mean_length": 1927.3125, "completions/mean_terminated_length": 324.36956787109375, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.3704422032583398, "grad_norm": 48.92866134643555, "kl": 23.0, "learning_rate": 3.768958556954314e-06, "loss": 0.5515, "num_tokens": 163923566.0, "reward": 0.18125002086162567, "reward_std": 0.23919788002967834, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.2650531232357025, "step": 955 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1828.0, "completions/mean_length": 2002.1500244140625, "completions/mean_terminated_length": 373.6000061035156, "completions/min_length": 144.0, "completions/min_terminated_length": 144.0, "epoch": 0.3708301008533747, "grad_norm": 32.31068801879883, "kl": 31.9375, "learning_rate": 3.7664368804200013e-06, "loss": 0.612, "num_tokens": 164132714.0, "reward": 0.14750000834465027, "reward_std": 0.2144201099872589, "rewards/format_reward/mean": 0.4749999940395355, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.22501757740974426, "step": 956 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1181.0, "completions/mean_length": 1481.5250244140625, "completions/mean_terminated_length": 293.12725830078125, "completions/min_length": 131.0, "completions/min_terminated_length": 131.0, "epoch": 0.37121799844840964, "grad_norm": 42.016441345214844, "kl": 36.8125, "learning_rate": 3.763913598793115e-06, "loss": 0.9058, "num_tokens": 164296580.0, "reward": 0.14874999225139618, "reward_std": 0.14175188541412354, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.16487433016300201, "step": 957 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 958.0, "completions/mean_length": 1925.2000732421875, "completions/mean_terminated_length": 320.6956481933594, "completions/min_length": 93.0, "completions/min_terminated_length": 93.0, "epoch": 0.3716058960434445, "grad_norm": 22.261119842529297, "kl": 48.5, "learning_rate": 3.761388716058261e-06, "loss": 0.8667, "num_tokens": 164497716.0, "reward": 0.13750000298023224, "reward_std": 0.19842331111431122, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.2201121747493744, "step": 958 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1455.0, "completions/mean_length": 1909.4124755859375, "completions/mean_terminated_length": 293.2391357421875, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.37199379363847945, "grad_norm": 73.07460021972656, "kl": 54.8125, "learning_rate": 3.7588622362025734e-06, "loss": 1.079, "num_tokens": 164700221.0, "reward": 0.1587500125169754, "reward_std": 0.17262431979179382, "rewards/format_reward/mean": 0.512499988079071, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.23372331261634827, "step": 959 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 524.0, "completions/mean_length": 1998.4625244140625, "completions/mean_terminated_length": 282.29547119140625, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "epoch": 0.37238169123351433, "grad_norm": 32.228668212890625, "kl": 38.1875, "learning_rate": 3.7563341632157087e-06, "loss": 0.7672, "num_tokens": 164900810.0, "reward": 0.16750000417232513, "reward_std": 0.17301371693611145, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.22466571629047394, "step": 960 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1696.0, "completions/mean_length": 2057.47509765625, "completions/mean_terminated_length": 303.3953552246094, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.37276958882854927, "grad_norm": 58.000885009765625, "kl": 29.40625, "learning_rate": 3.753804501089838e-06, "loss": 0.9045, "num_tokens": 165106992.0, "reward": 0.14874999225139618, "reward_std": 0.19111871719360352, "rewards/format_reward/mean": 0.512499988079071, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.23383159935474396, "step": 961 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7875, "completions/max_length": 4096.0, "completions/max_terminated_length": 311.0, "completions/mean_length": 3269.025146484375, "completions/mean_terminated_length": 204.35293579101562, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.37315748642358415, "grad_norm": 39.93343734741211, "kl": 26.375, "learning_rate": 3.7512732538196457e-06, "loss": 0.4271, "num_tokens": 165413626.0, "reward": 0.10375000536441803, "reward_std": 0.09579696506261826, "rewards/format_reward/mean": 0.23749999701976776, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.17888544499874115, "step": 962 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 367.0, "completions/mean_length": 3901.425048828125, "completions/mean_terminated_length": 204.5, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.3735453840186191, "grad_norm": 22.709333419799805, "kl": 14.890625, "learning_rate": 3.748740425402315e-06, "loss": 0.2448, "num_tokens": 165765316.0, "reward": 0.007500000298023224, "reward_std": 0.02121320180594921, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360680624842644, "step": 963 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 643.0, "completions/mean_length": 3856.41259765625, "completions/mean_terminated_length": 262.6000061035156, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.373933281613654, "grad_norm": 5.744748592376709, "kl": 3.69921875, "learning_rate": 3.7462060198375294e-06, "loss": 0.0625, "num_tokens": 166126565.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 964 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 575.0, "completions/mean_length": 3706.8125, "completions/mean_terminated_length": 204.125, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.3743211792086889, "grad_norm": 8.983959197998047, "kl": 0.77734375, "learning_rate": 3.7436700411274584e-06, "loss": 0.1015, "num_tokens": 166481374.0, "reward": 0.003750000149011612, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 965 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 204.0, "completions/mean_length": 3799.4375, "completions/mean_terminated_length": 141.83334350585938, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.37470907680372384, "grad_norm": 6.588636875152588, "kl": 0.8837890625, "learning_rate": 3.7411324932767605e-06, "loss": 0.0708, "num_tokens": 166833777.0, "reward": 0.003750000149011612, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 966 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 3954.0, "completions/mean_length": 3852.66259765625, "completions/mean_terminated_length": 1662.625, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.3750969743987587, "grad_norm": 19.079755783081055, "kl": 4.8828125, "learning_rate": 3.73859338029257e-06, "loss": 0.1687, "num_tokens": 167184342.0, "reward": 0.010000000707805157, "reward_std": 0.020411193370819092, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 967 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1349.0, "completions/mean_length": 3143.10009765625, "completions/mean_terminated_length": 465.90478515625, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.37548487199379366, "grad_norm": 22.384628295898438, "kl": 11.578125, "learning_rate": 3.736052706184492e-06, "loss": 0.4219, "num_tokens": 167485814.0, "reward": 0.021250000223517418, "reward_std": 0.03274238854646683, "rewards/format_reward/mean": 0.21250000596046448, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 968 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3667.0, "completions/mean_length": 3534.987548828125, "completions/mean_terminated_length": 1103.933349609375, "completions/min_length": 75.0, "completions/min_terminated_length": 75.0, "epoch": 0.37587276958882854, "grad_norm": 24.872840881347656, "kl": 28.40625, "learning_rate": 3.7335104749645994e-06, "loss": 0.4228, "num_tokens": 167819021.0, "reward": 0.0062500000931322575, "reward_std": 0.012246559374034405, "rewards/format_reward/mean": 0.0625, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 969 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3759.0, "completions/mean_length": 3819.800048828125, "completions/mean_terminated_length": 1640.888916015625, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.3762606671838635, "grad_norm": 23.631383895874023, "kl": 19.03125, "learning_rate": 3.730966690647421e-06, "loss": 0.2411, "num_tokens": 168174493.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 970 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1268.0, "completions/mean_length": 3763.800048828125, "completions/mean_terminated_length": 299.4285888671875, "completions/min_length": 50.0, "completions/min_terminated_length": 50.0, "epoch": 0.37664856477889835, "grad_norm": 21.809337615966797, "kl": 9.3984375, "learning_rate": 3.728421357249939e-06, "loss": 0.094, "num_tokens": 168521629.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 971 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1566.0, "completions/mean_length": 1280.237548828125, "completions/mean_terminated_length": 278.0169372558594, "completions/min_length": 30.0, "completions/min_terminated_length": 30.0, "epoch": 0.3770364623739333, "grad_norm": 7.762630939483643, "kl": 4.046875, "learning_rate": 3.725874478791583e-06, "loss": 0.0348, "num_tokens": 168672688.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 972 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1991.0, "completions/mean_length": 316.6750183105469, "completions/mean_terminated_length": 268.83544921875, "completions/min_length": 33.0, "completions/min_terminated_length": 33.0, "epoch": 0.37742435996896817, "grad_norm": 4.116335391998291, "kl": 1.46484375, "learning_rate": 3.723326059294222e-06, "loss": -0.0703, "num_tokens": 168740478.0, "reward": 0.044999998062849045, "reward_std": 0.03202547878026962, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.0603785514831543, "step": 973 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 753.0, "completions/max_terminated_length": 753.0, "completions/mean_length": 198.6750030517578, "completions/mean_terminated_length": 198.6750030517578, "completions/min_length": 21.0, "completions/min_terminated_length": 21.0, "epoch": 0.3778122575640031, "grad_norm": 3.695690393447876, "kl": 1.34375, "learning_rate": 3.720776102782158e-06, "loss": -0.0968, "num_tokens": 168808308.0, "reward": 0.07875000685453415, "reward_std": 0.039473988115787506, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 974 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1096.0, "completions/mean_length": 335.9750061035156, "completions/mean_terminated_length": 288.3797607421875, "completions/min_length": 46.0, "completions/min_terminated_length": 46.0, "epoch": 0.378200155159038, "grad_norm": 5.529967308044434, "kl": 1.412109375, "learning_rate": 3.7182246132821205e-06, "loss": -0.0955, "num_tokens": 168881546.0, "reward": 0.06875000149011612, "reward_std": 0.039180394262075424, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 975 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 1940.0, "completions/mean_length": 545.2125244140625, "completions/mean_terminated_length": 454.16668701171875, "completions/min_length": 50.0, "completions/min_terminated_length": 50.0, "epoch": 0.3785880527540729, "grad_norm": 4.100252151489258, "kl": 1.5703125, "learning_rate": 3.7156715948232584e-06, "loss": -0.0637, "num_tokens": 168970675.0, "reward": 0.07250000536441803, "reward_std": 0.046289440244436264, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 976 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 1692.0, "completions/mean_length": 578.9000244140625, "completions/mean_terminated_length": 488.71795654296875, "completions/min_length": 45.0, "completions/min_terminated_length": 45.0, "epoch": 0.37897595034910786, "grad_norm": 3.4438869953155518, "kl": 1.65625, "learning_rate": 3.713117051437138e-06, "loss": -0.0428, "num_tokens": 169062563.0, "reward": 0.08000000566244125, "reward_std": 0.03648562729358673, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 977 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3630.0, "completions/mean_length": 722.7875366210938, "completions/mean_terminated_length": 680.088623046875, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.37936384794414274, "grad_norm": 3.8930888175964355, "kl": 1.544921875, "learning_rate": 3.710560987157732e-06, "loss": 0.0229, "num_tokens": 169171498.0, "reward": 0.09000000357627869, "reward_std": 0.023400334641337395, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 978 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1798.0, "completions/mean_length": 677.5375366210938, "completions/mean_terminated_length": 544.3506469726562, "completions/min_length": 49.0, "completions/min_terminated_length": 49.0, "epoch": 0.3797517455391777, "grad_norm": 4.389645099639893, "kl": 1.4609375, "learning_rate": 3.7080034060214136e-06, "loss": 0.0901, "num_tokens": 169267949.0, "reward": 0.08625000715255737, "reward_std": 0.031017795205116272, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 979 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3555.0, "completions/mean_length": 934.4625244140625, "completions/mean_terminated_length": 853.3974609375, "completions/min_length": 89.0, "completions/min_terminated_length": 89.0, "epoch": 0.38013964313421256, "grad_norm": 3.980117082595825, "kl": 1.3046875, "learning_rate": 3.7054443120669532e-06, "loss": 0.0602, "num_tokens": 169395506.0, "reward": 0.0949999988079071, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 980 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1634.0, "completions/mean_length": 467.7375183105469, "completions/mean_terminated_length": 421.8101501464844, "completions/min_length": 42.0, "completions/min_terminated_length": 42.0, "epoch": 0.3805275407292475, "grad_norm": 1.1754422187805176, "kl": 1.357421875, "learning_rate": 3.7028837093355096e-06, "loss": 0.0087, "num_tokens": 169481229.0, "reward": 0.11749999970197678, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.060378555208444595, "step": 981 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2683.0, "completions/mean_length": 647.1625366210938, "completions/mean_terminated_length": 512.792236328125, "completions/min_length": 40.0, "completions/min_terminated_length": 40.0, "epoch": 0.3809154383242824, "grad_norm": 13.446356773376465, "kl": 1.73046875, "learning_rate": 3.7003216018706235e-06, "loss": 0.1307, "num_tokens": 169581930.0, "reward": 0.09125000238418579, "reward_std": 0.022306768223643303, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 982 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 1171.0, "completions/mean_length": 464.20001220703125, "completions/mean_terminated_length": 371.0769348144531, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.3813033359193173, "grad_norm": 5.676938533782959, "kl": 2.287109375, "learning_rate": 3.6977579937182136e-06, "loss": 0.0514, "num_tokens": 169664266.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 983 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 980.0, "completions/mean_length": 614.6000366210938, "completions/mean_terminated_length": 382.5066833496094, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.3816912335143522, "grad_norm": 7.676105499267578, "kl": 7.083984375, "learning_rate": 3.695192888926565e-06, "loss": 0.2073, "num_tokens": 169765962.0, "reward": 0.0925000011920929, "reward_std": 0.018771234899759293, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 984 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1007.0, "completions/mean_length": 454.2875061035156, "completions/mean_terminated_length": 408.18988037109375, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.3820791311093871, "grad_norm": 3.8934402465820312, "kl": 3.650390625, "learning_rate": 3.6926262915463297e-06, "loss": 0.1204, "num_tokens": 169849281.0, "reward": 0.0949999988079071, "reward_std": 0.011700167320668697, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 985 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1938.0, "completions/mean_length": 408.8000183105469, "completions/mean_terminated_length": 362.1265869140625, "completions/min_length": 76.0, "completions/min_terminated_length": 76.0, "epoch": 0.382467028704422, "grad_norm": 5.238800048828125, "kl": 6.08984375, "learning_rate": 3.690058205630514e-06, "loss": 0.1164, "num_tokens": 169931641.0, "reward": 0.09375, "reward_std": 0.015235701575875282, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 986 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 781.0, "completions/mean_length": 496.20001220703125, "completions/mean_terminated_length": 306.7368469238281, "completions/min_length": 37.0, "completions/min_terminated_length": 37.0, "epoch": 0.38285492629945694, "grad_norm": 6.710990905761719, "kl": 12.125, "learning_rate": 3.6874886352344784e-06, "loss": 0.3021, "num_tokens": 170025649.0, "reward": 0.0950000062584877, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 987 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1400.0, "completions/max_terminated_length": 1400.0, "completions/mean_length": 380.8125, "completions/mean_terminated_length": 380.8125, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.3832428238944919, "grad_norm": 7.087944030761719, "kl": 2.39453125, "learning_rate": 3.6849175844159236e-06, "loss": 0.0089, "num_tokens": 170100162.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 988 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 4052.0, "completions/mean_length": 1601.5750732421875, "completions/mean_terminated_length": 1362.383544921875, "completions/min_length": 101.0, "completions/min_terminated_length": 101.0, "epoch": 0.38363072148952676, "grad_norm": 0.22042106091976166, "kl": 0.7578125, "learning_rate": 3.68234505723489e-06, "loss": 0.0076, "num_tokens": 170287472.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 989 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 4091.0, "completions/mean_length": 3411.449951171875, "completions/mean_terminated_length": 2207.586181640625, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.3840186190845617, "grad_norm": 1.2900699377059937, "kl": 0.5859375, "learning_rate": 3.6797710577537486e-06, "loss": 0.0093, "num_tokens": 170615724.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 990 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.825, "completions/max_length": 4096.0, "completions/max_terminated_length": 3911.0, "completions/mean_length": 3790.5, "completions/mean_terminated_length": 2350.285888671875, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.3844065166795966, "grad_norm": 0.2800968289375305, "kl": 0.5693359375, "learning_rate": 3.6771955900371975e-06, "loss": 0.0042, "num_tokens": 170963268.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 991 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 3846.0, "completions/mean_length": 4041.525146484375, "completions/mean_terminated_length": 3006.5, "completions/min_length": 1681.0, "completions/min_terminated_length": 1681.0, "epoch": 0.3847944142746315, "grad_norm": 0.11734374612569809, "kl": 0.5634765625, "learning_rate": 3.674618658152251e-06, "loss": 0.0056, "num_tokens": 171336814.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 992 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3895.0, "completions/mean_length": 3967.262451171875, "completions/mean_terminated_length": 2036.2000732421875, "completions/min_length": 764.0, "completions/min_terminated_length": 764.0, "epoch": 0.3851823118696664, "grad_norm": 0.3570474684238434, "kl": 0.55859375, "learning_rate": 3.672040266168236e-06, "loss": -0.02, "num_tokens": 171704995.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 993 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3870.0, "completions/mean_length": 4093.175048828125, "completions/mean_terminated_length": 3870.0, "completions/min_length": 3870.0, "completions/min_terminated_length": 3870.0, "epoch": 0.38557020946470133, "grad_norm": 0.09672918170690536, "kl": 0.5390625, "learning_rate": 3.669460418156786e-06, "loss": 0.0054, "num_tokens": 172077425.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 994 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3859581070597362, "grad_norm": 13.484615325927734, "kl": 1.0927734375, "learning_rate": 3.6668791181918316e-06, "loss": 0.0109, "num_tokens": 172450241.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 995 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.38634600465477115, "grad_norm": 0.19731418788433075, "kl": 0.4716796875, "learning_rate": 3.6642963703495997e-06, "loss": 0.0047, "num_tokens": 172833145.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 996 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.386733902249806, "grad_norm": 0.08544236421585083, "kl": 0.462890625, "learning_rate": 3.6617121787086006e-06, "loss": 0.0046, "num_tokens": 173212921.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 997 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1045.0, "completions/mean_length": 4057.862548828125, "completions/mean_terminated_length": 1045.0, "completions/min_length": 1045.0, "completions/min_terminated_length": 1045.0, "epoch": 0.38712179984484096, "grad_norm": 0.09712952375411987, "kl": 0.45849609375, "learning_rate": 3.659126547349626e-06, "loss": 0.0046, "num_tokens": 173584078.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 998 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3483.0, "completions/mean_length": 4088.337646484375, "completions/mean_terminated_length": 3483.0, "completions/min_length": 3483.0, "completions/min_terminated_length": 3483.0, "epoch": 0.3875096974398759, "grad_norm": 0.2137012481689453, "kl": 0.44287109375, "learning_rate": 3.6565394803557413e-06, "loss": 0.0044, "num_tokens": 173957353.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 999 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3878975950349108, "grad_norm": 0.12327276915311813, "kl": 0.43115234375, "learning_rate": 3.653950981812277e-06, "loss": 0.0043, "num_tokens": 174330889.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1000 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.3882854926299457, "grad_norm": 0.15808407962322235, "kl": 0.40380859375, "learning_rate": 3.651361055806826e-06, "loss": 0.004, "num_tokens": 174711497.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1001 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 187.0, "completions/mean_length": 4047.137451171875, "completions/mean_terminated_length": 187.0, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.3886733902249806, "grad_norm": 0.16617237031459808, "kl": 0.3662109375, "learning_rate": 3.6487697064292366e-06, "loss": 0.0037, "num_tokens": 175086500.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1002 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 130.0, "completions/mean_length": 4046.425048828125, "completions/mean_terminated_length": 130.0, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.38906128782001553, "grad_norm": 0.25548940896987915, "kl": 0.296875, "learning_rate": 3.6461769377716017e-06, "loss": 0.003, "num_tokens": 175468254.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1003 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 509.0, "completions/mean_length": 4051.16259765625, "completions/mean_terminated_length": 509.0, "completions/min_length": 509.0, "completions/min_terminated_length": 509.0, "epoch": 0.3894491854150504, "grad_norm": 0.19577419757843018, "kl": 0.229248046875, "learning_rate": 3.6435827539282564e-06, "loss": 0.0023, "num_tokens": 175847795.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1004 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3222.0, "completions/mean_length": 4021.9375, "completions/mean_terminated_length": 2121.0, "completions/min_length": 1489.0, "completions/min_terminated_length": 1489.0, "epoch": 0.38983708301008535, "grad_norm": 0.4053300619125366, "kl": 0.183349609375, "learning_rate": 3.6409871589957725e-06, "loss": 0.0027, "num_tokens": 176222390.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1005 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3894.0, "completions/mean_length": 4006.800048828125, "completions/mean_terminated_length": 2668.800048828125, "completions/min_length": 859.0, "completions/min_terminated_length": 859.0, "epoch": 0.39022498060512023, "grad_norm": 1.4361892938613892, "kl": 0.138916015625, "learning_rate": 3.638390157072948e-06, "loss": 0.0004, "num_tokens": 176592062.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1006 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 4040.0, "completions/mean_length": 3695.525146484375, "completions/mean_terminated_length": 2093.625, "completions/min_length": 110.0, "completions/min_terminated_length": 110.0, "epoch": 0.39061287820015517, "grad_norm": 0.4187167286872864, "kl": 0.142578125, "learning_rate": 3.6357917522608033e-06, "loss": -0.0855, "num_tokens": 176933696.0, "reward": 0.0949999988079071, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1007 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3656.0, "completions/mean_length": 3891.625, "completions/mean_terminated_length": 2609.636474609375, "completions/min_length": 242.0, "completions/min_terminated_length": 242.0, "epoch": 0.39100077579519005, "grad_norm": 0.04140333831310272, "kl": 0.140869140625, "learning_rate": 3.6331919486625744e-06, "loss": 0.0014, "num_tokens": 177295602.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1008 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 4065.0, "completions/mean_length": 3832.512451171875, "completions/mean_terminated_length": 3092.238037109375, "completions/min_length": 1396.0, "completions/min_terminated_length": 1396.0, "epoch": 0.391388673390225, "grad_norm": 0.057090021669864655, "kl": 0.1455078125, "learning_rate": 3.630590750383706e-06, "loss": 0.0015, "num_tokens": 177651163.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1009 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 3916.0, "completions/mean_length": 3703.824951171875, "completions/mean_terminated_length": 2889.307861328125, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.39177657098525986, "grad_norm": 0.09326933324337006, "kl": 0.159423828125, "learning_rate": 3.627988161531845e-06, "loss": 0.0016, "num_tokens": 177995261.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1010 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 4060.0, "completions/mean_length": 3850.03759765625, "completions/mean_terminated_length": 3201.591064453125, "completions/min_length": 1105.0, "completions/min_terminated_length": 1105.0, "epoch": 0.3921644685802948, "grad_norm": 0.07064224779605865, "kl": 0.15087890625, "learning_rate": 3.6253841862168354e-06, "loss": 0.0015, "num_tokens": 178344888.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1011 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 4037.0, "completions/mean_length": 3449.425048828125, "completions/mean_terminated_length": 2371.800048828125, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.39255236617532974, "grad_norm": 0.513639509677887, "kl": 0.151123046875, "learning_rate": 3.622778828550709e-06, "loss": 0.0039, "num_tokens": 178662930.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1012 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.775, "completions/max_length": 4096.0, "completions/max_terminated_length": 3930.0, "completions/mean_length": 3867.85009765625, "completions/mean_terminated_length": 3082.0, "completions/min_length": 469.0, "completions/min_terminated_length": 469.0, "epoch": 0.3929402637703646, "grad_norm": 0.02624860592186451, "kl": 0.127197265625, "learning_rate": 3.620172092647681e-06, "loss": 0.0013, "num_tokens": 179020398.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1013 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 4094.0, "completions/mean_length": 3659.962646484375, "completions/mean_terminated_length": 2642.541748046875, "completions/min_length": 200.0, "completions/min_terminated_length": 200.0, "epoch": 0.39332816136539955, "grad_norm": 0.0793406292796135, "kl": 0.12109375, "learning_rate": 3.6175639826241425e-06, "loss": -0.0171, "num_tokens": 179366771.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1014 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.825, "completions/max_length": 4096.0, "completions/max_terminated_length": 4023.0, "completions/mean_length": 3866.91259765625, "completions/mean_terminated_length": 2786.9287109375, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.39371605896043443, "grad_norm": 0.0768905058503151, "kl": 0.11279296875, "learning_rate": 3.614954502598657e-06, "loss": -0.028, "num_tokens": 179719020.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1015 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3897.0, "completions/mean_length": 3906.949951171875, "completions/mean_terminated_length": 2415.5556640625, "completions/min_length": 280.0, "completions/min_terminated_length": 280.0, "epoch": 0.39410395655546937, "grad_norm": 0.916217029094696, "kl": 0.10693359375, "learning_rate": 3.6123436566919485e-06, "loss": 0.0019, "num_tokens": 180080736.0, "reward": 0.09749999642372131, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1016 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 4095.0, "completions/mean_length": 3793.5625, "completions/mean_terminated_length": 2234.84619140625, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.39449185415050425, "grad_norm": 0.07840313017368317, "kl": 0.119140625, "learning_rate": 3.6097314490268996e-06, "loss": -0.0021, "num_tokens": 180428709.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1017 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3857.0, "completions/mean_length": 3605.574951171875, "completions/mean_terminated_length": 2227.71435546875, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.3948797517455392, "grad_norm": 0.2829349935054779, "kl": 0.128662109375, "learning_rate": 3.607117883728542e-06, "loss": -0.0664, "num_tokens": 180762307.0, "reward": 0.0962500050663948, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1018 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4007.0, "completions/mean_length": 3523.4375, "completions/mean_terminated_length": 2618.419189453125, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.39526764934057407, "grad_norm": 0.10593443363904953, "kl": 0.1376953125, "learning_rate": 3.6045029649240534e-06, "loss": -0.0381, "num_tokens": 181084974.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1019 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 4044.0, "completions/mean_length": 3029.987548828125, "completions/mean_terminated_length": 1963.9749755859375, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.395655546935609, "grad_norm": 0.06813572347164154, "kl": 0.1875, "learning_rate": 3.601886696742745e-06, "loss": 0.0019, "num_tokens": 181372269.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1020 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 4000.0, "completions/mean_length": 2909.137451171875, "completions/mean_terminated_length": 1722.2750244140625, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.3960434445306439, "grad_norm": 0.0997515469789505, "kl": 0.23681640625, "learning_rate": 3.599269083316064e-06, "loss": 0.0024, "num_tokens": 181649432.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1021 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 4057.0, "completions/mean_length": 2505.3125, "completions/mean_terminated_length": 1498.9591064453125, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.3964313421256788, "grad_norm": 0.4407622814178467, "kl": 0.20068359375, "learning_rate": 3.596650128777579e-06, "loss": -0.0281, "num_tokens": 181902281.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1022 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 4025.0, "completions/mean_length": 2381.962646484375, "completions/mean_terminated_length": 1353.5399169921875, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.39681923972071376, "grad_norm": 0.1622203141450882, "kl": 0.229248046875, "learning_rate": 3.594029837262975e-06, "loss": -0.0619, "num_tokens": 182140766.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1023 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3888.0, "completions/mean_length": 2224.712646484375, "completions/mean_terminated_length": 1641.8524169921875, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.39720713731574864, "grad_norm": 0.042107757180929184, "kl": 0.213623046875, "learning_rate": 3.5914082129100507e-06, "loss": 0.0021, "num_tokens": 182368895.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1024 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3937.0, "completions/mean_length": 2351.800048828125, "completions/mean_terminated_length": 1558.9818115234375, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.3975950349107836, "grad_norm": 0.044583022594451904, "kl": 0.21484375, "learning_rate": 3.5887852598587104e-06, "loss": 0.0021, "num_tokens": 182602183.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1025 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4063.0, "completions/mean_length": 2530.025146484375, "completions/mean_terminated_length": 1898.140380859375, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.39798293250581845, "grad_norm": 0.116180919110775, "kl": 0.187255859375, "learning_rate": 3.586160982250952e-06, "loss": -0.0035, "num_tokens": 182842257.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1026 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 4054.0, "completions/mean_length": 2723.212646484375, "completions/mean_terminated_length": 1854.7142333984375, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.3983708301008534, "grad_norm": 0.12252399325370789, "kl": 0.17529296875, "learning_rate": 3.5835353842308712e-06, "loss": -0.0093, "num_tokens": 183107978.0, "reward": 0.11874999850988388, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.0603785514831543, "step": 1027 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 4024.0, "completions/mean_length": 3556.125, "completions/mean_terminated_length": 2039.3333740234375, "completions/min_length": 422.0, "completions/min_terminated_length": 422.0, "epoch": 0.39875872769588827, "grad_norm": 0.03329033404588699, "kl": 0.13671875, "learning_rate": 3.5809084699446438e-06, "loss": 0.0014, "num_tokens": 183441156.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1028 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2752.0, "completions/mean_length": 3230.087646484375, "completions/mean_terminated_length": 1084.1304931640625, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.3991466252909232, "grad_norm": 0.25008830428123474, "kl": 0.130126953125, "learning_rate": 3.578280243540526e-06, "loss": 0.0076, "num_tokens": 183747963.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1029 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 3623.0, "completions/mean_length": 3557.5, "completions/mean_terminated_length": 1403.5, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.3995345228859581, "grad_norm": 0.6503325700759888, "kl": 0.1273193359375, "learning_rate": 3.5756507091688464e-06, "loss": 0.0042, "num_tokens": 184080387.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1030 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 4000.0, "completions/mean_length": 3248.012451171875, "completions/mean_terminated_length": 1269.375, "completions/min_length": 239.0, "completions/min_terminated_length": 239.0, "epoch": 0.399922420480993, "grad_norm": 0.039302993565797806, "kl": 0.167724609375, "learning_rate": 3.5730198709819985e-06, "loss": 0.0017, "num_tokens": 184390620.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1031 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3151.0, "completions/mean_length": 3402.3125, "completions/mean_terminated_length": 1453.3809814453125, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "epoch": 0.4003103180760279, "grad_norm": 0.052705299109220505, "kl": 0.166259765625, "learning_rate": 3.570387733134436e-06, "loss": 0.0017, "num_tokens": 184716541.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1032 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4028.0, "completions/mean_length": 2982.762451171875, "completions/mean_terminated_length": 1397.242431640625, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.40069821567106284, "grad_norm": 0.13681885600090027, "kl": 0.2041015625, "learning_rate": 3.567754299782662e-06, "loss": 0.002, "num_tokens": 184993194.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1033 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 3874.0, "completions/mean_length": 2694.824951171875, "completions/mean_terminated_length": 1427.09521484375, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.4010861132660978, "grad_norm": 0.11345701664686203, "kl": 0.237548828125, "learning_rate": 3.565119575085228e-06, "loss": -0.0252, "num_tokens": 185246980.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1034 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 3873.0, "completions/mean_length": 2752.16259765625, "completions/mean_terminated_length": 1109.6944580078125, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.40147401086113266, "grad_norm": 0.040714651346206665, "kl": 0.227783203125, "learning_rate": 3.562483563202724e-06, "loss": 0.0023, "num_tokens": 185510041.0, "reward": 0.12000000476837158, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.060378555208444595, "step": 1035 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3954.0, "completions/mean_length": 2754.699951171875, "completions/mean_terminated_length": 1195.8919677734375, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.4018619084561676, "grad_norm": 0.0915834829211235, "kl": 0.250244140625, "learning_rate": 3.5598462682977743e-06, "loss": 0.0025, "num_tokens": 185772177.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1036 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 4022.0, "completions/mean_length": 2656.862548828125, "completions/mean_terminated_length": 984.3513793945312, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.4022498060512025, "grad_norm": 0.026184018701314926, "kl": 0.240966796875, "learning_rate": 3.5572076945350247e-06, "loss": 0.0024, "num_tokens": 186034694.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1037 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 4045.0, "completions/mean_length": 2591.762451171875, "completions/mean_terminated_length": 1297.4185791015625, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.4026377036462374, "grad_norm": 0.038184333592653275, "kl": 0.252197265625, "learning_rate": 3.554567846081144e-06, "loss": 0.0025, "num_tokens": 186297043.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1038 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 3202.0, "completions/mean_length": 2290.074951171875, "completions/mean_terminated_length": 656.1428833007812, "completions/min_length": 93.0, "completions/min_terminated_length": 93.0, "epoch": 0.4030256012412723, "grad_norm": 0.0503123514354229, "kl": 0.30126953125, "learning_rate": 3.5519267271048143e-06, "loss": 0.003, "num_tokens": 186526985.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1039 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3951.0, "completions/mean_length": 2367.97509765625, "completions/mean_terminated_length": 1090.7391357421875, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.4034134988363072, "grad_norm": 0.0963863730430603, "kl": 0.30322265625, "learning_rate": 3.5492843417767223e-06, "loss": -0.0211, "num_tokens": 186768215.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1040 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4042.0, "completions/mean_length": 2404.66259765625, "completions/mean_terminated_length": 1217.1275634765625, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.4038013964313421, "grad_norm": 0.05069610849022865, "kl": 0.3310546875, "learning_rate": 3.5466406942695556e-06, "loss": 0.0033, "num_tokens": 187002820.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1041 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 3244.0, "completions/mean_length": 2662.16259765625, "completions/mean_terminated_length": 1077.394775390625, "completions/min_length": 144.0, "completions/min_terminated_length": 144.0, "epoch": 0.40418929402637704, "grad_norm": 0.13604938983917236, "kl": 0.271240234375, "learning_rate": 3.543995788757994e-06, "loss": -0.0174, "num_tokens": 187269225.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1042 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 3342.0, "completions/mean_length": 2706.462646484375, "completions/mean_terminated_length": 826.5, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.4045771916214119, "grad_norm": 0.034338679164648056, "kl": 0.2724609375, "learning_rate": 3.541349629418704e-06, "loss": 0.0027, "num_tokens": 187538854.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1043 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 4094.0, "completions/mean_length": 2866.875, "completions/mean_terminated_length": 1023.1875, "completions/min_length": 258.0, "completions/min_terminated_length": 258.0, "epoch": 0.40496508921644686, "grad_norm": 0.09516379982233047, "kl": 0.28271484375, "learning_rate": 3.5387022204303327e-06, "loss": 0.0054, "num_tokens": 187809364.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1044 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 2918.0, "completions/mean_length": 2703.60009765625, "completions/mean_terminated_length": 1001.7777709960938, "completions/min_length": 231.0, "completions/min_terminated_length": 231.0, "epoch": 0.40535298681148174, "grad_norm": 0.03963429108262062, "kl": 0.30908203125, "learning_rate": 3.536053565973501e-06, "loss": 0.0031, "num_tokens": 188064164.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1045 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3780.0, "completions/mean_length": 2644.47509765625, "completions/mean_terminated_length": 1118.5128173828125, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.4057408844065167, "grad_norm": 0.03482300043106079, "kl": 0.328125, "learning_rate": 3.533403670230795e-06, "loss": 0.0033, "num_tokens": 188324074.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1046 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2854.0, "completions/mean_length": 2398.97509765625, "completions/mean_terminated_length": 1010.5, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "epoch": 0.4061287820015516, "grad_norm": 0.038678914308547974, "kl": 0.34228515625, "learning_rate": 3.530752537386762e-06, "loss": 0.0034, "num_tokens": 188570896.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1047 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4096.0, "completions/mean_length": 2307.525146484375, "completions/mean_terminated_length": 1051.7872314453125, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.4065166795965865, "grad_norm": 0.11774434894323349, "kl": 0.39599609375, "learning_rate": 3.528100171627902e-06, "loss": 0.004, "num_tokens": 188808114.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1048 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3031.0, "completions/mean_length": 2320.66259765625, "completions/mean_terminated_length": 1197.48974609375, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "epoch": 0.40690457719162143, "grad_norm": 0.03744364157319069, "kl": 0.39208984375, "learning_rate": 3.5254465771426632e-06, "loss": 0.0039, "num_tokens": 189047103.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1049 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 3663.0, "completions/mean_length": 2196.28759765625, "completions/mean_terminated_length": 1173.365478515625, "completions/min_length": 310.0, "completions/min_terminated_length": 310.0, "epoch": 0.4072924747866563, "grad_norm": 0.04245365038514137, "kl": 0.421875, "learning_rate": 3.522791758121434e-06, "loss": 0.0042, "num_tokens": 189271374.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1050 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3902.0, "completions/mean_length": 1845.4124755859375, "completions/mean_terminated_length": 1192.01611328125, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "epoch": 0.40768037238169125, "grad_norm": 0.04482458159327507, "kl": 0.48486328125, "learning_rate": 3.5201357187565357e-06, "loss": 0.0048, "num_tokens": 189471143.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1051 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 3947.0, "completions/mean_length": 2057.175048828125, "completions/mean_terminated_length": 1283.82763671875, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "epoch": 0.40806826997672613, "grad_norm": 0.1850571632385254, "kl": 0.447265625, "learning_rate": 3.517478463242216e-06, "loss": -0.006, "num_tokens": 189681381.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1052 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3410.0, "completions/mean_length": 1849.0, "completions/mean_terminated_length": 1287.25, "completions/min_length": 277.0, "completions/min_terminated_length": 277.0, "epoch": 0.40845616757176106, "grad_norm": 0.04459524527192116, "kl": 0.47216796875, "learning_rate": 3.514819995774646e-06, "loss": 0.0047, "num_tokens": 189872181.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1053 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3701.0, "completions/mean_length": 1940.7000732421875, "completions/mean_terminated_length": 1314.9676513671875, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.40884406516679594, "grad_norm": 0.936102032661438, "kl": 0.48828125, "learning_rate": 3.512160320551906e-06, "loss": 0.0103, "num_tokens": 190070061.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1054 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 4000.0, "completions/mean_length": 1609.112548828125, "completions/mean_terminated_length": 1212.6522216796875, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.4092319627618309, "grad_norm": 0.04905008524656296, "kl": 0.5380859375, "learning_rate": 3.5094994417739886e-06, "loss": 0.0054, "num_tokens": 190234830.0, "reward": 0.12000000476837158, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.060378555208444595, "step": 1055 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3218.0, "completions/mean_length": 1203.487548828125, "completions/mean_terminated_length": 1051.25, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "epoch": 0.40961986035686576, "grad_norm": 1.0962153673171997, "kl": 0.6240234375, "learning_rate": 3.5068373636427834e-06, "loss": 0.0444, "num_tokens": 190380173.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1056 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 2749.0, "completions/mean_length": 1122.8499755859375, "completions/mean_terminated_length": 966.368408203125, "completions/min_length": 330.0, "completions/min_terminated_length": 330.0, "epoch": 0.4100077579519007, "grad_norm": 10.999507904052734, "kl": 1.6455078125, "learning_rate": 3.5041740903620753e-06, "loss": 0.0071, "num_tokens": 190510657.0, "reward": 0.0950000062584877, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1057 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.775, "completions/max_length": 4096.0, "completions/max_terminated_length": 2910.0, "completions/mean_length": 3394.150146484375, "completions/mean_terminated_length": 976.6666870117188, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "epoch": 0.41039565554693563, "grad_norm": 70.02967834472656, "kl": 28.65625, "learning_rate": 3.501509626137537e-06, "loss": 0.3928, "num_tokens": 190828837.0, "reward": 0.05874999985098839, "reward_std": 0.048947010189294815, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1058 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3352.0, "completions/mean_length": 3099.487548828125, "completions/mean_terminated_length": 907.1599731445312, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.4107835531419705, "grad_norm": 55.242549896240234, "kl": 40.25, "learning_rate": 3.49884397517672e-06, "loss": 0.6324, "num_tokens": 191131852.0, "reward": 0.05750000476837158, "reward_std": 0.048268865793943405, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1059 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1865.0, "completions/mean_length": 1855.1500244140625, "completions/mean_terminated_length": 894.7857666015625, "completions/min_length": 336.0, "completions/min_terminated_length": 336.0, "epoch": 0.41117145073700545, "grad_norm": 15.481045722961426, "kl": 12.296875, "learning_rate": 3.496177141689053e-06, "loss": 0.2102, "num_tokens": 191321432.0, "reward": 0.08125000447034836, "reward_std": 0.02722664549946785, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1060 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3390.0, "completions/mean_length": 1539.3250732421875, "completions/mean_terminated_length": 1439.7142333984375, "completions/min_length": 370.0, "completions/min_terminated_length": 370.0, "epoch": 0.41155934833204033, "grad_norm": 9.414000511169434, "kl": 2.0390625, "learning_rate": 3.4935091298858275e-06, "loss": 0.0596, "num_tokens": 191484650.0, "reward": 0.0962500050663948, "reward_std": 0.008164633996784687, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1061 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3274.0, "completions/mean_length": 1637.550048828125, "completions/mean_terminated_length": 1508.157958984375, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.41194724592707527, "grad_norm": 8.071102142333984, "kl": 2.6162109375, "learning_rate": 3.4908399439802003e-06, "loss": 0.0564, "num_tokens": 191654094.0, "reward": 0.0962500050663948, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1062 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3692.0, "completions/mean_length": 2054.862548828125, "completions/mean_terminated_length": 1621.89404296875, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.41233514352211015, "grad_norm": 16.393205642700195, "kl": 16.234375, "learning_rate": 3.488169588187179e-06, "loss": 0.1876, "num_tokens": 191855403.0, "reward": 0.08624999970197678, "reward_std": 0.025210026651620865, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1063 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4043.0, "completions/mean_length": 2251.612548828125, "completions/mean_terminated_length": 1825.984619140625, "completions/min_length": 428.0, "completions/min_terminated_length": 428.0, "epoch": 0.4127230411171451, "grad_norm": 8.949953079223633, "kl": 10.125, "learning_rate": 3.4854980667236206e-06, "loss": 0.1127, "num_tokens": 192078820.0, "reward": 0.08999999612569809, "reward_std": 0.023400334641337395, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1064 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3496.0, "completions/mean_length": 2253.987548828125, "completions/mean_terminated_length": 1990.8428955078125, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.41311093871217996, "grad_norm": 0.7554440498352051, "kl": 0.5009765625, "learning_rate": 3.4828253838082225e-06, "loss": 0.005, "num_tokens": 192309851.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1065 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4080.0, "completions/mean_length": 2791.487548828125, "completions/mean_terminated_length": 2198.527099609375, "completions/min_length": 434.0, "completions/min_terminated_length": 434.0, "epoch": 0.4134988363072149, "grad_norm": 0.05444490909576416, "kl": 0.3447265625, "learning_rate": 3.480151543661515e-06, "loss": 0.0034, "num_tokens": 192580354.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1066 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3953.0, "completions/mean_length": 2980.925048828125, "completions/mean_terminated_length": 2113.64453125, "completions/min_length": 271.0, "completions/min_terminated_length": 271.0, "epoch": 0.4138867339022498, "grad_norm": 0.0710318386554718, "kl": 0.3291015625, "learning_rate": 3.477476550505858e-06, "loss": 0.0033, "num_tokens": 192864556.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1067 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3913.0, "completions/mean_length": 3212.762451171875, "completions/mean_terminated_length": 2186.29736328125, "completions/min_length": 460.0, "completions/min_terminated_length": 460.0, "epoch": 0.4142746314972847, "grad_norm": 0.0810510516166687, "kl": 0.31396484375, "learning_rate": 3.4748004085654288e-06, "loss": 0.0031, "num_tokens": 193165769.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1068 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3955.0, "completions/mean_length": 3423.5625, "completions/mean_terminated_length": 2559.0, "completions/min_length": 547.0, "completions/min_terminated_length": 547.0, "epoch": 0.41466252909231965, "grad_norm": 0.1084289699792862, "kl": 0.30517578125, "learning_rate": 3.4721231220662214e-06, "loss": -0.0147, "num_tokens": 193493902.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1069 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 3822.0, "completions/mean_length": 3293.275146484375, "completions/mean_terminated_length": 2207.2353515625, "completions/min_length": 419.0, "completions/min_terminated_length": 419.0, "epoch": 0.41505042668735453, "grad_norm": 0.11352741718292236, "kl": 0.296875, "learning_rate": 3.4694446952360382e-06, "loss": -0.0555, "num_tokens": 193801932.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1070 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 4095.0, "completions/mean_length": 3116.824951171875, "completions/mean_terminated_length": 1857.8857421875, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.41543832428238947, "grad_norm": 0.072853222489357, "kl": 0.2958984375, "learning_rate": 3.4667651323044793e-06, "loss": -0.0283, "num_tokens": 194090982.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1071 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3661.0, "completions/mean_length": 2796.6875, "completions/mean_terminated_length": 2057.86279296875, "completions/min_length": 713.0, "completions/min_terminated_length": 713.0, "epoch": 0.41582622187742435, "grad_norm": 0.04961065575480461, "kl": 0.30908203125, "learning_rate": 3.4640844375029404e-06, "loss": 0.0031, "num_tokens": 194361373.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1072 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 4071.0, "completions/mean_length": 2618.025146484375, "completions/mean_terminated_length": 2057.413818359375, "completions/min_length": 380.0, "completions/min_terminated_length": 380.0, "epoch": 0.4162141194724593, "grad_norm": 0.14512306451797485, "kl": 0.32861328125, "learning_rate": 3.4614026150646048e-06, "loss": 0.0125, "num_tokens": 194618095.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1073 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3973.0, "completions/mean_length": 2021.487548828125, "completions/mean_terminated_length": 1822.5616455078125, "completions/min_length": 382.0, "completions/min_terminated_length": 382.0, "epoch": 0.41660201706749417, "grad_norm": 0.19230450689792633, "kl": 0.3623046875, "learning_rate": 3.4587196692244363e-06, "loss": -0.0655, "num_tokens": 194830958.0, "reward": 0.0962500050663948, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1074 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4092.0, "completions/mean_length": 2144.72509765625, "completions/mean_terminated_length": 1865.971435546875, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.4169899146625291, "grad_norm": 0.06131698563694954, "kl": 0.35302734375, "learning_rate": 3.4560356042191723e-06, "loss": 0.0035, "num_tokens": 195052208.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1075 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3820.0, "completions/mean_length": 2165.462646484375, "completions/mean_terminated_length": 1920.7464599609375, "completions/min_length": 301.0, "completions/min_terminated_length": 301.0, "epoch": 0.417377812257564, "grad_norm": 0.14239546656608582, "kl": 0.34521484375, "learning_rate": 3.453350424287318e-06, "loss": -0.0495, "num_tokens": 195270509.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1076 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3867.0, "completions/mean_length": 1828.75, "completions/mean_terminated_length": 1644.9189453125, "completions/min_length": 294.0, "completions/min_terminated_length": 294.0, "epoch": 0.4177657098525989, "grad_norm": 0.04157046973705292, "kl": 0.37109375, "learning_rate": 3.450664133669139e-06, "loss": 0.0037, "num_tokens": 195469233.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1077 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3841.0, "completions/mean_length": 1810.425048828125, "completions/mean_terminated_length": 1690.131591796875, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.4181536074476338, "grad_norm": 0.20115560293197632, "kl": 0.38720703125, "learning_rate": 3.4479767366066553e-06, "loss": -0.0911, "num_tokens": 195663851.0, "reward": 0.0949999988079071, "reward_std": 0.011700167320668697, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1078 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3903.0, "completions/mean_length": 1835.4625244140625, "completions/mean_terminated_length": 1652.1756591796875, "completions/min_length": 289.0, "completions/min_terminated_length": 289.0, "epoch": 0.41854150504266874, "grad_norm": 0.14852842688560486, "kl": 0.375, "learning_rate": 3.445288237343632e-06, "loss": -0.0289, "num_tokens": 195862536.0, "reward": 0.11874999850988388, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.0603785514831543, "step": 1079 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3872.0, "completions/mean_length": 1410.7750244140625, "completions/mean_terminated_length": 1341.923095703125, "completions/min_length": 350.0, "completions/min_terminated_length": 350.0, "epoch": 0.4189294026377037, "grad_norm": 0.4013930559158325, "kl": 0.43115234375, "learning_rate": 3.4425986401255786e-06, "loss": -0.0009, "num_tokens": 196024750.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1080 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3663.0, "completions/max_terminated_length": 3663.0, "completions/mean_length": 1516.2625732421875, "completions/mean_terminated_length": 1516.2625732421875, "completions/min_length": 201.0, "completions/min_terminated_length": 201.0, "epoch": 0.41931730023273855, "grad_norm": 0.2762854993343353, "kl": 0.4208984375, "learning_rate": 3.439907949199735e-06, "loss": -0.0308, "num_tokens": 196191643.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1081 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3878.0, "completions/max_terminated_length": 3878.0, "completions/mean_length": 1361.237548828125, "completions/mean_terminated_length": 1361.237548828125, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "epoch": 0.4197051978277735, "grad_norm": 0.1624692678451538, "kl": 0.42333984375, "learning_rate": 3.4372161688150694e-06, "loss": -0.0082, "num_tokens": 196350214.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1082 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3684.0, "completions/mean_length": 1385.550048828125, "completions/mean_terminated_length": 1279.947998046875, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "epoch": 0.42009309542280837, "grad_norm": 0.17831285297870636, "kl": 0.4287109375, "learning_rate": 3.4345233032222706e-06, "loss": -0.0232, "num_tokens": 196508634.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1083 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3799.0, "completions/max_terminated_length": 3799.0, "completions/mean_length": 1269.612548828125, "completions/mean_terminated_length": 1269.612548828125, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.4204809930178433, "grad_norm": 0.19580578804016113, "kl": 0.45556640625, "learning_rate": 3.431829356673742e-06, "loss": -0.0177, "num_tokens": 196657163.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1084 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3627.0, "completions/mean_length": 1286.6624755859375, "completions/mean_terminated_length": 1214.628173828125, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.4208688906128782, "grad_norm": 0.06260217726230621, "kl": 0.44677734375, "learning_rate": 3.4291343334235917e-06, "loss": 0.0045, "num_tokens": 196814080.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1085 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3388.0, "completions/max_terminated_length": 3388.0, "completions/mean_length": 1126.4749755859375, "completions/mean_terminated_length": 1126.4749755859375, "completions/min_length": 311.0, "completions/min_terminated_length": 311.0, "epoch": 0.4212567882079131, "grad_norm": 0.12675805389881134, "kl": 0.47412109375, "learning_rate": 3.426438237727629e-06, "loss": -0.0171, "num_tokens": 196955638.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1086 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3029.0, "completions/max_terminated_length": 3029.0, "completions/mean_length": 1162.5125732421875, "completions/mean_terminated_length": 1162.5125732421875, "completions/min_length": 319.0, "completions/min_terminated_length": 319.0, "epoch": 0.421644685802948, "grad_norm": 0.04607558250427246, "kl": 0.47607421875, "learning_rate": 3.423741073843358e-06, "loss": 0.0048, "num_tokens": 197097775.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1087 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2947.0, "completions/max_terminated_length": 2947.0, "completions/mean_length": 977.0, "completions/mean_terminated_length": 977.0, "completions/min_length": 287.0, "completions/min_terminated_length": 287.0, "epoch": 0.42203258339798294, "grad_norm": 0.05711809918284416, "kl": 0.5439453125, "learning_rate": 3.4210428460299677e-06, "loss": 0.0054, "num_tokens": 197229063.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1088 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2188.0, "completions/max_terminated_length": 2188.0, "completions/mean_length": 916.2000122070312, "completions/mean_terminated_length": 916.2000122070312, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.4224204809930178, "grad_norm": 0.05438077822327614, "kl": 0.5556640625, "learning_rate": 3.4183435585483287e-06, "loss": 0.0056, "num_tokens": 197351719.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1089 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2695.0, "completions/max_terminated_length": 2695.0, "completions/mean_length": 957.0125122070312, "completions/mean_terminated_length": 957.0125122070312, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.42280837858805276, "grad_norm": 0.05709853395819664, "kl": 0.51806640625, "learning_rate": 3.415643215660985e-06, "loss": 0.0052, "num_tokens": 197476984.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1090 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2327.0, "completions/max_terminated_length": 2327.0, "completions/mean_length": 1026.0250244140625, "completions/mean_terminated_length": 1026.0250244140625, "completions/min_length": 233.0, "completions/min_terminated_length": 233.0, "epoch": 0.42319627618308764, "grad_norm": 0.22619746625423431, "kl": 0.4892578125, "learning_rate": 3.412941821632145e-06, "loss": -0.014, "num_tokens": 197610282.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1091 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3143.0, "completions/max_terminated_length": 3143.0, "completions/mean_length": 1080.9749755859375, "completions/mean_terminated_length": 1080.9749755859375, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.4235841737781226, "grad_norm": 0.050646036863327026, "kl": 0.515625, "learning_rate": 3.4102393807276798e-06, "loss": 0.0052, "num_tokens": 197737904.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1092 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2548.0, "completions/max_terminated_length": 2548.0, "completions/mean_length": 1122.4375, "completions/mean_terminated_length": 1122.4375, "completions/min_length": 271.0, "completions/min_terminated_length": 271.0, "epoch": 0.4239720713731575, "grad_norm": 0.04718707129359245, "kl": 0.51171875, "learning_rate": 3.4075358972151126e-06, "loss": 0.0051, "num_tokens": 197875019.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1093 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2333.0, "completions/max_terminated_length": 2333.0, "completions/mean_length": 1155.7750244140625, "completions/mean_terminated_length": 1155.7750244140625, "completions/min_length": 310.0, "completions/min_terminated_length": 310.0, "epoch": 0.4243599689681924, "grad_norm": 0.048457518219947815, "kl": 0.517578125, "learning_rate": 3.4048313753636113e-06, "loss": 0.0052, "num_tokens": 198015953.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1094 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2544.0, "completions/max_terminated_length": 2544.0, "completions/mean_length": 1133.3250732421875, "completions/mean_terminated_length": 1133.3250732421875, "completions/min_length": 275.0, "completions/min_terminated_length": 275.0, "epoch": 0.4247478665632273, "grad_norm": 0.6740795373916626, "kl": 0.5517578125, "learning_rate": 3.4021258194439877e-06, "loss": -0.0035, "num_tokens": 198144755.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1095 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2503.0, "completions/max_terminated_length": 2503.0, "completions/mean_length": 1183.0374755859375, "completions/mean_terminated_length": 1183.0374755859375, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.4251357641582622, "grad_norm": 0.22858251631259918, "kl": 0.51806640625, "learning_rate": 3.399419233728682e-06, "loss": -0.0059, "num_tokens": 198289654.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1096 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3055.0, "completions/max_terminated_length": 3055.0, "completions/mean_length": 1313.050048828125, "completions/mean_terminated_length": 1313.050048828125, "completions/min_length": 315.0, "completions/min_terminated_length": 315.0, "epoch": 0.42552366175329714, "grad_norm": 0.04499928653240204, "kl": 0.51123046875, "learning_rate": 3.3967116224917624e-06, "loss": 0.0051, "num_tokens": 198435738.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1097 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3045.0, "completions/max_terminated_length": 3045.0, "completions/mean_length": 1323.2874755859375, "completions/mean_terminated_length": 1323.2874755859375, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.425911559348332, "grad_norm": 0.20956075191497803, "kl": 0.537109375, "learning_rate": 3.3940029900089172e-06, "loss": -0.0224, "num_tokens": 198584057.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1098 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4012.0, "completions/max_terminated_length": 4012.0, "completions/mean_length": 1436.5250244140625, "completions/mean_terminated_length": 1436.5250244140625, "completions/min_length": 338.0, "completions/min_terminated_length": 338.0, "epoch": 0.42629945694336696, "grad_norm": 0.051601920276880264, "kl": 0.5205078125, "learning_rate": 3.391293340557447e-06, "loss": 0.0052, "num_tokens": 198743811.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1099 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3235.0, "completions/mean_length": 1637.0374755859375, "completions/mean_terminated_length": 1605.911376953125, "completions/min_length": 569.0, "completions/min_terminated_length": 569.0, "epoch": 0.42668735453840184, "grad_norm": 0.05540595203638077, "kl": 0.4970703125, "learning_rate": 3.3885826784162575e-06, "loss": 0.005, "num_tokens": 198930790.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1100 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3717.0, "completions/mean_length": 1716.8875732421875, "completions/mean_terminated_length": 1655.8846435546875, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.4270752521334368, "grad_norm": 0.04591596499085426, "kl": 0.4736328125, "learning_rate": 3.3858710078658545e-06, "loss": 0.0047, "num_tokens": 199116493.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1101 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3962.0, "completions/max_terminated_length": 3962.0, "completions/mean_length": 1656.7249755859375, "completions/mean_terminated_length": 1656.7249755859375, "completions/min_length": 426.0, "completions/min_terminated_length": 426.0, "epoch": 0.42746314972847166, "grad_norm": 0.04877853766083717, "kl": 0.4794921875, "learning_rate": 3.383158333188336e-06, "loss": 0.0048, "num_tokens": 199292087.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1102 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3444.0, "completions/mean_length": 1775.5875244140625, "completions/mean_terminated_length": 1716.0897216796875, "completions/min_length": 373.0, "completions/min_terminated_length": 373.0, "epoch": 0.4278510473235066, "grad_norm": 0.22994352877140045, "kl": 0.4619140625, "learning_rate": 3.380444658667386e-06, "loss": 0.0115, "num_tokens": 199479574.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1103 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3978.0, "completions/mean_length": 1815.0, "completions/mean_terminated_length": 1786.1265869140625, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.42823894491854153, "grad_norm": 0.29369881749153137, "kl": 0.46337890625, "learning_rate": 3.3777299885882663e-06, "loss": -0.0397, "num_tokens": 199672558.0, "reward": 0.09749999642372131, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1104 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 4087.0, "completions/mean_length": 1932.237548828125, "completions/mean_terminated_length": 1876.7564697265625, "completions/min_length": 467.0, "completions/min_terminated_length": 467.0, "epoch": 0.4286268425135764, "grad_norm": 0.15992389619350433, "kl": 0.43212890625, "learning_rate": 3.3750143272378124e-06, "loss": -0.0096, "num_tokens": 199877681.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1105 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 4031.0, "completions/mean_length": 2166.862548828125, "completions/mean_terminated_length": 1922.3238525390625, "completions/min_length": 218.0, "completions/min_terminated_length": 218.0, "epoch": 0.42901474010861135, "grad_norm": 0.036413684487342834, "kl": 0.3974609375, "learning_rate": 3.3722976789044225e-06, "loss": 0.004, "num_tokens": 200094062.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1106 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 4010.0, "completions/mean_length": 2609.9375, "completions/mean_terminated_length": 2294.712158203125, "completions/min_length": 556.0, "completions/min_terminated_length": 556.0, "epoch": 0.42940263770364623, "grad_norm": 0.12381068617105484, "kl": 0.35107421875, "learning_rate": 3.3695800478780564e-06, "loss": -0.0017, "num_tokens": 200348865.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1107 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3979.0, "completions/mean_length": 2358.137451171875, "completions/mean_terminated_length": 1923.671875, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.42979053529868116, "grad_norm": 0.4364805519580841, "kl": 0.33984375, "learning_rate": 3.3668614384502247e-06, "loss": -0.019, "num_tokens": 200578772.0, "reward": 0.09375, "reward_std": 0.015235701575875282, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1108 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3794.0, "completions/mean_length": 2380.85009765625, "completions/mean_terminated_length": 1918.0318603515625, "completions/min_length": 612.0, "completions/min_terminated_length": 612.0, "epoch": 0.43017843289371605, "grad_norm": 0.7446850538253784, "kl": 0.333984375, "learning_rate": 3.364141854913982e-06, "loss": 0.0487, "num_tokens": 200816552.0, "reward": 0.09375, "reward_std": 0.0176776684820652, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1109 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3743.0, "completions/mean_length": 2226.712646484375, "completions/mean_terminated_length": 1928.710205078125, "completions/min_length": 286.0, "completions/min_terminated_length": 286.0, "epoch": 0.430566330488751, "grad_norm": 0.3587658405303955, "kl": 0.3564453125, "learning_rate": 3.3614213015639226e-06, "loss": 0.0396, "num_tokens": 201035577.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1110 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3839.0, "completions/max_terminated_length": 3839.0, "completions/mean_length": 1740.862548828125, "completions/mean_terminated_length": 1740.862548828125, "completions/min_length": 579.0, "completions/min_terminated_length": 579.0, "epoch": 0.43095422808378586, "grad_norm": 0.18191875517368317, "kl": 0.427734375, "learning_rate": 3.3586997826961713e-06, "loss": -0.0061, "num_tokens": 201218638.0, "reward": 0.09750000387430191, "reward_std": 0.004629100672900677, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1111 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3255.0, "completions/max_terminated_length": 3255.0, "completions/mean_length": 1526.4375, "completions/mean_terminated_length": 1526.4375, "completions/min_length": 283.0, "completions/min_terminated_length": 283.0, "epoch": 0.4313421256788208, "grad_norm": 0.5339266061782837, "kl": 0.45751953125, "learning_rate": 3.3559773026083774e-06, "loss": 0.0063, "num_tokens": 201392945.0, "reward": 0.0962500050663948, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1112 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3046.0, "completions/max_terminated_length": 3046.0, "completions/mean_length": 1190.0, "completions/mean_terminated_length": 1190.0, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "epoch": 0.4317300232738557, "grad_norm": 0.5046526193618774, "kl": 0.52734375, "learning_rate": 3.3532538655997095e-06, "loss": -0.0307, "num_tokens": 201527657.0, "reward": 0.0950000062584877, "reward_std": 0.011700168251991272, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1113 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3246.0, "completions/max_terminated_length": 3246.0, "completions/mean_length": 1149.2874755859375, "completions/mean_terminated_length": 1149.2874755859375, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.4321179208688906, "grad_norm": 0.8791932463645935, "kl": 0.5126953125, "learning_rate": 3.3505294759708457e-06, "loss": -0.061, "num_tokens": 201676352.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1114 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3576.0, "completions/mean_length": 1915.425048828125, "completions/mean_terminated_length": 1603.914306640625, "completions/min_length": 380.0, "completions/min_terminated_length": 380.0, "epoch": 0.43250581846392555, "grad_norm": 0.30588680505752563, "kl": 0.40234375, "learning_rate": 3.3478041380239684e-06, "loss": -0.0271, "num_tokens": 201867818.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1115 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3713.0, "completions/mean_length": 2245.462646484375, "completions/mean_terminated_length": 1981.0999755859375, "completions/min_length": 400.0, "completions/min_terminated_length": 400.0, "epoch": 0.43289371605896043, "grad_norm": 0.15560652315616608, "kl": 0.33447265625, "learning_rate": 3.34507785606276e-06, "loss": 0.0166, "num_tokens": 202105215.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1116 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 4090.0, "completions/mean_length": 2952.4375, "completions/mean_terminated_length": 2462.33935546875, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.43328161365399537, "grad_norm": 0.04854455962777138, "kl": 0.26416015625, "learning_rate": 3.342350634392391e-06, "loss": 0.0026, "num_tokens": 202388738.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1117 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3919.0, "completions/mean_length": 3269.637451171875, "completions/mean_terminated_length": 2309.270263671875, "completions/min_length": 772.0, "completions/min_terminated_length": 772.0, "epoch": 0.43366951124903025, "grad_norm": 0.15172088146209717, "kl": 0.243896484375, "learning_rate": 3.3396224773195155e-06, "loss": -0.0083, "num_tokens": 202693733.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1118 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.65, "completions/max_length": 4096.0, "completions/max_terminated_length": 4082.0, "completions/mean_length": 3574.712646484375, "completions/mean_terminated_length": 2606.607177734375, "completions/min_length": 392.0, "completions/min_terminated_length": 392.0, "epoch": 0.4340574088440652, "grad_norm": 0.052076712250709534, "kl": 0.2294921875, "learning_rate": 3.336893389152269e-06, "loss": 0.0023, "num_tokens": 203026782.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1119 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3784.0, "completions/mean_length": 3790.262451171875, "completions/mean_terminated_length": 2657.2353515625, "completions/min_length": 1079.0, "completions/min_terminated_length": 1079.0, "epoch": 0.43444530643910007, "grad_norm": 0.05025310441851616, "kl": 0.209716796875, "learning_rate": 3.3341633742002506e-06, "loss": 0.0021, "num_tokens": 203372635.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1120 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 3889.0, "completions/mean_length": 3469.1875, "completions/mean_terminated_length": 1816.681884765625, "completions/min_length": 304.0, "completions/min_terminated_length": 304.0, "epoch": 0.434833204034135, "grad_norm": 0.13782063126564026, "kl": 0.219482421875, "learning_rate": 3.331432436774528e-06, "loss": -0.0087, "num_tokens": 203703354.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1121 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 4049.0, "completions/mean_length": 3708.574951171875, "completions/mean_terminated_length": 2620.09521484375, "completions/min_length": 758.0, "completions/min_terminated_length": 758.0, "epoch": 0.4352211016291699, "grad_norm": 0.05137169361114502, "kl": 0.210693359375, "learning_rate": 3.328700581187625e-06, "loss": 0.0021, "num_tokens": 204041000.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1122 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3984.0, "completions/mean_length": 3676.300048828125, "completions/mean_terminated_length": 1857.60009765625, "completions/min_length": 583.0, "completions/min_terminated_length": 583.0, "epoch": 0.4356089992242048, "grad_norm": 0.05078550800681114, "kl": 0.2138671875, "learning_rate": 3.325967811753514e-06, "loss": 0.0021, "num_tokens": 204388984.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1123 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3772.0, "completions/mean_length": 3689.237548828125, "completions/mean_terminated_length": 2383.31591796875, "completions/min_length": 692.0, "completions/min_terminated_length": 692.0, "epoch": 0.4359968968192397, "grad_norm": 0.07391182333230972, "kl": 0.214111328125, "learning_rate": 3.32323413278761e-06, "loss": -0.0265, "num_tokens": 204735331.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1124 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3882.0, "completions/mean_length": 3533.53759765625, "completions/mean_terminated_length": 2296.119873046875, "completions/min_length": 296.0, "completions/min_terminated_length": 296.0, "epoch": 0.43638479441427463, "grad_norm": 0.060898955911397934, "kl": 0.216796875, "learning_rate": 3.3204995486067664e-06, "loss": 0.0022, "num_tokens": 205064718.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1125 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3660.0, "completions/mean_length": 3527.050048828125, "completions/mean_terminated_length": 2117.04345703125, "completions/min_length": 520.0, "completions/min_terminated_length": 520.0, "epoch": 0.4367726920093095, "grad_norm": 0.04852789267897606, "kl": 0.211669921875, "learning_rate": 3.3177640635292647e-06, "loss": 0.0021, "num_tokens": 205391298.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1126 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.775, "completions/max_length": 4096.0, "completions/max_terminated_length": 3999.0, "completions/mean_length": 3741.75, "completions/mean_terminated_length": 2521.5556640625, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "epoch": 0.43716058960434445, "grad_norm": 0.05163509026169777, "kl": 0.22119140625, "learning_rate": 3.3150276818748084e-06, "loss": 0.0022, "num_tokens": 205736366.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1127 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 3916.0, "completions/mean_length": 3635.074951171875, "completions/mean_terminated_length": 2419.9091796875, "completions/min_length": 951.0, "completions/min_terminated_length": 951.0, "epoch": 0.4375484871993794, "grad_norm": 0.047186169773340225, "kl": 0.231689453125, "learning_rate": 3.312290407964519e-06, "loss": 0.0023, "num_tokens": 206069860.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1128 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4037.0, "completions/mean_length": 3444.5, "completions/mean_terminated_length": 2414.709716796875, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "epoch": 0.43793638479441427, "grad_norm": 0.0450146459043026, "kl": 0.2373046875, "learning_rate": 3.309552246120924e-06, "loss": 0.0024, "num_tokens": 206393692.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1129 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 4041.0, "completions/mean_length": 3269.91259765625, "completions/mean_terminated_length": 2309.864990234375, "completions/min_length": 625.0, "completions/min_terminated_length": 625.0, "epoch": 0.4383242823894492, "grad_norm": 0.06758179515600204, "kl": 0.250244140625, "learning_rate": 3.3068132006679552e-06, "loss": -0.0192, "num_tokens": 206703909.0, "reward": 0.11874999850988388, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.0603785514831543, "step": 1130 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3983.0, "completions/mean_length": 2873.925048828125, "completions/mean_terminated_length": 1923.4222412109375, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "epoch": 0.4387121799844841, "grad_norm": 0.1071464791893959, "kl": 0.2705078125, "learning_rate": 3.304073275930941e-06, "loss": -0.0167, "num_tokens": 206980567.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1131 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4087.0, "completions/mean_length": 3122.074951171875, "completions/mean_terminated_length": 2438.255126953125, "completions/min_length": 821.0, "completions/min_terminated_length": 821.0, "epoch": 0.439100077579519, "grad_norm": 0.25645583868026733, "kl": 0.252685546875, "learning_rate": 3.3013324762365934e-06, "loss": 0.0104, "num_tokens": 207281629.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1132 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 3932.0, "completions/mean_length": 2521.387451171875, "completions/mean_terminated_length": 1996.5167236328125, "completions/min_length": 337.0, "completions/min_terminated_length": 337.0, "epoch": 0.4394879751745539, "grad_norm": 0.0383426770567894, "kl": 0.30224609375, "learning_rate": 3.2985908059130122e-06, "loss": 0.003, "num_tokens": 207531500.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1133 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4025.0, "completions/mean_length": 1874.6375732421875, "completions/mean_terminated_length": 1557.300048828125, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "epoch": 0.43987587276958884, "grad_norm": 0.0429861955344677, "kl": 0.359375, "learning_rate": 3.295848269289667e-06, "loss": 0.0036, "num_tokens": 207721607.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1134 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3876.0, "completions/mean_length": 1463.6375732421875, "completions/mean_terminated_length": 1396.1409912109375, "completions/min_length": 294.0, "completions/min_terminated_length": 294.0, "epoch": 0.4402637703646237, "grad_norm": 0.04598909243941307, "kl": 0.4111328125, "learning_rate": 3.293104870697399e-06, "loss": 0.0041, "num_tokens": 207897642.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1135 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3788.0, "completions/mean_length": 1409.2625732421875, "completions/mean_terminated_length": 1340.371826171875, "completions/min_length": 397.0, "completions/min_terminated_length": 397.0, "epoch": 0.44065166795965865, "grad_norm": 0.04903113469481468, "kl": 0.44580078125, "learning_rate": 3.2903606144684075e-06, "loss": 0.0045, "num_tokens": 208052303.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1136 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3314.0, "completions/mean_length": 1122.5750732421875, "completions/mean_terminated_length": 1084.936767578125, "completions/min_length": 282.0, "completions/min_terminated_length": 282.0, "epoch": 0.44103956555469354, "grad_norm": 0.19562876224517822, "kl": 0.462890625, "learning_rate": 3.2876155049362492e-06, "loss": -0.0043, "num_tokens": 208190277.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1137 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3066.0, "completions/max_terminated_length": 3066.0, "completions/mean_length": 1020.4750366210938, "completions/mean_terminated_length": 1020.4750366210938, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "epoch": 0.44142746314972847, "grad_norm": 0.0467405803501606, "kl": 0.4990234375, "learning_rate": 3.284869546435826e-06, "loss": 0.005, "num_tokens": 208317507.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1138 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2519.0, "completions/max_terminated_length": 2519.0, "completions/mean_length": 810.1500244140625, "completions/mean_terminated_length": 810.1500244140625, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "epoch": 0.4418153607447634, "grad_norm": 0.04640684276819229, "kl": 0.5400390625, "learning_rate": 3.282122743303381e-06, "loss": 0.0054, "num_tokens": 208421287.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1139 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1587.0, "completions/max_terminated_length": 1587.0, "completions/mean_length": 754.4375, "completions/mean_terminated_length": 754.4375, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "epoch": 0.4422032583397983, "grad_norm": 0.11399303376674652, "kl": 0.5048828125, "learning_rate": 3.2793750998764914e-06, "loss": 0.0051, "num_tokens": 208538314.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1140 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1459.0, "completions/mean_length": 783.2250366210938, "completions/mean_terminated_length": 741.2911376953125, "completions/min_length": 309.0, "completions/min_terminated_length": 309.0, "epoch": 0.4425911559348332, "grad_norm": 0.04029253497719765, "kl": 0.50439453125, "learning_rate": 3.2766266204940617e-06, "loss": 0.005, "num_tokens": 208654540.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1141 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1562.0, "completions/max_terminated_length": 1562.0, "completions/mean_length": 735.2250366210938, "completions/mean_terminated_length": 735.2250366210938, "completions/min_length": 284.0, "completions/min_terminated_length": 284.0, "epoch": 0.4429790535298681, "grad_norm": 0.03984124958515167, "kl": 0.529296875, "learning_rate": 3.273877309496315e-06, "loss": 0.0053, "num_tokens": 208763006.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1142 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1715.0, "completions/mean_length": 732.8125, "completions/mean_terminated_length": 690.2405395507812, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.44336695112490304, "grad_norm": 0.046227000653743744, "kl": 0.5439453125, "learning_rate": 3.27112717122479e-06, "loss": 0.0054, "num_tokens": 208874215.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1143 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1709.0, "completions/max_terminated_length": 1709.0, "completions/mean_length": 699.9625244140625, "completions/mean_terminated_length": 699.9625244140625, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "epoch": 0.4437548487199379, "grad_norm": 0.23076042532920837, "kl": 0.5810546875, "learning_rate": 3.2683762100223293e-06, "loss": -0.0118, "num_tokens": 208982812.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1387.0, "completions/max_terminated_length": 1387.0, "completions/mean_length": 718.0750122070312, "completions/mean_terminated_length": 718.0750122070312, "completions/min_length": 294.0, "completions/min_terminated_length": 294.0, "epoch": 0.44414274631497286, "grad_norm": 0.2204977422952652, "kl": 0.59375, "learning_rate": 3.2656244302330764e-06, "loss": 0.0059, "num_tokens": 209092786.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1145 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1653.0, "completions/max_terminated_length": 1653.0, "completions/mean_length": 693.3250122070312, "completions/mean_terminated_length": 693.3250122070312, "completions/min_length": 152.0, "completions/min_terminated_length": 152.0, "epoch": 0.44453064391000774, "grad_norm": 0.04437399283051491, "kl": 0.568359375, "learning_rate": 3.2628718362024667e-06, "loss": 0.0057, "num_tokens": 209195892.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1146 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1519.0, "completions/max_terminated_length": 1519.0, "completions/mean_length": 647.7125244140625, "completions/mean_terminated_length": 647.7125244140625, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.4449185415050427, "grad_norm": 0.0533582977950573, "kl": 0.6044921875, "learning_rate": 3.2601184322772244e-06, "loss": 0.006, "num_tokens": 209298717.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1147 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1689.0, "completions/max_terminated_length": 1689.0, "completions/mean_length": 741.6000366210938, "completions/mean_terminated_length": 741.6000366210938, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "epoch": 0.44530643910007756, "grad_norm": 0.04430073872208595, "kl": 0.57421875, "learning_rate": 3.2573642228053466e-06, "loss": 0.0057, "num_tokens": 209398829.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1148 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1464.0, "completions/mean_length": 779.7750244140625, "completions/mean_terminated_length": 737.7974853515625, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.4456943366951125, "grad_norm": 24.385791778564453, "kl": 1.2919921875, "learning_rate": 3.2546092121361077e-06, "loss": 0.0712, "num_tokens": 209504451.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1149 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1756.0, "completions/max_terminated_length": 1756.0, "completions/mean_length": 717.5250244140625, "completions/mean_terminated_length": 717.5250244140625, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.44608223429014743, "grad_norm": 0.33190250396728516, "kl": 0.6015625, "learning_rate": 3.251853404620047e-06, "loss": -0.0007, "num_tokens": 209605309.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1150 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1775.0, "completions/mean_length": 875.7875366210938, "completions/mean_terminated_length": 750.3246459960938, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.4464701318851823, "grad_norm": 1.2180579900741577, "kl": 0.8193359375, "learning_rate": 3.249096804608957e-06, "loss": 0.0082, "num_tokens": 209732372.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1151 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2019.0, "completions/max_terminated_length": 2019.0, "completions/mean_length": 843.4625244140625, "completions/mean_terminated_length": 843.4625244140625, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.44685802948021724, "grad_norm": 0.055416885763406754, "kl": 0.5673828125, "learning_rate": 3.246339416455889e-06, "loss": 0.0057, "num_tokens": 209848545.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1152 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2845.0, "completions/mean_length": 998.6500244140625, "completions/mean_terminated_length": 919.2307739257812, "completions/min_length": 328.0, "completions/min_terminated_length": 328.0, "epoch": 0.4472459270752521, "grad_norm": 0.3631346523761749, "kl": 0.51953125, "learning_rate": 3.243581244515134e-06, "loss": -0.0228, "num_tokens": 209975061.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1153 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2370.0, "completions/mean_length": 1181.0374755859375, "completions/mean_terminated_length": 1106.294921875, "completions/min_length": 391.0, "completions/min_terminated_length": 391.0, "epoch": 0.44763382467028706, "grad_norm": 0.3132648766040802, "kl": 0.4814453125, "learning_rate": 3.2408222931422213e-06, "loss": -0.0068, "num_tokens": 210115664.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1154 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2451.0, "completions/max_terminated_length": 2451.0, "completions/mean_length": 1239.3875732421875, "completions/mean_terminated_length": 1239.3875732421875, "completions/min_length": 423.0, "completions/min_terminated_length": 423.0, "epoch": 0.44802172226532194, "grad_norm": 0.04190245270729065, "kl": 0.44189453125, "learning_rate": 3.2380625666939126e-06, "loss": 0.0044, "num_tokens": 210257831.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1155 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2945.0, "completions/mean_length": 1251.7625732421875, "completions/mean_terminated_length": 1215.759521484375, "completions/min_length": 402.0, "completions/min_terminated_length": 402.0, "epoch": 0.4484096198603569, "grad_norm": 0.044455043971538544, "kl": 0.44580078125, "learning_rate": 3.235302069528192e-06, "loss": 0.0045, "num_tokens": 210402276.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1156 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3064.0, "completions/mean_length": 1549.9625244140625, "completions/mean_terminated_length": 1484.6795654296875, "completions/min_length": 297.0, "completions/min_terminated_length": 297.0, "epoch": 0.44879751745539176, "grad_norm": 0.04589772969484329, "kl": 0.41064453125, "learning_rate": 3.2325408060042614e-06, "loss": 0.0041, "num_tokens": 210573345.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1157 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3513.0, "completions/mean_length": 1438.6875, "completions/mean_terminated_length": 1370.55126953125, "completions/min_length": 449.0, "completions/min_terminated_length": 449.0, "epoch": 0.4491854150504267, "grad_norm": 0.037321120500564575, "kl": 0.40966796875, "learning_rate": 3.2297787804825336e-06, "loss": 0.0041, "num_tokens": 210738624.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1158 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 4041.0, "completions/mean_length": 1683.5374755859375, "completions/mean_terminated_length": 1653.0, "completions/min_length": 515.0, "completions/min_terminated_length": 515.0, "epoch": 0.4495733126454616, "grad_norm": 0.15176838636398315, "kl": 0.3837890625, "learning_rate": 3.2270159973246234e-06, "loss": 0.0028, "num_tokens": 210918363.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1159 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3587.0, "completions/mean_length": 1675.7874755859375, "completions/mean_terminated_length": 1613.7308349609375, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "epoch": 0.4499612102404965, "grad_norm": 0.036723703145980835, "kl": 0.37060546875, "learning_rate": 3.2242524608933425e-06, "loss": 0.0037, "num_tokens": 211095074.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1160 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3645.0, "completions/mean_length": 1538.5625, "completions/mean_terminated_length": 1438.922119140625, "completions/min_length": 319.0, "completions/min_terminated_length": 319.0, "epoch": 0.4503491078355314, "grad_norm": 0.44512757658958435, "kl": 0.3857421875, "learning_rate": 3.221488175552691e-06, "loss": 0.0055, "num_tokens": 211255935.0, "reward": 0.09749999642372131, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1161 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3836.0, "completions/mean_length": 1741.9000244140625, "completions/mean_terminated_length": 1618.0, "completions/min_length": 300.0, "completions/min_terminated_length": 300.0, "epoch": 0.45073700543056633, "grad_norm": 0.10875444859266281, "kl": 0.3681640625, "learning_rate": 3.218723145667856e-06, "loss": -0.0268, "num_tokens": 211437711.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1162 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3456.0, "completions/mean_length": 1701.8875732421875, "completions/mean_terminated_length": 1472.3150634765625, "completions/min_length": 307.0, "completions/min_terminated_length": 307.0, "epoch": 0.45112490302560126, "grad_norm": 20.328022003173828, "kl": 0.36865234375, "learning_rate": 3.215957375605193e-06, "loss": 0.0353, "num_tokens": 211614862.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1163 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3969.0, "completions/mean_length": 1673.7249755859375, "completions/mean_terminated_length": 1643.0633544921875, "completions/min_length": 399.0, "completions/min_terminated_length": 399.0, "epoch": 0.45151280062063615, "grad_norm": 0.1476094126701355, "kl": 0.36181640625, "learning_rate": 3.2131908697322335e-06, "loss": -0.0033, "num_tokens": 211814824.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1164 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3540.0, "completions/mean_length": 1457.0999755859375, "completions/mean_terminated_length": 1389.4359130859375, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "epoch": 0.4519006982156711, "grad_norm": 0.9041408896446228, "kl": 0.42919921875, "learning_rate": 3.2104236324176664e-06, "loss": 0.0043, "num_tokens": 211985624.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1165 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3069.0, "completions/mean_length": 1347.4749755859375, "completions/mean_terminated_length": 1277.0, "completions/min_length": 371.0, "completions/min_terminated_length": 371.0, "epoch": 0.45228859581070596, "grad_norm": 0.2033308744430542, "kl": 0.4013671875, "learning_rate": 3.207655668031337e-06, "loss": -0.0182, "num_tokens": 212137238.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1166 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3606.0, "completions/mean_length": 1506.2625732421875, "completions/mean_terminated_length": 1369.9605712890625, "completions/min_length": 257.0, "completions/min_terminated_length": 257.0, "epoch": 0.4526764934057409, "grad_norm": 0.44294464588165283, "kl": 0.353515625, "learning_rate": 3.2048869809442408e-06, "loss": -0.041, "num_tokens": 212303275.0, "reward": 0.0949999988079071, "reward_std": 0.011700168251991272, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1167 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3546.0, "completions/mean_length": 1567.0625, "completions/mean_terminated_length": 1398.4666748046875, "completions/min_length": 420.0, "completions/min_terminated_length": 420.0, "epoch": 0.4530643910007758, "grad_norm": 0.22635884582996368, "kl": 0.3623046875, "learning_rate": 3.2021175755285106e-06, "loss": -0.028, "num_tokens": 212481216.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1168 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3754.0, "completions/mean_length": 1633.0374755859375, "completions/mean_terminated_length": 1601.86083984375, "completions/min_length": 405.0, "completions/min_terminated_length": 405.0, "epoch": 0.4534522885958107, "grad_norm": 0.2664501965045929, "kl": 0.3388671875, "learning_rate": 3.1993474561574158e-06, "loss": -0.0249, "num_tokens": 212652811.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1169 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3594.0, "completions/mean_length": 1588.9625244140625, "completions/mean_terminated_length": 1524.6795654296875, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.4538401861908456, "grad_norm": 0.21072901785373688, "kl": 0.31982421875, "learning_rate": 3.196576627205354e-06, "loss": -0.0061, "num_tokens": 212830088.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1170 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3790.0, "completions/mean_length": 1835.6375732421875, "completions/mean_terminated_length": 1549.1126708984375, "completions/min_length": 328.0, "completions/min_terminated_length": 328.0, "epoch": 0.45422808378588053, "grad_norm": 0.41282835602760315, "kl": 0.3369140625, "learning_rate": 3.193805093047842e-06, "loss": -0.0619, "num_tokens": 213029011.0, "reward": 0.11375000327825546, "reward_std": 0.04344821348786354, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.1111220195889473, "step": 1171 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4072.0, "completions/mean_length": 2312.22509765625, "completions/mean_terminated_length": 1900.5845947265625, "completions/min_length": 506.0, "completions/min_terminated_length": 506.0, "epoch": 0.4546159813809154, "grad_norm": 0.035579320043325424, "kl": 0.29345703125, "learning_rate": 3.19103285806151e-06, "loss": 0.0029, "num_tokens": 213266957.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1172 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3902.0, "completions/mean_length": 2586.837646484375, "completions/mean_terminated_length": 1900.8544921875, "completions/min_length": 815.0, "completions/min_terminated_length": 815.0, "epoch": 0.45500387897595035, "grad_norm": 0.029414674267172813, "kl": 0.271484375, "learning_rate": 3.1882599266240953e-06, "loss": 0.0027, "num_tokens": 213524376.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1173 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 4028.0, "completions/mean_length": 2512.0, "completions/mean_terminated_length": 2084.571533203125, "completions/min_length": 407.0, "completions/min_terminated_length": 407.0, "epoch": 0.4553917765709853, "grad_norm": 0.14121979475021362, "kl": 0.26953125, "learning_rate": 3.185486303114436e-06, "loss": -0.0126, "num_tokens": 213766672.0, "reward": 0.10374999046325684, "reward_std": 0.01767767034471035, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472136124968529, "step": 1174 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4055.0, "completions/mean_length": 2776.550048828125, "completions/mean_terminated_length": 1850.1275634765625, "completions/min_length": 535.0, "completions/min_terminated_length": 535.0, "epoch": 0.45577967416602017, "grad_norm": 10.113360404968262, "kl": 0.26806640625, "learning_rate": 3.1827119919124616e-06, "loss": 0.0125, "num_tokens": 214028268.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1175 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 4068.0, "completions/mean_length": 2901.387451171875, "completions/mean_terminated_length": 2104.979248046875, "completions/min_length": 452.0, "completions/min_terminated_length": 452.0, "epoch": 0.4561675717610551, "grad_norm": 0.3163341283798218, "kl": 0.257568359375, "learning_rate": 3.1799369973991885e-06, "loss": 0.0166, "num_tokens": 214315283.0, "reward": 0.0962500050663948, "reward_std": 0.008164633996784687, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1176 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3970.0, "completions/mean_length": 2771.97509765625, "completions/mean_terminated_length": 1977.5599365234375, "completions/min_length": 522.0, "completions/min_terminated_length": 522.0, "epoch": 0.45655546935609, "grad_norm": 0.1584012359380722, "kl": 0.252197265625, "learning_rate": 3.177161323956712e-06, "loss": 0.0187, "num_tokens": 214585281.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1177 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3849.0, "completions/mean_length": 2952.112548828125, "completions/mean_terminated_length": 2228.428466796875, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "epoch": 0.4569433669511249, "grad_norm": 0.5490285158157349, "kl": 0.242919921875, "learning_rate": 3.1743849759681988e-06, "loss": 0.018, "num_tokens": 214878466.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1178 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4093.0, "completions/mean_length": 2456.425048828125, "completions/mean_terminated_length": 2078.0615234375, "completions/min_length": 491.0, "completions/min_terminated_length": 491.0, "epoch": 0.4573312645461598, "grad_norm": 0.14504960179328918, "kl": 0.258056640625, "learning_rate": 3.171607957817881e-06, "loss": -0.0206, "num_tokens": 215121524.0, "reward": 0.10249999910593033, "reward_std": 0.007071068976074457, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360680624842644, "step": 1179 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 4005.0, "completions/mean_length": 2149.587646484375, "completions/mean_terminated_length": 1736.712158203125, "completions/min_length": 353.0, "completions/min_terminated_length": 353.0, "epoch": 0.45771916214119474, "grad_norm": 0.203142911195755, "kl": 0.279296875, "learning_rate": 3.1688302738910504e-06, "loss": 0.0305, "num_tokens": 215345379.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1180 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 4069.0, "completions/mean_length": 2765.487548828125, "completions/mean_terminated_length": 2124.870361328125, "completions/min_length": 364.0, "completions/min_terminated_length": 364.0, "epoch": 0.4581070597362296, "grad_norm": 0.04543924704194069, "kl": 0.260986328125, "learning_rate": 3.1660519285740475e-06, "loss": 0.0026, "num_tokens": 215622970.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1181 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4036.0, "completions/mean_length": 2068.612548828125, "completions/mean_terminated_length": 1778.9857177734375, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "epoch": 0.45849495733126455, "grad_norm": 0.0441645011305809, "kl": 0.28759765625, "learning_rate": 3.163272926254261e-06, "loss": 0.0029, "num_tokens": 215837699.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1182 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3623.0, "completions/mean_length": 2124.72509765625, "completions/mean_terminated_length": 1631.90625, "completions/min_length": 308.0, "completions/min_terminated_length": 308.0, "epoch": 0.45888285492629943, "grad_norm": 0.030590318143367767, "kl": 0.28857421875, "learning_rate": 3.1604932713201136e-06, "loss": 0.0029, "num_tokens": 216049965.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1183 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3867.0, "completions/mean_length": 1860.6624755859375, "completions/mean_terminated_length": 1466.191162109375, "completions/min_length": 410.0, "completions/min_terminated_length": 410.0, "epoch": 0.45927075252133437, "grad_norm": 0.07480499893426895, "kl": 0.30322265625, "learning_rate": 3.157712968161059e-06, "loss": -0.0229, "num_tokens": 216244082.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1184 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 4092.0, "completions/mean_length": 1771.6875, "completions/mean_terminated_length": 1583.229736328125, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.4596586501163693, "grad_norm": 0.23475708067417145, "kl": 0.3095703125, "learning_rate": 3.1549320211675775e-06, "loss": -0.0144, "num_tokens": 216440489.0, "reward": 0.10499999672174454, "reward_std": 0.02828427217900753, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.04974619299173355, "step": 1185 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 4075.0, "completions/mean_length": 1587.2750244140625, "completions/mean_terminated_length": 1420.0267333984375, "completions/min_length": 319.0, "completions/min_terminated_length": 319.0, "epoch": 0.4600465477114042, "grad_norm": 0.16350385546684265, "kl": 0.33203125, "learning_rate": 3.1521504347311623e-06, "loss": 0.0076, "num_tokens": 216617247.0, "reward": 0.10374999046325684, "reward_std": 0.01767767034471035, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.044721364974975586, "step": 1186 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 4036.0, "completions/mean_length": 1623.2000732421875, "completions/mean_terminated_length": 1348.4444580078125, "completions/min_length": 403.0, "completions/min_terminated_length": 403.0, "epoch": 0.4604344453064391, "grad_norm": 0.03716783970594406, "kl": 0.314453125, "learning_rate": 3.1493682132443187e-06, "loss": 0.0032, "num_tokens": 216795607.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1187 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3591.0, "completions/mean_length": 1734.5999755859375, "completions/mean_terminated_length": 1472.2222900390625, "completions/min_length": 295.0, "completions/min_terminated_length": 295.0, "epoch": 0.460822342901474, "grad_norm": 0.03140425309538841, "kl": 0.3076171875, "learning_rate": 3.1465853611005536e-06, "loss": 0.0031, "num_tokens": 216979847.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1188 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3916.0, "completions/mean_length": 1603.3875732421875, "completions/mean_terminated_length": 1437.21337890625, "completions/min_length": 410.0, "completions/min_terminated_length": 410.0, "epoch": 0.46121024049650894, "grad_norm": 0.030238060280680656, "kl": 0.33251953125, "learning_rate": 3.143801882694372e-06, "loss": 0.0033, "num_tokens": 217159454.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1189 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3353.0, "completions/mean_length": 1493.0875244140625, "completions/mean_terminated_length": 1163.140869140625, "completions/min_length": 275.0, "completions/min_terminated_length": 275.0, "epoch": 0.4615981380915438, "grad_norm": 0.10919120907783508, "kl": 0.34814453125, "learning_rate": 3.1410177824212634e-06, "loss": 0.0035, "num_tokens": 217320285.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1190 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3621.0, "completions/mean_length": 1394.5125732421875, "completions/mean_terminated_length": 1214.413330078125, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.46198603568657876, "grad_norm": 0.10138589143753052, "kl": 0.3359375, "learning_rate": 3.1382330646777055e-06, "loss": 0.0034, "num_tokens": 217478422.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1191 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2829.0, "completions/mean_length": 1242.0, "completions/mean_terminated_length": 1168.820556640625, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.46237393328161364, "grad_norm": 0.18154403567314148, "kl": 0.35205078125, "learning_rate": 3.135447733861145e-06, "loss": -0.0047, "num_tokens": 217622430.0, "reward": 0.10125000774860382, "reward_std": 0.010606602765619755, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1192 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3238.0, "completions/mean_length": 1406.3125, "completions/mean_terminated_length": 1188.229736328125, "completions/min_length": 386.0, "completions/min_terminated_length": 386.0, "epoch": 0.46276183087664857, "grad_norm": 0.24058547616004944, "kl": 0.36181640625, "learning_rate": 3.1326617943700007e-06, "loss": 0.0087, "num_tokens": 217784199.0, "reward": 0.10499999672174454, "reward_std": 0.014142136089503765, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472136124968529, "step": 1193 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2910.0, "completions/mean_length": 1047.5, "completions/mean_terminated_length": 1008.9114379882812, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "epoch": 0.46314972847168345, "grad_norm": 0.30454543232917786, "kl": 0.37109375, "learning_rate": 3.1298752506036513e-06, "loss": -0.0117, "num_tokens": 217913695.0, "reward": 0.09749999642372131, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1194 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3964.0, "completions/mean_length": 1349.4375, "completions/mean_terminated_length": 1204.881591796875, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.4635376260667184, "grad_norm": 0.48139485716819763, "kl": 0.390625, "learning_rate": 3.1270881069624293e-06, "loss": 0.0064, "num_tokens": 218071298.0, "reward": 0.09749999642372131, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1195 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 4072.0, "completions/mean_length": 1514.5, "completions/mean_terminated_length": 1227.6666259765625, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.46392552366175327, "grad_norm": 2.88093638420105, "kl": 4.94189453125, "learning_rate": 3.124300367847616e-06, "loss": 0.036, "num_tokens": 218234370.0, "reward": 0.0949999988079071, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1196 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3405.0, "completions/mean_length": 1189.237548828125, "completions/mean_terminated_length": 995.453369140625, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.4643134212567882, "grad_norm": 0.2878471910953522, "kl": 0.37451171875, "learning_rate": 3.1215120376614307e-06, "loss": 0.0304, "num_tokens": 218367157.0, "reward": 0.10625000298023224, "reward_std": 0.02474873699247837, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 1197 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3746.0, "completions/mean_length": 1255.0374755859375, "completions/mean_terminated_length": 894.9154663085938, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.46470131885182314, "grad_norm": 3.0192201137542725, "kl": 8.34521484375, "learning_rate": 3.1187231208070283e-06, "loss": 0.1376, "num_tokens": 218516528.0, "reward": 0.12000000476837158, "reward_std": 0.07937442511320114, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.1364096701145172, "step": 1198 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3168.0, "completions/mean_length": 1742.0250244140625, "completions/mean_terminated_length": 1106.825439453125, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.465089216446858, "grad_norm": 7.62866735458374, "kl": 4.63916015625, "learning_rate": 3.1159336216884883e-06, "loss": 0.115, "num_tokens": 218709906.0, "reward": 0.10375000536441803, "reward_std": 0.029377836734056473, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 1199 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3782.0, "completions/mean_length": 1864.0, "completions/mean_terminated_length": 849.4545288085938, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.46547711404189296, "grad_norm": 5.951154708862305, "kl": 25.0625, "learning_rate": 3.1131435447108117e-06, "loss": 0.4367, "num_tokens": 218906362.0, "reward": 0.09250000864267349, "reward_std": 0.030471403151750565, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360680624842644, "step": 1200 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3302.0, "completions/mean_length": 2087.97509765625, "completions/mean_terminated_length": 883.1599731445312, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "epoch": 0.46586501163692784, "grad_norm": 10.396591186523438, "kl": 30.1044921875, "learning_rate": 3.1103528942799096e-06, "loss": 0.5414, "num_tokens": 219119672.0, "reward": 0.09875001013278961, "reward_std": 0.04153930023312569, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.04386400431394577, "step": 1201 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2833.0, "completions/mean_length": 1616.9000244140625, "completions/mean_terminated_length": 844.7212524414062, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.4662529092319628, "grad_norm": 8.85931396484375, "kl": 35.140625, "learning_rate": 3.1075616748026004e-06, "loss": 0.5691, "num_tokens": 219288680.0, "reward": 0.12250000238418579, "reward_std": 0.08144745975732803, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.11066542565822601, "step": 1202 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 3903.0, "completions/mean_length": 2036.7125244140625, "completions/mean_terminated_length": 927.8654174804688, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.46664080682699766, "grad_norm": 6.129714488983154, "kl": 39.640625, "learning_rate": 3.1047698906866013e-06, "loss": 0.6451, "num_tokens": 219493489.0, "reward": 0.0950000062584877, "reward_std": 0.07203599065542221, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.09426598995923996, "step": 1203 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2387.0, "completions/mean_length": 2232.387451171875, "completions/mean_terminated_length": 782.9111328125, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "epoch": 0.4670287044220326, "grad_norm": 18.460859298706055, "kl": 37.78125, "learning_rate": 3.10197754634052e-06, "loss": 0.5692, "num_tokens": 219719496.0, "reward": 0.13375000655651093, "reward_std": 0.10895007103681564, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.05250000208616257, "rewards/unicoder_reward_fn/std": 0.182094007730484, "step": 1204 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3087.0, "completions/mean_length": 2135.85009765625, "completions/mean_terminated_length": 895.7550659179688, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.4674166020170675, "grad_norm": 11.47602367401123, "kl": 42.25, "learning_rate": 3.0991846461738488e-06, "loss": 0.6607, "num_tokens": 219936428.0, "reward": 0.10750000923871994, "reward_std": 0.09471943229436874, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.10919730365276337, "step": 1205 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2506.0, "completions/mean_length": 1899.112548828125, "completions/mean_terminated_length": 841.3518676757812, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.4678044996121024, "grad_norm": 10.980424880981445, "kl": 23.75, "learning_rate": 3.0963911945969597e-06, "loss": 0.5707, "num_tokens": 220140029.0, "reward": 0.11000001430511475, "reward_std": 0.08929470181465149, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.030000001192092896, "rewards/unicoder_reward_fn/std": 0.12365734577178955, "step": 1206 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2066.0, "completions/mean_length": 1115.25, "completions/mean_terminated_length": 784.0555419921875, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.4681923972071373, "grad_norm": 5.500183582305908, "kl": 4.08203125, "learning_rate": 3.0935971960210927e-06, "loss": 0.1404, "num_tokens": 220262825.0, "reward": 0.14875002205371857, "reward_std": 0.10290957987308502, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.05250000208616257, "rewards/unicoder_reward_fn/std": 0.15829390287399292, "step": 1207 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2290.0, "completions/mean_length": 1078.6500244140625, "completions/mean_terminated_length": 647.5999755859375, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.4685802948021722, "grad_norm": 6.139678955078125, "kl": 9.65771484375, "learning_rate": 3.090802654858353e-06, "loss": 0.3056, "num_tokens": 220398637.0, "reward": 0.13625000417232513, "reward_std": 0.09321650862693787, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.042500000447034836, "rewards/unicoder_reward_fn/std": 0.13758771121501923, "step": 1208 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2956.0, "completions/mean_length": 1217.300048828125, "completions/mean_terminated_length": 897.4444580078125, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "epoch": 0.46896819239720716, "grad_norm": 3.4526076316833496, "kl": 12.81298828125, "learning_rate": 3.0880075755217037e-06, "loss": 0.3082, "num_tokens": 220549645.0, "reward": 0.10875000804662704, "reward_std": 0.03923286870121956, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.0725223645567894, "step": 1209 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3355.0, "completions/mean_length": 1020.9750366210938, "completions/mean_terminated_length": 679.3055419921875, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.46935608999224204, "grad_norm": 1.7739555835723877, "kl": 5.50732421875, "learning_rate": 3.0852119624249553e-06, "loss": 0.1172, "num_tokens": 220674683.0, "reward": 0.14625000953674316, "reward_std": 0.09239669144153595, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.18418481945991516, "step": 1210 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2110.0, "completions/mean_length": 886.4625244140625, "completions/mean_terminated_length": 578.6986083984375, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.469743987587277, "grad_norm": 1.340546727180481, "kl": 4.5537109375, "learning_rate": 3.0824158199827634e-06, "loss": 0.1553, "num_tokens": 220786336.0, "reward": 0.20499999821186066, "reward_std": 0.14003805816173553, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.28452715277671814, "step": 1211 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2972.0, "completions/mean_length": 981.0499877929688, "completions/mean_terminated_length": 773.3866577148438, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.47013188518231186, "grad_norm": 0.9002012014389038, "kl": 3.119140625, "learning_rate": 3.0796191526106186e-06, "loss": 0.1483, "num_tokens": 220910340.0, "reward": 0.16375000774860382, "reward_std": 0.12453608959913254, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.20853619277477264, "step": 1212 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3833.0, "completions/mean_length": 868.4249877929688, "completions/mean_terminated_length": 742.6753540039062, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.4705197827773468, "grad_norm": 1.5576125383377075, "kl": 5.5263671875, "learning_rate": 3.0768219647248387e-06, "loss": 0.2518, "num_tokens": 221024462.0, "reward": 0.1862500160932541, "reward_std": 0.09941747039556503, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.08750000596046448, "rewards/unicoder_reward_fn/std": 0.1851274073123932, "step": 1213 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2229.0, "completions/mean_length": 817.1500244140625, "completions/mean_terminated_length": 502.7397155761719, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.4709076803723817, "grad_norm": 3.685164213180542, "kl": 20.60546875, "learning_rate": 3.0740242607425676e-06, "loss": 0.4446, "num_tokens": 221127978.0, "reward": 0.17499999701976776, "reward_std": 0.09832517057657242, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.08000000566244125, "rewards/unicoder_reward_fn/std": 0.2125195562839508, "step": 1214 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 2999.0, "completions/mean_length": 1425.2625732421875, "completions/mean_terminated_length": 535.0167236328125, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.4712955779674166, "grad_norm": 19.69581413269043, "kl": 73.6875, "learning_rate": 3.07122604508176e-06, "loss": 1.095, "num_tokens": 221292535.0, "reward": 0.08624999970197678, "reward_std": 0.053599368780851364, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.054191410541534424, "step": 1215 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1883.0, "completions/mean_length": 1690.987548828125, "completions/mean_terminated_length": 465.7924499511719, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.4716834755624515, "grad_norm": 29.395030975341797, "kl": 72.0, "learning_rate": 3.0684273221611814e-06, "loss": 0.9736, "num_tokens": 221480238.0, "reward": 0.12250000238418579, "reward_std": 0.1244409829378128, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.14755390584468842, "step": 1216 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1548.0, "completions/mean_length": 1586.0250244140625, "completions/mean_terminated_length": 445.12725830078125, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.47207137315748643, "grad_norm": 22.605745315551758, "kl": 44.875, "learning_rate": 3.065628096400397e-06, "loss": 0.6011, "num_tokens": 221652600.0, "reward": 0.13250000774860382, "reward_std": 0.12136118859052658, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.16098982095718384, "step": 1217 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3556.0, "completions/mean_length": 2031.0125732421875, "completions/mean_terminated_length": 581.1276245117188, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.4724592707525213, "grad_norm": 30.198638916015625, "kl": 51.6875, "learning_rate": 3.0628283722197662e-06, "loss": 0.7864, "num_tokens": 221870281.0, "reward": 0.13625000417232513, "reward_std": 0.1710900068283081, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.18891765177249908, "step": 1218 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 1911.0, "completions/mean_length": 2200.925048828125, "completions/mean_terminated_length": 486.3333435058594, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.47284716834755625, "grad_norm": 6.199952602386475, "kl": 41.625, "learning_rate": 3.060028154040436e-06, "loss": 0.8734, "num_tokens": 222093963.0, "reward": 0.1550000011920929, "reward_std": 0.17740723490715027, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.2158292680978775, "step": 1219 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1558.0, "completions/mean_length": 1751.8250732421875, "completions/mean_terminated_length": 345.32000732421875, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.4732350659425912, "grad_norm": 15.309721946716309, "kl": 20.3125, "learning_rate": 3.057227446284333e-06, "loss": 0.68, "num_tokens": 222286669.0, "reward": 0.26624998450279236, "reward_std": 0.2249164879322052, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.31341323256492615, "step": 1220 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 2394.0, "completions/mean_length": 1675.3375244140625, "completions/mean_terminated_length": 371.90386962890625, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.47362296353762606, "grad_norm": 40.70164108276367, "kl": 13.9375, "learning_rate": 3.054426253374156e-06, "loss": 0.538, "num_tokens": 222478464.0, "reward": 0.2587500214576721, "reward_std": 0.17030470073223114, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.1899999976158142, "rewards/unicoder_reward_fn/std": 0.3132516145706177, "step": 1221 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3031.0, "completions/mean_length": 1711.675048828125, "completions/mean_terminated_length": 355.8823547363281, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.474010861132661, "grad_norm": 69.12137603759766, "kl": 17.140625, "learning_rate": 3.051624579733372e-06, "loss": 0.6017, "num_tokens": 222663438.0, "reward": 0.32624998688697815, "reward_std": 0.27807363867759705, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.26249998807907104, "rewards/unicoder_reward_fn/std": 0.3331381380558014, "step": 1222 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1336.0, "completions/mean_length": 1251.8375244140625, "completions/mean_terminated_length": 365.9508056640625, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.4743987587276959, "grad_norm": 84.33834838867188, "kl": 16.296875, "learning_rate": 3.048822429786207e-06, "loss": 0.6572, "num_tokens": 222812321.0, "reward": 0.19500000774860382, "reward_std": 0.1914529800415039, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.19251829385757446, "step": 1223 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3994.0, "completions/mean_length": 990.5375366210938, "completions/mean_terminated_length": 546.9000244140625, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.4747866563227308, "grad_norm": 8.592666625976562, "kl": 1.447265625, "learning_rate": 3.0460198079576356e-06, "loss": 0.1665, "num_tokens": 222933516.0, "reward": 0.15375001728534698, "reward_std": 0.12115573137998581, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.05499999597668648, "rewards/unicoder_reward_fn/std": 0.15903186798095703, "step": 1224 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3993.0, "completions/mean_length": 1736.9625244140625, "completions/mean_terminated_length": 785.0701904296875, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.4751745539177657, "grad_norm": 0.14703363180160522, "kl": 0.3232421875, "learning_rate": 3.0432167186733834e-06, "loss": 0.0297, "num_tokens": 223120929.0, "reward": 0.10875000059604645, "reward_std": 0.03181980550289154, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.07044164091348648, "step": 1225 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 3993.0, "completions/mean_length": 2632.737548828125, "completions/mean_terminated_length": 1169.4749755859375, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.47556245151280063, "grad_norm": 0.13500066101551056, "kl": 0.28955078125, "learning_rate": 3.0404131663599084e-06, "loss": -0.0436, "num_tokens": 223382508.0, "reward": 0.09750000387430191, "reward_std": 0.004629100672900677, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1226 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3650.0, "completions/mean_length": 2420.462646484375, "completions/mean_terminated_length": 1049.5682373046875, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.4759503491078355, "grad_norm": 0.09706425666809082, "kl": 0.29541015625, "learning_rate": 3.0376091554444037e-06, "loss": 0.0304, "num_tokens": 223619529.0, "reward": 0.11250000447034836, "reward_std": 0.0353553406894207, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 1227 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 3467.0, "completions/mean_length": 3342.875, "completions/mean_terminated_length": 1585.5833740234375, "completions/min_length": 303.0, "completions/min_terminated_length": 303.0, "epoch": 0.47633824670287045, "grad_norm": 2.0636346340179443, "kl": 0.5654296875, "learning_rate": 3.0348046903547844e-06, "loss": 0.0035, "num_tokens": 223937719.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1228 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 3909.0, "completions/mean_length": 3305.449951171875, "completions/mean_terminated_length": 1663.53857421875, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.47672614429790533, "grad_norm": 0.04585544392466545, "kl": 0.28564453125, "learning_rate": 3.031999775519685e-06, "loss": 0.0029, "num_tokens": 224253323.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1229 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2685.0, "completions/mean_length": 3484.5625, "completions/mean_terminated_length": 1218.6470947265625, "completions/min_length": 394.0, "completions/min_terminated_length": 394.0, "epoch": 0.47711404189294027, "grad_norm": 0.030072100460529327, "kl": 0.2685546875, "learning_rate": 3.0291944153684482e-06, "loss": 0.0027, "num_tokens": 224582712.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1230 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3827.0, "completions/mean_length": 3486.362548828125, "completions/mean_terminated_length": 1773.5714111328125, "completions/min_length": 282.0, "completions/min_terminated_length": 282.0, "epoch": 0.4775019394879752, "grad_norm": 0.032058872282505035, "kl": 0.27001953125, "learning_rate": 3.02638861433112e-06, "loss": 0.0027, "num_tokens": 224906581.0, "reward": 0.14000000059604645, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.12075711041688919, "step": 1231 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.775, "completions/max_length": 4096.0, "completions/max_terminated_length": 3856.0, "completions/mean_length": 3610.050048828125, "completions/mean_terminated_length": 1936.2222900390625, "completions/min_length": 233.0, "completions/min_terminated_length": 233.0, "epoch": 0.4778898370830101, "grad_norm": 0.09866160154342651, "kl": 0.259765625, "learning_rate": 3.0235823768384438e-06, "loss": -0.002, "num_tokens": 225249593.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1232 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2900.0, "completions/mean_length": 3445.800048828125, "completions/mean_terminated_length": 1619.047607421875, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.478277734678045, "grad_norm": 0.03399457409977913, "kl": 0.27197265625, "learning_rate": 3.0207757073218514e-06, "loss": 0.0027, "num_tokens": 225574441.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1233 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 4088.0, "completions/mean_length": 3257.425048828125, "completions/mean_terminated_length": 1611.3333740234375, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.4786656322730799, "grad_norm": 0.029711756855249405, "kl": 0.2763671875, "learning_rate": 3.0179686102134574e-06, "loss": 0.0028, "num_tokens": 225884427.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1234 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 3829.0, "completions/mean_length": 3171.487548828125, "completions/mean_terminated_length": 1251.34619140625, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "epoch": 0.47905352986811484, "grad_norm": 0.18904626369476318, "kl": 0.2744140625, "learning_rate": 3.015161089946051e-06, "loss": -0.0168, "num_tokens": 226186842.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1235 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 4056.0, "completions/mean_length": 3330.762451171875, "completions/mean_terminated_length": 1545.2083740234375, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.4794414274631497, "grad_norm": 1.6451194286346436, "kl": 0.94921875, "learning_rate": 3.01235315095309e-06, "loss": 0.0148, "num_tokens": 226494439.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1236 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2587.0, "completions/mean_length": 2938.087646484375, "completions/mean_terminated_length": 1107.838623046875, "completions/min_length": 257.0, "completions/min_terminated_length": 257.0, "epoch": 0.47982932505818465, "grad_norm": 1.6745362281799316, "kl": 1.01220703125, "learning_rate": 3.0095447976686935e-06, "loss": 0.0161, "num_tokens": 226778350.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1237 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4065.0, "completions/mean_length": 3034.087646484375, "completions/mean_terminated_length": 1521.666748046875, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "epoch": 0.48021722265321953, "grad_norm": 0.10845030099153519, "kl": 0.275390625, "learning_rate": 3.0067360345276355e-06, "loss": -0.0327, "num_tokens": 227070413.0, "reward": 0.09750000387430191, "reward_std": 0.004629100672900677, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1238 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 3941.0, "completions/mean_length": 2745.0, "completions/mean_terminated_length": 1522.666748046875, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.48060512024825447, "grad_norm": 0.08795331418514252, "kl": 0.2861328125, "learning_rate": 3.0039268659653335e-06, "loss": -0.0284, "num_tokens": 227334933.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1239 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3797.0, "completions/mean_length": 2247.6875, "completions/mean_terminated_length": 1196.686279296875, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.48099301784328935, "grad_norm": 0.038482051342725754, "kl": 0.31103515625, "learning_rate": 3.001117296417852e-06, "loss": 0.0031, "num_tokens": 227556964.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1240 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3892.0, "completions/mean_length": 2353.5, "completions/mean_terminated_length": 1514.5185546875, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.4813809154383243, "grad_norm": 0.09808485209941864, "kl": 0.3017578125, "learning_rate": 2.9983073303218817e-06, "loss": 0.0384, "num_tokens": 227787452.0, "reward": 0.10125000774860382, "reward_std": 0.0035355340223759413, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1241 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3228.0, "completions/mean_length": 2057.050048828125, "completions/mean_terminated_length": 1234.3157958984375, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.48176881303335917, "grad_norm": 11.523490905761719, "kl": 1.76611328125, "learning_rate": 2.9954969721147437e-06, "loss": 0.0527, "num_tokens": 227993424.0, "reward": 0.10249999910593033, "reward_std": 0.014142136089503765, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472135752439499, "step": 1242 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3655.0, "completions/mean_length": 2038.800048828125, "completions/mean_terminated_length": 1483.6826171875, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "epoch": 0.4821567106283941, "grad_norm": 0.02715626358985901, "kl": 0.30615234375, "learning_rate": 2.9926862262343782e-06, "loss": 0.0031, "num_tokens": 228196624.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1243 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3727.0, "completions/mean_length": 1556.550048828125, "completions/mean_terminated_length": 1193.771484375, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.48254460822342904, "grad_norm": 0.02964499033987522, "kl": 0.36572265625, "learning_rate": 2.9898750971193352e-06, "loss": 0.0037, "num_tokens": 228372852.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1244 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3222.0, "completions/mean_length": 1430.112548828125, "completions/mean_terminated_length": 1133.90283203125, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.4829325058184639, "grad_norm": 1.4048491716384888, "kl": 4.5380859375, "learning_rate": 2.9870635892087717e-06, "loss": 0.1298, "num_tokens": 228528829.0, "reward": 0.09749999642372131, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1245 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3772.0, "completions/mean_length": 1301.3250732421875, "completions/mean_terminated_length": 1192.4415283203125, "completions/min_length": 218.0, "completions/min_terminated_length": 218.0, "epoch": 0.48332040341349886, "grad_norm": 0.18537817895412445, "kl": 0.3623046875, "learning_rate": 2.984251706942443e-06, "loss": 0.0134, "num_tokens": 228682103.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1246 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3499.0, "completions/mean_length": 1164.1624755859375, "completions/mean_terminated_length": 1088.9871826171875, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.48370830100853374, "grad_norm": 0.17967569828033447, "kl": 0.39111328125, "learning_rate": 2.981439454760694e-06, "loss": -0.0168, "num_tokens": 228826788.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1247 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3877.0, "completions/mean_length": 1115.1500244140625, "completions/mean_terminated_length": 1038.718017578125, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.4840961986035687, "grad_norm": 0.16516458988189697, "kl": 0.39404296875, "learning_rate": 2.9786268371044553e-06, "loss": 0.024, "num_tokens": 228968320.0, "reward": 0.10125000774860382, "reward_std": 0.0035355340223759413, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1248 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3216.0, "completions/mean_length": 1139.8250732421875, "completions/mean_terminated_length": 1024.6492919921875, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.48448409619860355, "grad_norm": 1.8460294008255005, "kl": 7.89990234375, "learning_rate": 2.9758138584152334e-06, "loss": 0.1802, "num_tokens": 229113114.0, "reward": 0.09875000268220901, "reward_std": 0.029377836734056473, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472135752439499, "step": 1249 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3632.0, "completions/mean_length": 865.0875244140625, "completions/mean_terminated_length": 824.1898803710938, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "epoch": 0.4848719937936385, "grad_norm": 0.7225891351699829, "kl": 3.5908203125, "learning_rate": 2.9730005231351077e-06, "loss": 0.0567, "num_tokens": 229233817.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1250 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3976.0, "completions/mean_length": 1133.25, "completions/mean_terminated_length": 893.0270385742188, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.48525989138867337, "grad_norm": 21.460229873657227, "kl": 25.97509765625, "learning_rate": 2.9701868357067166e-06, "loss": 0.5233, "num_tokens": 229375589.0, "reward": 0.10249999910593033, "reward_std": 0.042103853076696396, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.09192044287919998, "step": 1251 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3018.0, "completions/mean_length": 794.3875122070312, "completions/mean_terminated_length": 752.594970703125, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.4856477889837083, "grad_norm": 4.915255069732666, "kl": 5.63525390625, "learning_rate": 2.9673728005732564e-06, "loss": 0.0952, "num_tokens": 229480508.0, "reward": 0.0962500050663948, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1252 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2208.0, "completions/mean_length": 892.7750244140625, "completions/mean_terminated_length": 810.6410522460938, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.4860356865787432, "grad_norm": 1.8891613483428955, "kl": 9.7099609375, "learning_rate": 2.9645584221784746e-06, "loss": 0.1407, "num_tokens": 229600346.0, "reward": 0.09624999761581421, "reward_std": 0.008164633996784687, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1253 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2804.0, "completions/mean_length": 785.9249877929688, "completions/mean_terminated_length": 701.05126953125, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.4864235841737781, "grad_norm": 12.244430541992188, "kl": 3.67626953125, "learning_rate": 2.9617437049666582e-06, "loss": 0.0918, "num_tokens": 229709228.0, "reward": 0.11000000685453415, "reward_std": 0.04502127692103386, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.08282969892024994, "step": 1254 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2441.0, "completions/mean_length": 1035.875, "completions/mean_terminated_length": 742.4383544921875, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.48681148176881306, "grad_norm": 3.8908462524414062, "kl": 23.29052734375, "learning_rate": 2.958928653382628e-06, "loss": 0.4347, "num_tokens": 229849122.0, "reward": 0.10375000536441803, "reward_std": 0.048903051763772964, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.07647287845611572, "step": 1255 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3005.0, "completions/mean_length": 971.6500244140625, "completions/mean_terminated_length": 763.3600463867188, "completions/min_length": 101.0, "completions/min_terminated_length": 101.0, "epoch": 0.48719937936384794, "grad_norm": 3.099421739578247, "kl": 15.041015625, "learning_rate": 2.9561132718717366e-06, "loss": 0.2214, "num_tokens": 229976206.0, "reward": 0.1262500137090683, "reward_std": 0.07998128235340118, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.12231935560703278, "step": 1256 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3627.0, "completions/mean_length": 1049.0250244140625, "completions/mean_terminated_length": 613.7428588867188, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.4875872769588829, "grad_norm": 4.320045471191406, "kl": 20.140625, "learning_rate": 2.9532975648798553e-06, "loss": 0.406, "num_tokens": 230109816.0, "reward": 0.12124999612569809, "reward_std": 0.1074451357126236, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.1529250591993332, "step": 1257 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2957.0, "completions/mean_length": 1168.5250244140625, "completions/mean_terminated_length": 701.8261108398438, "completions/min_length": 98.0, "completions/min_terminated_length": 98.0, "epoch": 0.48797517455391776, "grad_norm": 12.218875885009766, "kl": 22.859375, "learning_rate": 2.9504815368533695e-06, "loss": 0.5764, "num_tokens": 230247170.0, "reward": 0.1275000125169754, "reward_std": 0.09229806810617447, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.04749999940395355, "rewards/unicoder_reward_fn/std": 0.17208696901798248, "step": 1258 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1797.0, "completions/mean_length": 1372.125, "completions/mean_terminated_length": 581.3225708007812, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.4883630721489527, "grad_norm": 6.446746349334717, "kl": 27.46875, "learning_rate": 2.947665192239173e-06, "loss": 0.5026, "num_tokens": 230402236.0, "reward": 0.13500000536441803, "reward_std": 0.13260193169116974, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.18308192491531372, "step": 1259 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1899.0, "completions/mean_length": 1523.2125244140625, "completions/mean_terminated_length": 485.0701904296875, "completions/min_length": 67.0, "completions/min_terminated_length": 67.0, "epoch": 0.4887509697439876, "grad_norm": 10.15000057220459, "kl": 33.375, "learning_rate": 2.9448485354846563e-06, "loss": 0.7625, "num_tokens": 230572517.0, "reward": 0.22750000655651093, "reward_std": 0.2121163159608841, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.2918188273906708, "step": 1260 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 1301.0, "completions/mean_length": 1462.8250732421875, "completions/mean_terminated_length": 464.03448486328125, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.4891388673390225, "grad_norm": 8.852174758911133, "kl": 25.15625, "learning_rate": 2.9420315710377063e-06, "loss": 0.621, "num_tokens": 230740431.0, "reward": 0.19375000894069672, "reward_std": 0.1931668072938919, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.2830216884613037, "step": 1261 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1569.0, "completions/mean_length": 1522.5250244140625, "completions/mean_terminated_length": 484.1052551269531, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.4895267649340574, "grad_norm": 8.698474884033203, "kl": 25.703125, "learning_rate": 2.9392143033466936e-06, "loss": 0.6317, "num_tokens": 230896689.0, "reward": 0.2474999874830246, "reward_std": 0.2056581974029541, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.3137260675430298, "step": 1262 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1725.0, "completions/mean_length": 1576.9124755859375, "completions/mean_terminated_length": 431.8727111816406, "completions/min_length": 75.0, "completions/min_terminated_length": 75.0, "epoch": 0.4899146625290923, "grad_norm": 13.060507774353027, "kl": 48.125, "learning_rate": 2.936396736860469e-06, "loss": 0.8649, "num_tokens": 231070546.0, "reward": 0.23749999701976776, "reward_std": 0.2210523635149002, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.2777736783027649, "step": 1263 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2101.0, "completions/mean_length": 1982.1375732421875, "completions/mean_terminated_length": 338.022216796875, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.4903025601241272, "grad_norm": 27.558944702148438, "kl": 67.25, "learning_rate": 2.933578876028355e-06, "loss": 1.0922, "num_tokens": 231281037.0, "reward": 0.19499999284744263, "reward_std": 0.14907896518707275, "rewards/format_reward/mean": 0.4000000059604645, "rewards/format_reward/std": 0.49298882484436035, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.26428791880607605, "step": 1264 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1291.0, "completions/mean_length": 2391.262451171875, "completions/mean_terminated_length": 410.0810852050781, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.49069045771916214, "grad_norm": 22.304277420043945, "kl": 55.9375, "learning_rate": 2.9307607253001373e-06, "loss": 0.8917, "num_tokens": 231522498.0, "reward": 0.2512499988079071, "reward_std": 0.2510490119457245, "rewards/format_reward/mean": 0.38749998807907104, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.21250000596046448, "rewards/unicoder_reward_fn/std": 0.3357873857021332, "step": 1265 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1022.0, "completions/mean_length": 1311.7000732421875, "completions/mean_terminated_length": 320.677978515625, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.4910783553141971, "grad_norm": 11.334142684936523, "kl": 21.59375, "learning_rate": 2.9279422891260613e-06, "loss": 0.7189, "num_tokens": 231672234.0, "reward": 0.3212500512599945, "reward_std": 0.28281065821647644, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.3447618782520294, "step": 1266 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 2374.0, "completions/mean_length": 2367.462646484375, "completions/mean_terminated_length": 456.97369384765625, "completions/min_length": 58.0, "completions/min_terminated_length": 58.0, "epoch": 0.49146625290923196, "grad_norm": 3.811044931411743, "kl": 29.125, "learning_rate": 2.925123571956822e-06, "loss": 0.6668, "num_tokens": 231907927.0, "reward": 0.16875000298023224, "reward_std": 0.17950479686260223, "rewards/format_reward/mean": 0.38749998807907104, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.22747932374477386, "step": 1267 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2117.0, "completions/mean_length": 1718.2750244140625, "completions/mean_terminated_length": 366.2353210449219, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.4918541505042669, "grad_norm": 10.236571311950684, "kl": 24.34375, "learning_rate": 2.92230457824356e-06, "loss": 0.6027, "num_tokens": 232092957.0, "reward": 0.20375001430511475, "reward_std": 0.1950017362833023, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.27369946241378784, "step": 1268 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 846.0, "completions/mean_length": 1940.612548828125, "completions/mean_terminated_length": 264.20001220703125, "completions/min_length": 76.0, "completions/min_terminated_length": 76.0, "epoch": 0.4922420480993018, "grad_norm": 14.796316146850586, "kl": 34.0625, "learning_rate": 2.9194853124378494e-06, "loss": 0.6206, "num_tokens": 232299678.0, "reward": 0.2787500023841858, "reward_std": 0.24097982048988342, "rewards/format_reward/mean": 0.3375000059604645, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.24500000476837158, "rewards/unicoder_reward_fn/std": 0.30352362990379333, "step": 1269 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1451.0, "completions/mean_length": 2357.387451171875, "completions/mean_terminated_length": 336.8378601074219, "completions/min_length": 50.0, "completions/min_terminated_length": 50.0, "epoch": 0.4926299456943367, "grad_norm": 8.187853813171387, "kl": 38.375, "learning_rate": 2.916665778991696e-06, "loss": 0.7785, "num_tokens": 232531141.0, "reward": 0.13625000417232513, "reward_std": 0.179449200630188, "rewards/format_reward/mean": 0.36250001192092896, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.2505689561367035, "step": 1270 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 921.0, "completions/mean_length": 2434.60009765625, "completions/mean_terminated_length": 404.0, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.4930178432893716, "grad_norm": 2.250920534133911, "kl": 34.75, "learning_rate": 2.913845982357528e-06, "loss": 0.7003, "num_tokens": 232772397.0, "reward": 0.12250000238418579, "reward_std": 0.14416757225990295, "rewards/format_reward/mean": 0.3499999940395355, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.19576208293437958, "step": 1271 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 894.0, "completions/mean_length": 2443.887451171875, "completions/mean_terminated_length": 319.74285888671875, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.49340574088440653, "grad_norm": 7.6935601234436035, "kl": 33.75, "learning_rate": 2.91102592698819e-06, "loss": 0.6076, "num_tokens": 233016172.0, "reward": 0.14875000715255737, "reward_std": 0.12912200391292572, "rewards/format_reward/mean": 0.2874999940395355, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.20524761080741882, "step": 1272 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 971.0, "completions/mean_length": 2486.10009765625, "completions/mean_terminated_length": 308.0, "completions/min_length": 56.0, "completions/min_terminated_length": 56.0, "epoch": 0.4937936384794414, "grad_norm": 12.581275939941406, "kl": 27.71875, "learning_rate": 2.9082056173369344e-06, "loss": 0.7227, "num_tokens": 233261436.0, "reward": 0.1287500113248825, "reward_std": 0.1554878205060959, "rewards/format_reward/mean": 0.26249998807907104, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.20865754783153534, "step": 1273 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 830.0, "completions/mean_length": 2084.650146484375, "completions/mean_terminated_length": 353.9534912109375, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.49418153607447635, "grad_norm": 18.60162353515625, "kl": 20.21875, "learning_rate": 2.905385057857414e-06, "loss": 0.6127, "num_tokens": 233472000.0, "reward": 0.16500000655651093, "reward_std": 0.16575920581817627, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.20276570320129395, "step": 1274 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1046.0, "completions/mean_length": 2392.150146484375, "completions/mean_terminated_length": 412.0, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.4945694336695112, "grad_norm": 13.28848648071289, "kl": 26.75, "learning_rate": 2.9025642530036776e-06, "loss": 0.6919, "num_tokens": 233708836.0, "reward": 0.1575000137090683, "reward_std": 0.1750430315732956, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.21012353897094727, "step": 1275 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 981.0, "completions/mean_length": 2175.28759765625, "completions/mean_terminated_length": 348.2682800292969, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.49495733126454616, "grad_norm": 9.976146697998047, "kl": 28.75, "learning_rate": 2.8997432072301614e-06, "loss": 0.6266, "num_tokens": 233928035.0, "reward": 0.15375001728534698, "reward_std": 0.20489095151424408, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.27540531754493713, "step": 1276 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 1435.0, "completions/mean_length": 2596.949951171875, "completions/mean_terminated_length": 348.375, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.49534522885958104, "grad_norm": 10.06446361541748, "kl": 40.6875, "learning_rate": 2.896921924991681e-06, "loss": 0.823, "num_tokens": 234184135.0, "reward": 0.13375000655651093, "reward_std": 0.14429230988025665, "rewards/format_reward/mean": 0.3375000059604645, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.16838772594928741, "step": 1277 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 773.0, "completions/mean_length": 3023.212646484375, "completions/mean_terminated_length": 364.5652160644531, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.495733126454616, "grad_norm": 18.356447219848633, "kl": 51.25, "learning_rate": 2.894100410743427e-06, "loss": 0.8182, "num_tokens": 234476896.0, "reward": 0.07375000417232513, "reward_std": 0.12381631135940552, "rewards/format_reward/mean": 0.23749999701976776, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.13312219083309174, "step": 1278 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 1478.0, "completions/mean_length": 2127.91259765625, "completions/mean_terminated_length": 347.26190185546875, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.4961210240496509, "grad_norm": 9.742002487182617, "kl": 38.71875, "learning_rate": 2.8912786689409556e-06, "loss": 0.6174, "num_tokens": 234700809.0, "reward": 0.24250002205371857, "reward_std": 0.25417858362197876, "rewards/format_reward/mean": 0.4749999940395355, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.19499999284744263, "rewards/unicoder_reward_fn/std": 0.2827979624271393, "step": 1279 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 835.0, "completions/mean_length": 2277.375, "completions/mean_terminated_length": 267.3157958984375, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.4965089216446858, "grad_norm": 4.278537750244141, "kl": 36.59375, "learning_rate": 2.8884567040401816e-06, "loss": 0.751, "num_tokens": 234925255.0, "reward": 0.1850000023841858, "reward_std": 0.18040205538272858, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.2912108898162842, "step": 1280 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 1323.0, "completions/mean_length": 3078.03759765625, "completions/mean_terminated_length": 394.3182067871094, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.49689681923972073, "grad_norm": 2.8901636600494385, "kl": 28.46875, "learning_rate": 2.8856345204973733e-06, "loss": 0.5716, "num_tokens": 235217146.0, "reward": 0.10375000536441803, "reward_std": 0.11243756115436554, "rewards/format_reward/mean": 0.23749999701976776, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.20769988000392914, "step": 1281 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 811.0, "completions/mean_length": 2318.03759765625, "completions/mean_terminated_length": 352.9210510253906, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.4972847168347556, "grad_norm": 13.306024551391602, "kl": 16.0, "learning_rate": 2.8828121227691435e-06, "loss": 0.4876, "num_tokens": 235459373.0, "reward": 0.13500002026557922, "reward_std": 0.14247643947601318, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.09000000357627869, "rewards/unicoder_reward_fn/std": 0.17689301073551178, "step": 1282 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 855.0, "completions/mean_length": 3018.925048828125, "completions/mean_terminated_length": 349.6521911621094, "completions/min_length": 33.0, "completions/min_terminated_length": 33.0, "epoch": 0.49767261442979055, "grad_norm": 10.498726844787598, "kl": 17.796875, "learning_rate": 2.8799895153124435e-06, "loss": 0.4907, "num_tokens": 235747295.0, "reward": 0.0650000050663948, "reward_std": 0.08958332240581512, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.1353616863489151, "step": 1283 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 648.0, "completions/mean_length": 1645.6875, "completions/mean_terminated_length": 326.2884826660156, "completions/min_length": 50.0, "completions/min_terminated_length": 50.0, "epoch": 0.49806051202482543, "grad_norm": 20.152179718017578, "kl": 18.125, "learning_rate": 2.8771667025845555e-06, "loss": 0.6476, "num_tokens": 235923110.0, "reward": 0.20500002801418304, "reward_std": 0.21132779121398926, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.24955657124519348, "step": 1284 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1146.0, "completions/mean_length": 2553.112548828125, "completions/mean_terminated_length": 355.66668701171875, "completions/min_length": 72.0, "completions/min_terminated_length": 72.0, "epoch": 0.49844840961986037, "grad_norm": 8.330194473266602, "kl": 40.25, "learning_rate": 2.8743436890430866e-06, "loss": 0.7826, "num_tokens": 236178927.0, "reward": 0.1250000149011612, "reward_std": 0.17417804896831512, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.08750000596046448, "rewards/unicoder_reward_fn/std": 0.2201121747493744, "step": 1285 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 761.0, "completions/mean_length": 3069.837646484375, "completions/mean_terminated_length": 364.5, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.49883630721489525, "grad_norm": 18.963008880615234, "kl": 48.8125, "learning_rate": 2.87152047914596e-06, "loss": 0.801, "num_tokens": 236477154.0, "reward": 0.08875000476837158, "reward_std": 0.111746646463871, "rewards/format_reward/mean": 0.26249998807907104, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.16410478949546814, "step": 1286 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 810.0, "completions/mean_length": 3485.087646484375, "completions/mean_terminated_length": 336.5384826660156, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.4992242048099302, "grad_norm": 23.623929977416992, "kl": 47.0, "learning_rate": 2.86869707735141e-06, "loss": 0.6891, "num_tokens": 236811961.0, "reward": 0.06750000268220901, "reward_std": 0.05654168128967285, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.126265749335289, "step": 1287 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 961.0, "completions/mean_length": 2893.91259765625, "completions/mean_terminated_length": 397.2692565917969, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.49961210240496506, "grad_norm": 2.408736228942871, "kl": 34.9375, "learning_rate": 2.865873488117973e-06, "loss": 0.7261, "num_tokens": 237089138.0, "reward": 0.125, "reward_std": 0.1875668615102768, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.22713126242160797, "step": 1288 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 3131.0, "completions/mean_length": 3123.925048828125, "completions/mean_terminated_length": 561.1818237304688, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.5, "grad_norm": 9.79652214050293, "kl": 21.84375, "learning_rate": 2.8630497159044817e-06, "loss": 0.5441, "num_tokens": 237383156.0, "reward": 0.09250000864267349, "reward_std": 0.1411263793706894, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.06499999761581421, "rewards/unicoder_reward_fn/std": 0.17938007414340973, "step": 1289 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 397.0, "completions/mean_length": 3566.137451171875, "completions/mean_terminated_length": 242.45455932617188, "completions/min_length": 92.0, "completions/min_terminated_length": 92.0, "epoch": 0.5003878975950349, "grad_norm": 5.606987476348877, "kl": 21.25, "learning_rate": 2.8602257651700583e-06, "loss": 0.3835, "num_tokens": 237721319.0, "reward": 0.046250004321336746, "reward_std": 0.09312813729047775, "rewards/format_reward/mean": 0.13750000298023224, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.032499998807907104, "rewards/unicoder_reward_fn/std": 0.1507575809955597, "step": 1290 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.825, "completions/max_length": 4096.0, "completions/max_terminated_length": 496.0, "completions/mean_length": 3431.125, "completions/mean_terminated_length": 296.71429443359375, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 0.5007757951900699, "grad_norm": 3.823667049407959, "kl": 13.78125, "learning_rate": 2.857401640374107e-06, "loss": 0.3406, "num_tokens": 238042337.0, "reward": 0.05000000074505806, "reward_std": 0.06668306887149811, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.09426598995923996, "step": 1291 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 585.0, "completions/mean_length": 3720.425048828125, "completions/mean_terminated_length": 340.25, "completions/min_length": 152.0, "completions/min_terminated_length": 152.0, "epoch": 0.5011636927851048, "grad_norm": 1.6298160552978516, "kl": 14.53125, "learning_rate": 2.854577345976305e-06, "loss": 0.3292, "num_tokens": 238390683.0, "reward": 0.04375000298023224, "reward_std": 0.09568715840578079, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434914350509644, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.1450447291135788, "step": 1292 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 924.0, "completions/mean_length": 3490.41259765625, "completions/mean_terminated_length": 369.3077087402344, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.5015515903801396, "grad_norm": 4.240476608276367, "kl": 17.34375, "learning_rate": 2.8517528864365984e-06, "loss": 0.4171, "num_tokens": 238720084.0, "reward": 0.07500000298023224, "reward_std": 0.11938637495040894, "rewards/format_reward/mean": 0.125, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.0625000074505806, "rewards/unicoder_reward_fn/std": 0.20523220300674438, "step": 1293 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.85, "completions/max_length": 4096.0, "completions/max_terminated_length": 1048.0, "completions/mean_length": 3541.47509765625, "completions/mean_terminated_length": 399.16668701171875, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.5019394879751745, "grad_norm": 2.518167734146118, "kl": 27.0, "learning_rate": 2.848928266215196e-06, "loss": 0.5173, "num_tokens": 239060026.0, "reward": 0.04750000312924385, "reward_std": 0.09582620114088058, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.032499998807907104, "rewards/unicoder_reward_fn/std": 0.16971680521965027, "step": 1294 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.825, "completions/max_length": 4096.0, "completions/max_terminated_length": 1182.0, "completions/mean_length": 3445.324951171875, "completions/mean_terminated_length": 377.8571472167969, "completions/min_length": 58.0, "completions/min_terminated_length": 58.0, "epoch": 0.5023273855702095, "grad_norm": 8.60047435760498, "kl": 30.25, "learning_rate": 2.846103489772559e-06, "loss": 0.5018, "num_tokens": 239391372.0, "reward": 0.05125000700354576, "reward_std": 0.10949565470218658, "rewards/format_reward/mean": 0.13750000298023224, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.1353616863489151, "step": 1295 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1651.0, "completions/mean_length": 3311.775146484375, "completions/mean_terminated_length": 405.5294189453125, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.5027152831652444, "grad_norm": 4.015797138214111, "kl": 27.40625, "learning_rate": 2.8432785615693925e-06, "loss": 0.59, "num_tokens": 239702130.0, "reward": 0.046250004321336746, "reward_std": 0.08416152000427246, "rewards/format_reward/mean": 0.1875, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.027500003576278687, "rewards/unicoder_reward_fn/std": 0.10430602729320526, "step": 1296 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.825, "completions/max_length": 4096.0, "completions/max_terminated_length": 912.0, "completions/mean_length": 3446.78759765625, "completions/mean_terminated_length": 386.21429443359375, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.5031031807602793, "grad_norm": 8.243019104003906, "kl": 30.875, "learning_rate": 2.8404534860666448e-06, "loss": 0.5201, "num_tokens": 240024153.0, "reward": 0.057500001043081284, "reward_std": 0.11878859996795654, "rewards/format_reward/mean": 0.17499999701976776, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.12887045741081238, "step": 1297 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 753.0, "completions/mean_length": 3337.087646484375, "completions/mean_terminated_length": 301.4375, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.5034910783553141, "grad_norm": 3.218961477279663, "kl": 26.28125, "learning_rate": 2.8376282677254976e-06, "loss": 0.5221, "num_tokens": 240338520.0, "reward": 0.08000000566244125, "reward_std": 0.1250842660665512, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.20964105427265167, "step": 1298 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 732.0, "completions/mean_length": 3818.887451171875, "completions/mean_terminated_length": 401.16668701171875, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "epoch": 0.5038789759503491, "grad_norm": 9.655828475952148, "kl": 20.6875, "learning_rate": 2.834802911007353e-06, "loss": 0.325, "num_tokens": 240695039.0, "reward": 0.02500000037252903, "reward_std": 0.04497981071472168, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.1111220121383667, "step": 1299 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.85, "completions/max_length": 4096.0, "completions/max_terminated_length": 597.0, "completions/mean_length": 3529.699951171875, "completions/mean_terminated_length": 320.66668701171875, "completions/min_length": 196.0, "completions/min_terminated_length": 196.0, "epoch": 0.504266873545384, "grad_norm": 8.401337623596191, "kl": 10.765625, "learning_rate": 2.8319774203738376e-06, "loss": 0.3165, "num_tokens": 241022687.0, "reward": 0.07125000655651093, "reward_std": 0.12764286994934082, "rewards/format_reward/mean": 0.13750000298023224, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.19144684076309204, "step": 1300 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.85, "completions/max_length": 4096.0, "completions/max_terminated_length": 799.0, "completions/mean_length": 3527.6875, "completions/mean_terminated_length": 307.25, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.5046547711404189, "grad_norm": 10.577078819274902, "kl": 8.96875, "learning_rate": 2.829151800286785e-06, "loss": 0.3098, "num_tokens": 241349830.0, "reward": 0.08125000447034836, "reward_std": 0.131562277674675, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.16868816316127777, "step": 1301 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2241.0, "completions/mean_length": 3600.175048828125, "completions/mean_terminated_length": 490.0, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.5050426687354539, "grad_norm": 11.54315185546875, "kl": 9.796875, "learning_rate": 2.8263260552082372e-06, "loss": 0.3181, "num_tokens": 241682804.0, "reward": 0.03999999910593033, "reward_std": 0.07132973521947861, "rewards/format_reward/mean": 0.125, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.09933321923017502, "step": 1302 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 552.0, "completions/mean_length": 3626.8125, "completions/mean_terminated_length": 342.5, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.5054305663304888, "grad_norm": 4.471291542053223, "kl": 15.828125, "learning_rate": 2.823500189600429e-06, "loss": 0.3719, "num_tokens": 242028053.0, "reward": 0.06499999761581421, "reward_std": 0.13213156163692474, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.20617063343524933, "step": 1303 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 627.0, "completions/mean_length": 3913.212646484375, "completions/mean_terminated_length": 440.25, "completions/min_length": 365.0, "completions/min_terminated_length": 365.0, "epoch": 0.5058184639255237, "grad_norm": 18.849712371826172, "kl": 27.59375, "learning_rate": 2.820674207925789e-06, "loss": 0.3359, "num_tokens": 242395766.0, "reward": 0.01125000137835741, "reward_std": 0.028030434623360634, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 1304 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8125, "completions/max_length": 4096.0, "completions/max_terminated_length": 944.0, "completions/mean_length": 3379.949951171875, "completions/mean_terminated_length": 277.0666809082031, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.5062063615205585, "grad_norm": 7.300143718719482, "kl": 36.125, "learning_rate": 2.817848114646928e-06, "loss": 0.658, "num_tokens": 242706002.0, "reward": 0.10625000298023224, "reward_std": 0.19997137784957886, "rewards/format_reward/mean": 0.1875, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.2313280552625656, "step": 1305 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 857.0, "completions/mean_length": 3487.150146484375, "completions/mean_terminated_length": 349.23077392578125, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.5065942591155935, "grad_norm": 12.11135482788086, "kl": 30.40625, "learning_rate": 2.815021914226633e-06, "loss": 0.476, "num_tokens": 243028814.0, "reward": 0.03875000402331352, "reward_std": 0.05985304340720177, "rewards/format_reward/mean": 0.16249999403953552, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.02250000089406967, "rewards/unicoder_reward_fn/std": 0.10059948265552521, "step": 1306 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 744.0, "completions/mean_length": 3685.85009765625, "completions/mean_terminated_length": 450.22222900390625, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.5069821567106284, "grad_norm": 7.343531608581543, "kl": 26.75, "learning_rate": 2.81219561112786e-06, "loss": 0.4421, "num_tokens": 243373594.0, "reward": 0.036250002682209015, "reward_std": 0.07592368125915527, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.10732654482126236, "step": 1307 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1100.0, "completions/mean_length": 3682.525146484375, "completions/mean_terminated_length": 420.6666564941406, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.5073700543056633, "grad_norm": 5.6435441970825195, "kl": 15.9375, "learning_rate": 2.809369209813728e-06, "loss": 0.2636, "num_tokens": 243717916.0, "reward": 0.055000003427267075, "reward_std": 0.09679190069437027, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.1799437254667282, "step": 1308 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1075.0, "completions/mean_length": 3043.60009765625, "completions/mean_terminated_length": 435.478271484375, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.5077579519006982, "grad_norm": 13.667505264282227, "kl": 7.859375, "learning_rate": 2.80654271474751e-06, "loss": 0.3235, "num_tokens": 244012780.0, "reward": 0.10250001400709152, "reward_std": 0.1690264344215393, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.21435804665088654, "step": 1309 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 1500.0, "completions/mean_length": 2605.962646484375, "completions/mean_terminated_length": 370.90625, "completions/min_length": 72.0, "completions/min_terminated_length": 72.0, "epoch": 0.5081458494957332, "grad_norm": 16.763399124145508, "kl": 5.921875, "learning_rate": 2.803716130392627e-06, "loss": 0.2834, "num_tokens": 244270569.0, "reward": 0.10375000536441803, "reward_std": 0.13197766244411469, "rewards/format_reward/mean": 0.36250001192092896, "rewards/format_reward/std": 0.48375511169433594, "rewards/unicoder_reward_fn/mean": 0.06750001013278961, "rewards/unicoder_reward_fn/std": 0.1620868742465973, "step": 1310 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1891.0, "completions/mean_length": 3052.837646484375, "completions/mean_terminated_length": 467.60870361328125, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.508533747090768, "grad_norm": 22.102375030517578, "kl": 5.65625, "learning_rate": 2.8008894612126425e-06, "loss": 0.368, "num_tokens": 244560228.0, "reward": 0.0962500050663948, "reward_std": 0.13552942872047424, "rewards/format_reward/mean": 0.26249998807907104, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.19119875133037567, "step": 1311 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1260.0, "completions/mean_length": 3133.887451171875, "completions/mean_terminated_length": 430.8095397949219, "completions/min_length": 131.0, "completions/min_terminated_length": 131.0, "epoch": 0.5089216446858029, "grad_norm": 22.112380981445312, "kl": 5.703125, "learning_rate": 2.7980627116712523e-06, "loss": 0.3834, "num_tokens": 244853427.0, "reward": 0.04750000312924385, "reward_std": 0.0719771608710289, "rewards/format_reward/mean": 0.22499999403953552, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.09743547439575195, "step": 1312 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1278.0, "completions/mean_length": 3307.712646484375, "completions/mean_terminated_length": 386.4117736816406, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.5093095422808379, "grad_norm": 17.119770050048828, "kl": 10.015625, "learning_rate": 2.7952358862322796e-06, "loss": 0.418, "num_tokens": 245165140.0, "reward": 0.04374999925494194, "reward_std": 0.07210393995046616, "rewards/format_reward/mean": 0.21250000596046448, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.02250000089406967, "rewards/unicoder_reward_fn/std": 0.08415628224611282, "step": 1313 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1477.0, "completions/mean_length": 3050.775146484375, "completions/mean_terminated_length": 460.4347839355469, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.5096974398758728, "grad_norm": 9.98918628692627, "kl": 16.9375, "learning_rate": 2.792408989359669e-06, "loss": 0.4803, "num_tokens": 245467434.0, "reward": 0.08500000089406967, "reward_std": 0.12859635055065155, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.16958625614643097, "step": 1314 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1162.0, "completions/mean_length": 3403.637451171875, "completions/mean_terminated_length": 403.4000244140625, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.5100853374709077, "grad_norm": 3.3700151443481445, "kl": 28.71875, "learning_rate": 2.7895820255174743e-06, "loss": 0.53, "num_tokens": 245784949.0, "reward": 0.0962500050663948, "reward_std": 0.15301989018917084, "rewards/format_reward/mean": 0.1875, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.214638352394104, "step": 1315 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 730.0, "completions/mean_length": 3637.6875, "completions/mean_terminated_length": 429.5, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.5104732350659426, "grad_norm": 17.743789672851562, "kl": 39.0, "learning_rate": 2.78675499916986e-06, "loss": 0.5348, "num_tokens": 246122556.0, "reward": 0.06000000238418579, "reward_std": 0.08637119084596634, "rewards/format_reward/mean": 0.125, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.17208696901798248, "step": 1316 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7875, "completions/max_length": 4096.0, "completions/max_terminated_length": 956.0, "completions/mean_length": 3308.699951171875, "completions/mean_terminated_length": 391.058837890625, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.5108611326609775, "grad_norm": 6.342831611633301, "kl": 43.4375, "learning_rate": 2.7839279147810838e-06, "loss": 0.7804, "num_tokens": 246429244.0, "reward": 0.08625000715255737, "reward_std": 0.1401432603597641, "rewards/format_reward/mean": 0.21250000596046448, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.0650000050663948, "rewards/unicoder_reward_fn/std": 0.1677098125219345, "step": 1317 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1282.0, "completions/mean_length": 3688.375, "completions/mean_terminated_length": 472.6666564941406, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.5112490302560124, "grad_norm": 20.763343811035156, "kl": 40.5625, "learning_rate": 2.7811007768155e-06, "loss": 0.5142, "num_tokens": 246774194.0, "reward": 0.05750000476837158, "reward_std": 0.07387562841176987, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.18621821701526642, "step": 1318 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2057.0, "completions/mean_length": 3711.824951171875, "completions/mean_terminated_length": 681.1111450195312, "completions/min_length": 211.0, "completions/min_terminated_length": 211.0, "epoch": 0.5116369278510473, "grad_norm": 11.045652389526367, "kl": 35.5, "learning_rate": 2.778273589737544e-06, "loss": 0.5359, "num_tokens": 247122308.0, "reward": 0.036249998956918716, "reward_std": 0.06834294646978378, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.10250039398670197, "step": 1319 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 765.0, "completions/mean_length": 2587.199951171875, "completions/mean_terminated_length": 324.0, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.5120248254460822, "grad_norm": 16.372770309448242, "kl": 28.3125, "learning_rate": 2.775446358011731e-06, "loss": 0.771, "num_tokens": 247364436.0, "reward": 0.1275000125169754, "reward_std": 0.19167213141918182, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.09000000357627869, "rewards/unicoder_reward_fn/std": 0.19066838920116425, "step": 1320 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 544.0, "completions/mean_length": 3013.5625, "completions/mean_terminated_length": 331.0, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.5124127230411172, "grad_norm": 7.530948638916016, "kl": 25.1875, "learning_rate": 2.7726190861026442e-06, "loss": 0.6056, "num_tokens": 247653721.0, "reward": 0.12250000238418579, "reward_std": 0.19708846509456635, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.24229909479618073, "step": 1321 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 660.0, "completions/mean_length": 3016.275146484375, "completions/mean_terminated_length": 340.4347839355469, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.5128006206361521, "grad_norm": 2.2287075519561768, "kl": 24.125, "learning_rate": 2.7697917784749345e-06, "loss": 0.4648, "num_tokens": 247935903.0, "reward": 0.11250001192092896, "reward_std": 0.10836410522460938, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.2006950080394745, "step": 1322 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 1612.0, "completions/mean_length": 2901.837646484375, "completions/mean_terminated_length": 421.65386962890625, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.5131885182311869, "grad_norm": 14.559593200683594, "kl": 24.3125, "learning_rate": 2.7669644395933036e-06, "loss": 0.6622, "num_tokens": 248218274.0, "reward": 0.11250000447034836, "reward_std": 0.16561762988567352, "rewards/format_reward/mean": 0.30000001192092896, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.18195493519306183, "step": 1323 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.75, "completions/max_length": 4096.0, "completions/max_terminated_length": 2748.0, "completions/mean_length": 3224.97509765625, "completions/mean_terminated_length": 611.9000244140625, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.5135764158262219, "grad_norm": 10.837560653686523, "kl": 26.09375, "learning_rate": 2.7641370739225083e-06, "loss": 0.5827, "num_tokens": 248538416.0, "reward": 0.11874999850988388, "reward_std": 0.180306538939476, "rewards/format_reward/mean": 0.23749999701976776, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.0949999988079071, "rewards/unicoder_reward_fn/std": 0.23379100859165192, "step": 1324 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.75, "completions/max_length": 4096.0, "completions/max_terminated_length": 1070.0, "completions/mean_length": 3183.137451171875, "completions/mean_terminated_length": 444.5500183105469, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.5139643134212568, "grad_norm": 9.525189399719238, "kl": 31.1875, "learning_rate": 2.761309685927343e-06, "loss": 0.5925, "num_tokens": 248846987.0, "reward": 0.0925000011920929, "reward_std": 0.1327081024646759, "rewards/format_reward/mean": 0.22499999403953552, "rewards/format_reward/std": 0.4202169179916382, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.16565941274166107, "step": 1325 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 1109.0, "completions/mean_length": 2989.91259765625, "completions/mean_terminated_length": 409.04168701171875, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.5143522110162917, "grad_norm": 12.910548210144043, "kl": 32.8125, "learning_rate": 2.7584822800726395e-06, "loss": 0.669, "num_tokens": 249137892.0, "reward": 0.13750000298023224, "reward_std": 0.17257161438465118, "rewards/format_reward/mean": 0.30000001192092896, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.21095022559165955, "step": 1326 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 994.0, "completions/mean_length": 2008.0625, "completions/mean_terminated_length": 384.1111145019531, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.5147401086113266, "grad_norm": 19.800342559814453, "kl": 36.1875, "learning_rate": 2.755654860823259e-06, "loss": 0.8518, "num_tokens": 249343833.0, "reward": 0.2874999940395355, "reward_std": 0.2741328179836273, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.23499998450279236, "rewards/unicoder_reward_fn/std": 0.34534886479377747, "step": 1327 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 1481.0, "completions/mean_length": 2981.887451171875, "completions/mean_terminated_length": 382.29168701171875, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 0.5151280062063616, "grad_norm": 16.218647003173828, "kl": 41.6875, "learning_rate": 2.7528274326440814e-06, "loss": 0.8435, "num_tokens": 249628496.0, "reward": 0.048750001937150955, "reward_std": 0.08091932535171509, "rewards/format_reward/mean": 0.2874999940395355, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.06825121492147446, "step": 1328 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1171.0, "completions/mean_length": 2669.212646484375, "completions/mean_terminated_length": 413.9677429199219, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.5155159038013964, "grad_norm": 12.437304496765137, "kl": 43.625, "learning_rate": 2.7500000000000004e-06, "loss": 0.8093, "num_tokens": 249898881.0, "reward": 0.23625002801418304, "reward_std": 0.2639484107494354, "rewards/format_reward/mean": 0.38749998807907104, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.19750002026557922, "rewards/unicoder_reward_fn/std": 0.30604666471481323, "step": 1329 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2148.0, "completions/mean_length": 2750.650146484375, "completions/mean_terminated_length": 384.6896667480469, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.5159038013964313, "grad_norm": 11.381559371948242, "kl": 45.5, "learning_rate": 2.7471725673559205e-06, "loss": 0.8049, "num_tokens": 250174205.0, "reward": 0.09125000238418579, "reward_std": 0.11132951080799103, "rewards/format_reward/mean": 0.2874999940395355, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.12566280364990234, "step": 1330 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1082.0, "completions/mean_length": 2716.9375, "completions/mean_terminated_length": 418.5000305175781, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.5162916989914662, "grad_norm": 2.8225932121276855, "kl": 32.125, "learning_rate": 2.7443451391767416e-06, "loss": 0.676, "num_tokens": 250437840.0, "reward": 0.16500000655651093, "reward_std": 0.2265765517950058, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.27786481380462646, "step": 1331 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 886.0, "completions/mean_length": 2254.087646484375, "completions/mean_terminated_length": 317.71795654296875, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.5166795965865012, "grad_norm": 21.277685165405273, "kl": 24.03125, "learning_rate": 2.7415177199273612e-06, "loss": 0.7431, "num_tokens": 250650727.0, "reward": 0.2262500375509262, "reward_std": 0.2046312838792801, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.18000000715255737, "rewards/unicoder_reward_fn/std": 0.25969797372817993, "step": 1332 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 904.0, "completions/mean_length": 2029.987548828125, "completions/mean_terminated_length": 339.6136474609375, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.5170674941815361, "grad_norm": 21.011154174804688, "kl": 22.71875, "learning_rate": 2.738690314072657e-06, "loss": 0.7393, "num_tokens": 250854414.0, "reward": 0.23375001549720764, "reward_std": 0.2241944819688797, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.2517784833908081, "step": 1333 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 777.0, "completions/mean_length": 2735.337646484375, "completions/mean_terminated_length": 342.4482727050781, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.517455391776571, "grad_norm": 11.884465217590332, "kl": 25.75, "learning_rate": 2.735862926077493e-06, "loss": 0.6822, "num_tokens": 251125369.0, "reward": 0.17499999701976776, "reward_std": 0.2021322101354599, "rewards/format_reward/mean": 0.3499999940395355, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.14000001549720764, "rewards/unicoder_reward_fn/std": 0.26889365911483765, "step": 1334 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1584.0, "completions/mean_length": 2195.574951171875, "completions/mean_terminated_length": 387.8536376953125, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.5178432893716058, "grad_norm": 15.70750617980957, "kl": 32.15625, "learning_rate": 2.733035560406697e-06, "loss": 0.8924, "num_tokens": 251343775.0, "reward": 0.23500001430511475, "reward_std": 0.16276052594184875, "rewards/format_reward/mean": 0.4749999940395355, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.1875000149011612, "rewards/unicoder_reward_fn/std": 0.29051461815834045, "step": 1335 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1005.0, "completions/mean_length": 1904.125, "completions/mean_terminated_length": 365.14892578125, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.5182311869666408, "grad_norm": 2.6777002811431885, "kl": 49.75, "learning_rate": 2.730208221525066e-06, "loss": 1.0253, "num_tokens": 251538801.0, "reward": 0.18125002086162567, "reward_std": 0.17149856686592102, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.2592589259147644, "step": 1336 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 1548.0, "completions/mean_length": 2902.962646484375, "completions/mean_terminated_length": 425.1153869628906, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.5186190845616757, "grad_norm": 26.29150390625, "kl": 76.25, "learning_rate": 2.727380913897357e-06, "loss": 1.2642, "num_tokens": 251814398.0, "reward": 0.09750001132488251, "reward_std": 0.16234822571277618, "rewards/format_reward/mean": 0.30000001192092896, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.19858677685260773, "step": 1337 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 961.0, "completions/mean_length": 1988.7249755859375, "completions/mean_terminated_length": 349.73333740234375, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.5190069821567106, "grad_norm": 20.790817260742188, "kl": 73.625, "learning_rate": 2.7245536419882706e-06, "loss": 1.2145, "num_tokens": 252031400.0, "reward": 0.2600000202655792, "reward_std": 0.2509167194366455, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.20499999821186066, "rewards/unicoder_reward_fn/std": 0.2916127145290375, "step": 1338 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1443.0, "completions/mean_length": 1756.5125732421875, "completions/mean_terminated_length": 426.2156982421875, "completions/min_length": 101.0, "completions/min_terminated_length": 101.0, "epoch": 0.5193948797517456, "grad_norm": 13.884754180908203, "kl": 58.625, "learning_rate": 2.721726410262457e-06, "loss": 0.9575, "num_tokens": 252210673.0, "reward": 0.36625003814697266, "reward_std": 0.23507416248321533, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.30750003457069397, "rewards/unicoder_reward_fn/std": 0.34265345335006714, "step": 1339 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 762.0, "completions/mean_length": 2395.5, "completions/mean_terminated_length": 317.1111145019531, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.5197827773467805, "grad_norm": 12.129125595092773, "kl": 60.5, "learning_rate": 2.7188992231845013e-06, "loss": 1.0632, "num_tokens": 252450577.0, "reward": 0.16875001788139343, "reward_std": 0.14703164994716644, "rewards/format_reward/mean": 0.38749998807907104, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.2739306092262268, "step": 1340 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1324.0, "completions/mean_length": 1283.5250244140625, "completions/mean_terminated_length": 407.5081787109375, "completions/min_length": 62.0, "completions/min_terminated_length": 62.0, "epoch": 0.5201706749418153, "grad_norm": 25.171478271484375, "kl": 36.1875, "learning_rate": 2.716072085218917e-06, "loss": 1.1094, "num_tokens": 252599499.0, "reward": 0.3812499940395355, "reward_std": 0.25621774792671204, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.30750003457069397, "rewards/unicoder_reward_fn/std": 0.3290857970714569, "step": 1341 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1867.0, "completions/mean_length": 2793.050048828125, "completions/mean_terminated_length": 501.6551818847656, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.5205585725368502, "grad_norm": 1.277010202407837, "kl": 37.40625, "learning_rate": 2.713245000830141e-06, "loss": 0.7393, "num_tokens": 252863127.0, "reward": 0.10250001400709152, "reward_std": 0.15649108588695526, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.1594293713569641, "step": 1342 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 1397.0, "completions/mean_length": 1739.0250244140625, "completions/mean_terminated_length": 469.8846435546875, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.5209464701318852, "grad_norm": 15.842048645019531, "kl": 27.71875, "learning_rate": 2.710417974482526e-06, "loss": 0.7438, "num_tokens": 253048153.0, "reward": 0.2787500321865082, "reward_std": 0.19016431272029877, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.2175000160932541, "rewards/unicoder_reward_fn/std": 0.3055499196052551, "step": 1343 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2254.0, "completions/mean_length": 1859.4625244140625, "completions/mean_terminated_length": 444.51019287109375, "completions/min_length": 41.0, "completions/min_terminated_length": 41.0, "epoch": 0.5213343677269201, "grad_norm": 13.889349937438965, "kl": 25.9375, "learning_rate": 2.707591010640332e-06, "loss": 0.5945, "num_tokens": 253242534.0, "reward": 0.21124999225139618, "reward_std": 0.20256347954273224, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.2804946005344391, "step": 1344 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1353.0, "completions/mean_length": 1850.8375244140625, "completions/mean_terminated_length": 430.4285583496094, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.521722265321955, "grad_norm": 7.573080539703369, "kl": 31.9375, "learning_rate": 2.7047641137677215e-06, "loss": 0.7192, "num_tokens": 253434361.0, "reward": 0.3412500321865082, "reward_std": 0.1923058182001114, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.2800000011920929, "rewards/unicoder_reward_fn/std": 0.3566369116306305, "step": 1345 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1153.0, "completions/mean_length": 2067.237548828125, "completions/mean_terminated_length": 321.5581359863281, "completions/min_length": 80.0, "completions/min_terminated_length": 80.0, "epoch": 0.5221101629169899, "grad_norm": 5.143522262573242, "kl": 40.53125, "learning_rate": 2.701937288328748e-06, "loss": 0.7937, "num_tokens": 253644108.0, "reward": 0.20125000178813934, "reward_std": 0.12048287689685822, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.22890770435333252, "step": 1346 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 1137.0, "completions/mean_length": 1366.9000244140625, "completions/mean_terminated_length": 331.7241516113281, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.5224980605120249, "grad_norm": 3.7428383827209473, "kl": 51.8125, "learning_rate": 2.699110538787359e-06, "loss": 1.0391, "num_tokens": 253806060.0, "reward": 0.26250001788139343, "reward_std": 0.1850130558013916, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.1925000101327896, "rewards/unicoder_reward_fn/std": 0.2494424283504486, "step": 1347 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 932.0, "completions/mean_length": 1609.9625244140625, "completions/mean_terminated_length": 343.4905700683594, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.5228859581070597, "grad_norm": 17.918598175048828, "kl": 52.625, "learning_rate": 2.6962838696073736e-06, "loss": 0.8084, "num_tokens": 253977801.0, "reward": 0.27375003695487976, "reward_std": 0.161136195063591, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.26897603273391724, "step": 1348 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1374.0, "completions/mean_length": 2287.75, "completions/mean_terminated_length": 386.76922607421875, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.5232738557020946, "grad_norm": 23.422637939453125, "kl": 81.875, "learning_rate": 2.693457285252491e-06, "loss": 1.3799, "num_tokens": 254211693.0, "reward": 0.17750000953674316, "reward_std": 0.16651330888271332, "rewards/format_reward/mean": 0.4749999940395355, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.22070084512233734, "step": 1349 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 902.0, "completions/mean_length": 2514.550048828125, "completions/mean_terminated_length": 374.9411926269531, "completions/min_length": 98.0, "completions/min_terminated_length": 98.0, "epoch": 0.5236617532971296, "grad_norm": 23.69940757751465, "kl": 76.0625, "learning_rate": 2.6906307901862734e-06, "loss": 1.315, "num_tokens": 254460929.0, "reward": 0.1275000125169754, "reward_std": 0.13359364867210388, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.1303354948759079, "step": 1350 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1699.0, "completions/mean_length": 1813.3375244140625, "completions/mean_terminated_length": 369.2040710449219, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "epoch": 0.5240496508921645, "grad_norm": 15.569211959838867, "kl": 60.125, "learning_rate": 2.6878043888721404e-06, "loss": 1.04, "num_tokens": 254650388.0, "reward": 0.26374998688697815, "reward_std": 0.20143543183803558, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.20499999821186066, "rewards/unicoder_reward_fn/std": 0.2661968469619751, "step": 1351 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 761.0, "completions/mean_length": 2018.2000732421875, "completions/mean_terminated_length": 318.18182373046875, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.5244375484871994, "grad_norm": 12.449051856994629, "kl": 47.6875, "learning_rate": 2.6849780857733675e-06, "loss": 1.0074, "num_tokens": 254860852.0, "reward": 0.2874999940395355, "reward_std": 0.21906740963459015, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.23749999701976776, "rewards/unicoder_reward_fn/std": 0.3020499646663666, "step": 1352 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 951.0, "completions/mean_length": 2318.625, "completions/mean_terminated_length": 354.15789794921875, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.5248254460822342, "grad_norm": 8.835307121276855, "kl": 47.375, "learning_rate": 2.6821518853530727e-06, "loss": 1.1023, "num_tokens": 255096542.0, "reward": 0.1600000113248825, "reward_std": 0.1933038979768753, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.19297799468040466, "step": 1353 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2015.0, "completions/mean_length": 1924.2000732421875, "completions/mean_terminated_length": 399.3191223144531, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.5252133436772692, "grad_norm": 12.385886192321777, "kl": 39.625, "learning_rate": 2.679325792074212e-06, "loss": 0.9611, "num_tokens": 255308718.0, "reward": 0.20250001549720764, "reward_std": 0.16516248881816864, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.23652280867099762, "step": 1354 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 950.0, "completions/mean_length": 2655.8125, "completions/mean_terminated_length": 379.3870849609375, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.5256012412723041, "grad_norm": 6.904433250427246, "kl": 40.6875, "learning_rate": 2.676499810399572e-06, "loss": 0.8352, "num_tokens": 255574631.0, "reward": 0.13250000774860382, "reward_std": 0.15333232283592224, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.15903189778327942, "step": 1355 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 847.0, "completions/mean_length": 1519.6875, "completions/mean_terminated_length": 348.6363525390625, "completions/min_length": 52.0, "completions/min_terminated_length": 52.0, "epoch": 0.525989138867339, "grad_norm": 19.24216651916504, "kl": 38.875, "learning_rate": 2.673673944791764e-06, "loss": 0.9965, "num_tokens": 255750382.0, "reward": 0.21000002324581146, "reward_std": 0.18283693492412567, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.24644306302070618, "step": 1356 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 958.0, "completions/mean_length": 2686.925048828125, "completions/mean_terminated_length": 338.4666748046875, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.5263770364623739, "grad_norm": 10.810729026794434, "kl": 58.75, "learning_rate": 2.6708481997132152e-06, "loss": 1.0422, "num_tokens": 256010072.0, "reward": 0.16375000774860382, "reward_std": 0.18748866021633148, "rewards/format_reward/mean": 0.2874999940395355, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.24188081920146942, "step": 1357 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 637.0, "completions/mean_length": 2065.837646484375, "completions/mean_terminated_length": 229.02381896972656, "completions/min_length": 43.0, "completions/min_terminated_length": 43.0, "epoch": 0.5267649340574089, "grad_norm": 8.515369415283203, "kl": 62.875, "learning_rate": 2.6680225796261627e-06, "loss": 1.1974, "num_tokens": 256218003.0, "reward": 0.17000000178813934, "reward_std": 0.14801160991191864, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.195114403963089, "step": 1358 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.65, "completions/max_length": 4096.0, "completions/max_terminated_length": 609.0, "completions/mean_length": 2744.050048828125, "completions/mean_terminated_length": 233.2857208251953, "completions/min_length": 45.0, "completions/min_terminated_length": 45.0, "epoch": 0.5271528316524438, "grad_norm": 15.686687469482422, "kl": 71.75, "learning_rate": 2.6651970889926476e-06, "loss": 1.2702, "num_tokens": 256484199.0, "reward": 0.16750000417232513, "reward_std": 0.22428397834300995, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.14000001549720764, "rewards/unicoder_reward_fn/std": 0.236857071518898, "step": 1359 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 960.0, "completions/mean_length": 2908.85009765625, "completions/mean_terminated_length": 297.1199951171875, "completions/min_length": 33.0, "completions/min_terminated_length": 33.0, "epoch": 0.5275407292474786, "grad_norm": 15.064888000488281, "kl": 58.125, "learning_rate": 2.662371732274504e-06, "loss": 0.964, "num_tokens": 256765483.0, "reward": 0.10750000923871994, "reward_std": 0.12838111817836761, "rewards/format_reward/mean": 0.22499999403953552, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.17938008904457092, "step": 1360 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 861.0, "completions/mean_length": 2854.72509765625, "completions/mean_terminated_length": 276.69232177734375, "completions/min_length": 41.0, "completions/min_terminated_length": 41.0, "epoch": 0.5279286268425136, "grad_norm": 6.776430130004883, "kl": 50.625, "learning_rate": 2.6595465139333555e-06, "loss": 0.9461, "num_tokens": 257043261.0, "reward": 0.10750000923871994, "reward_std": 0.1407124549150467, "rewards/format_reward/mean": 0.22499999403953552, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.08499999344348907, "rewards/unicoder_reward_fn/std": 0.2032022327184677, "step": 1361 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1164.0, "completions/mean_length": 2553.112548828125, "completions/mean_terminated_length": 355.66668701171875, "completions/min_length": 41.0, "completions/min_terminated_length": 41.0, "epoch": 0.5283165244375485, "grad_norm": 12.096661567687988, "kl": 29.125, "learning_rate": 2.6567214384306082e-06, "loss": 0.7175, "num_tokens": 257290486.0, "reward": 0.18500001728534698, "reward_std": 0.20603306591510773, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.1599999964237213, "rewards/unicoder_reward_fn/std": 0.3075423240661621, "step": 1362 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 1405.0, "completions/mean_length": 2969.074951171875, "completions/mean_terminated_length": 339.5833435058594, "completions/min_length": 51.0, "completions/min_terminated_length": 51.0, "epoch": 0.5287044220325834, "grad_norm": 7.917473793029785, "kl": 23.375, "learning_rate": 2.6538965102274427e-06, "loss": 0.5355, "num_tokens": 257577924.0, "reward": 0.08000000566244125, "reward_std": 0.12230098247528076, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.12314446270465851, "step": 1363 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.75, "completions/max_length": 4096.0, "completions/max_terminated_length": 714.0, "completions/mean_length": 3165.875, "completions/mean_terminated_length": 375.5, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.5290923196276183, "grad_norm": 5.540017604827881, "kl": 25.84375, "learning_rate": 2.6510717337848046e-06, "loss": 0.5903, "num_tokens": 257872018.0, "reward": 0.10250001400709152, "reward_std": 0.1504027545452118, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.20792829990386963, "step": 1364 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 768.0, "completions/mean_length": 2607.824951171875, "completions/mean_terminated_length": 255.5483856201172, "completions/min_length": 47.0, "completions/min_terminated_length": 47.0, "epoch": 0.5294802172226533, "grad_norm": 11.235407829284668, "kl": 30.5625, "learning_rate": 2.648247113563402e-06, "loss": 0.7287, "num_tokens": 258134316.0, "reward": 0.125, "reward_std": 0.1773950606584549, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.22934965789318085, "step": 1365 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1180.0, "completions/mean_length": 2240.900146484375, "completions/mean_terminated_length": 290.6666564941406, "completions/min_length": 51.0, "completions/min_terminated_length": 51.0, "epoch": 0.5298681148176881, "grad_norm": 18.689979553222656, "kl": 34.375, "learning_rate": 2.645422654023696e-06, "loss": 0.9645, "num_tokens": 258356316.0, "reward": 0.1887500137090683, "reward_std": 0.23309972882270813, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.25605329871177673, "step": 1366 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1025.0, "completions/mean_length": 2840.612548828125, "completions/mean_terminated_length": 376.3333435058594, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.530256012412723, "grad_norm": 3.5819485187530518, "kl": 37.46875, "learning_rate": 2.6425983596258947e-06, "loss": 0.7874, "num_tokens": 258636725.0, "reward": 0.11250000447034836, "reward_std": 0.20645003020763397, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.20829325914382935, "step": 1367 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 798.0, "completions/mean_length": 2264.53759765625, "completions/mean_terminated_length": 339.1538391113281, "completions/min_length": 55.0, "completions/min_terminated_length": 55.0, "epoch": 0.5306439100077579, "grad_norm": 8.107589721679688, "kl": 47.75, "learning_rate": 2.639774234829943e-06, "loss": 0.807, "num_tokens": 258859544.0, "reward": 0.22500000894069672, "reward_std": 0.21325941383838654, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.29544001817703247, "step": 1368 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1225.0, "completions/mean_length": 2183.3125, "completions/mean_terminated_length": 363.92681884765625, "completions/min_length": 43.0, "completions/min_terminated_length": 43.0, "epoch": 0.5310318076027929, "grad_norm": 19.69232940673828, "kl": 59.0625, "learning_rate": 2.636950284095519e-06, "loss": 0.9304, "num_tokens": 259079769.0, "reward": 0.2225000113248825, "reward_std": 0.22755157947540283, "rewards/format_reward/mean": 0.4749999940395355, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.3188190162181854, "step": 1369 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 756.0, "completions/mean_length": 2201.362548828125, "completions/mean_terminated_length": 306.7250061035156, "completions/min_length": 49.0, "completions/min_terminated_length": 49.0, "epoch": 0.5314197051978278, "grad_norm": 13.149742126464844, "kl": 59.0, "learning_rate": 2.6341265118820286e-06, "loss": 0.9865, "num_tokens": 259310782.0, "reward": 0.15250001847743988, "reward_std": 0.17063985764980316, "rewards/format_reward/mean": 0.4000000059604645, "rewards/format_reward/std": 0.49298882484436035, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.22466571629047394, "step": 1370 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 795.0, "completions/mean_length": 2271.300048828125, "completions/mean_terminated_length": 254.5263214111328, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.5318076027928627, "grad_norm": 6.774125099182129, "kl": 63.9375, "learning_rate": 2.631302922648591e-06, "loss": 1.1746, "num_tokens": 259543510.0, "reward": 0.16625002026557922, "reward_std": 0.19079099595546722, "rewards/format_reward/mean": 0.38749998807907104, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.2164003551006317, "step": 1371 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 1968.0, "completions/mean_length": 2470.637451171875, "completions/mean_terminated_length": 271.6176452636719, "completions/min_length": 51.0, "completions/min_terminated_length": 51.0, "epoch": 0.5321955003878976, "grad_norm": 13.17680835723877, "kl": 56.625, "learning_rate": 2.628479520854041e-06, "loss": 0.8999, "num_tokens": 259794073.0, "reward": 0.14749999344348907, "reward_std": 0.10821715742349625, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.23079390823841095, "step": 1372 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 682.0, "completions/mean_length": 1854.550048828125, "completions/mean_terminated_length": 280.7659606933594, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.5325833979829325, "grad_norm": 16.730602264404297, "kl": 37.59375, "learning_rate": 2.625656310956915e-06, "loss": 0.9344, "num_tokens": 259985725.0, "reward": 0.22124998271465302, "reward_std": 0.18692617118358612, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.18676120042800903, "step": 1373 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 1148.0, "completions/mean_length": 1801.4749755859375, "completions/mean_terminated_length": 271.79168701171875, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.5329712955779674, "grad_norm": 9.58874797821045, "kl": 39.5, "learning_rate": 2.622833297415445e-06, "loss": 0.9279, "num_tokens": 260175923.0, "reward": 0.2487500160932541, "reward_std": 0.20178380608558655, "rewards/format_reward/mean": 0.512499988079071, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.19750000536441803, "rewards/unicoder_reward_fn/std": 0.263364315032959, "step": 1374 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2997.0, "completions/mean_length": 1890.3499755859375, "completions/mean_terminated_length": 341.7021179199219, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.5333591931730023, "grad_norm": 5.546108722686768, "kl": 37.8125, "learning_rate": 2.6200104846875572e-06, "loss": 0.8048, "num_tokens": 260381927.0, "reward": 0.23375001549720764, "reward_std": 0.13028672337532043, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.18000000715255737, "rewards/unicoder_reward_fn/std": 0.23728422820568085, "step": 1375 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1767.0, "completions/mean_length": 2029.7625732421875, "completions/mean_terminated_length": 339.2045593261719, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.5337470907680373, "grad_norm": 11.97602367401123, "kl": 41.75, "learning_rate": 2.6171878772308572e-06, "loss": 1.0596, "num_tokens": 260596372.0, "reward": 0.24500000476837158, "reward_std": 0.23772035539150238, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.1900000125169754, "rewards/unicoder_reward_fn/std": 0.25437939167022705, "step": 1376 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 1158.0, "completions/mean_length": 1628.8375244140625, "completions/mean_terminated_length": 300.3653869628906, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.5341349883630722, "grad_norm": 10.846090316772461, "kl": 41.875, "learning_rate": 2.6143654795026274e-06, "loss": 0.9209, "num_tokens": 260772703.0, "reward": 0.32249999046325684, "reward_std": 0.16081099212169647, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.25999999046325684, "rewards/unicoder_reward_fn/std": 0.33737632632255554, "step": 1377 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1897.0, "completions/mean_length": 1737.800048828125, "completions/mean_terminated_length": 396.8627624511719, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.534522885958107, "grad_norm": 7.749642848968506, "kl": 41.0625, "learning_rate": 2.6115432959598196e-06, "loss": 0.9547, "num_tokens": 260962135.0, "reward": 0.29625001549720764, "reward_std": 0.2304278463125229, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.23499998450279236, "rewards/unicoder_reward_fn/std": 0.2658161520957947, "step": 1378 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 2466.0, "completions/mean_length": 1863.5374755859375, "completions/mean_terminated_length": 375.22918701171875, "completions/min_length": 72.0, "completions/min_terminated_length": 72.0, "epoch": 0.5349107835531419, "grad_norm": 15.142977714538574, "kl": 63.375, "learning_rate": 2.6087213310590455e-06, "loss": 1.0677, "num_tokens": 261162786.0, "reward": 0.16375000774860382, "reward_std": 0.11272072046995163, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.1522822231054306, "step": 1379 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1762.0, "completions/mean_length": 1794.8875732421875, "completions/mean_terminated_length": 414.2200012207031, "completions/min_length": 40.0, "completions/min_terminated_length": 40.0, "epoch": 0.5352986811481769, "grad_norm": 10.382338523864746, "kl": 52.375, "learning_rate": 2.6058995892565743e-06, "loss": 0.8478, "num_tokens": 261350409.0, "reward": 0.3475000262260437, "reward_std": 0.10499675571918488, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.31477341055870056, "step": 1380 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2043.0, "completions/mean_length": 1812.0750732421875, "completions/mean_terminated_length": 441.7200012207031, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.5356865787432118, "grad_norm": 11.452803611755371, "kl": 59.9375, "learning_rate": 2.6030780750083194e-06, "loss": 1.058, "num_tokens": 261541567.0, "reward": 0.2587500214576721, "reward_std": 0.17649534344673157, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.2643358111381531, "step": 1381 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1003.0, "completions/mean_length": 1424.9375, "completions/mean_terminated_length": 280.1964416503906, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.5360744763382467, "grad_norm": 4.763754367828369, "kl": 47.40625, "learning_rate": 2.60025679276984e-06, "loss": 0.9642, "num_tokens": 261696642.0, "reward": 0.36250004172325134, "reward_std": 0.1962418407201767, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.30000001192092896, "rewards/unicoder_reward_fn/std": 0.35861900448799133, "step": 1382 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 814.0, "completions/mean_length": 1378.862548828125, "completions/mean_terminated_length": 282.47369384765625, "completions/min_length": 69.0, "completions/min_terminated_length": 69.0, "epoch": 0.5364623739332817, "grad_norm": 6.296304702758789, "kl": 57.125, "learning_rate": 2.597435746996324e-06, "loss": 1.2784, "num_tokens": 261852007.0, "reward": 0.382500022649765, "reward_std": 0.26535454392433167, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.3125, "rewards/unicoder_reward_fn/std": 0.3304676413536072, "step": 1383 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1077.0, "completions/mean_length": 1569.8125, "completions/mean_terminated_length": 282.8868103027344, "completions/min_length": 70.0, "completions/min_terminated_length": 70.0, "epoch": 0.5368502715283165, "grad_norm": 2.6444923877716064, "kl": 55.5, "learning_rate": 2.594614942142587e-06, "loss": 1.1087, "num_tokens": 262027496.0, "reward": 0.3212500214576721, "reward_std": 0.1623430699110031, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.2550000250339508, "rewards/unicoder_reward_fn/std": 0.2584765553474426, "step": 1384 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1212.0, "completions/mean_length": 2263.97509765625, "completions/mean_terminated_length": 338.0, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.5372381691233514, "grad_norm": 17.19717788696289, "kl": 64.25, "learning_rate": 2.591794382663067e-06, "loss": 1.0445, "num_tokens": 262252534.0, "reward": 0.36125001311302185, "reward_std": 0.19061842560768127, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.3149999976158142, "rewards/unicoder_reward_fn/std": 0.4313945472240448, "step": 1385 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1027.0, "completions/mean_length": 2444.212646484375, "completions/mean_terminated_length": 320.4857177734375, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.5376260667183863, "grad_norm": 15.569392204284668, "kl": 60.5625, "learning_rate": 2.5889740730118106e-06, "loss": 0.9939, "num_tokens": 262500119.0, "reward": 0.29250001907348633, "reward_std": 0.17556528747081757, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.25, "rewards/unicoder_reward_fn/std": 0.34786874055862427, "step": 1386 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5875, "completions/max_length": 4096.0, "completions/max_terminated_length": 599.0, "completions/mean_length": 2498.28759765625, "completions/mean_terminated_length": 222.75758361816406, "completions/min_length": 50.0, "completions/min_terminated_length": 50.0, "epoch": 0.5380139643134213, "grad_norm": 4.039833068847656, "kl": 42.875, "learning_rate": 2.5861540176424727e-06, "loss": 0.8254, "num_tokens": 262751806.0, "reward": 0.13250000774860382, "reward_std": 0.10560162365436554, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.182805135846138, "step": 1387 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 853.0, "completions/mean_length": 2080.212646484375, "completions/mean_terminated_length": 256.4047546386719, "completions/min_length": 56.0, "completions/min_terminated_length": 56.0, "epoch": 0.5384018619084562, "grad_norm": 18.131568908691406, "kl": 31.09375, "learning_rate": 2.583334221008305e-06, "loss": 0.8826, "num_tokens": 262970095.0, "reward": 0.22750000655651093, "reward_std": 0.19178970158100128, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.2988480031490326, "step": 1388 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5875, "completions/max_length": 4096.0, "completions/max_terminated_length": 634.0, "completions/mean_length": 2525.212646484375, "completions/mean_terminated_length": 288.0303039550781, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.5387897595034911, "grad_norm": 9.76858139038086, "kl": 28.8125, "learning_rate": 2.5805146875621518e-06, "loss": 0.686, "num_tokens": 263227928.0, "reward": 0.12375000864267349, "reward_std": 0.10225250571966171, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.09250000864267349, "rewards/unicoder_reward_fn/std": 0.20853617787361145, "step": 1389 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 673.0, "completions/mean_length": 2364.91259765625, "completions/mean_terminated_length": 249.13888549804688, "completions/min_length": 47.0, "completions/min_terminated_length": 47.0, "epoch": 0.5391776570985259, "grad_norm": 23.27193832397461, "kl": 26.84375, "learning_rate": 2.5776954217564408e-06, "loss": 0.8534, "num_tokens": 263459353.0, "reward": 0.17625001072883606, "reward_std": 0.20056524872779846, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953901171684265, "rewards/unicoder_reward_fn/mean": 0.13499999046325684, "rewards/unicoder_reward_fn/std": 0.2947215139865875, "step": 1390 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 978.0, "completions/mean_length": 2465.625, "completions/mean_terminated_length": 259.8235168457031, "completions/min_length": 64.0, "completions/min_terminated_length": 64.0, "epoch": 0.5395655546935609, "grad_norm": 13.728394508361816, "kl": 22.8125, "learning_rate": 2.574876428043178e-06, "loss": 0.628, "num_tokens": 263706603.0, "reward": 0.20000000298023224, "reward_std": 0.16780070960521698, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.27131885290145874, "step": 1391 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 709.0, "completions/mean_length": 2656.987548828125, "completions/mean_terminated_length": 258.63336181640625, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 0.5399534522885958, "grad_norm": 2.1439270973205566, "kl": 34.8125, "learning_rate": 2.5720577108739394e-06, "loss": 0.7012, "num_tokens": 263969314.0, "reward": 0.11125000566244125, "reward_std": 0.11546404659748077, "rewards/format_reward/mean": 0.38749998807907104, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.1566864252090454, "step": 1392 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 1747.0, "completions/mean_length": 2214.35009765625, "completions/mean_terminated_length": 332.70001220703125, "completions/min_length": 78.0, "completions/min_terminated_length": 78.0, "epoch": 0.5403413498836307, "grad_norm": 13.951356887817383, "kl": 30.78125, "learning_rate": 2.5692392746998634e-06, "loss": 0.7258, "num_tokens": 264192502.0, "reward": 0.17374999821186066, "reward_std": 0.13059750199317932, "rewards/format_reward/mean": 0.512499988079071, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.12250000238418579, "rewards/unicoder_reward_fn/std": 0.18139752745628357, "step": 1393 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 782.0, "completions/mean_length": 2279.087646484375, "completions/mean_terminated_length": 270.9210510253906, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.5407292474786657, "grad_norm": 3.52634859085083, "kl": 33.5625, "learning_rate": 2.5664211239716457e-06, "loss": 0.7503, "num_tokens": 264412421.0, "reward": 0.13125000894069672, "reward_std": 0.11714740842580795, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.16466307640075684, "step": 1394 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1314.0, "completions/mean_length": 2075.449951171875, "completions/mean_terminated_length": 336.83721923828125, "completions/min_length": 44.0, "completions/min_terminated_length": 44.0, "epoch": 0.5411171450737006, "grad_norm": 12.637396812438965, "kl": 40.0, "learning_rate": 2.5636032631395317e-06, "loss": 0.9386, "num_tokens": 264627705.0, "reward": 0.24250002205371857, "reward_std": 0.23272907733917236, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.1899999976158142, "rewards/unicoder_reward_fn/std": 0.303398460149765, "step": 1395 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1816.0, "completions/mean_length": 2128.400146484375, "completions/mean_terminated_length": 435.3488464355469, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.5415050426687354, "grad_norm": 12.542069435119629, "kl": 40.3125, "learning_rate": 2.5607856966533067e-06, "loss": 0.8473, "num_tokens": 264847529.0, "reward": 0.2587500214576721, "reward_std": 0.1781400442123413, "rewards/format_reward/mean": 0.512499988079071, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.2924254834651947, "step": 1396 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1824.0, "completions/mean_length": 2289.625, "completions/mean_terminated_length": 390.6153869628906, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.5418929402637703, "grad_norm": 5.1371636390686035, "kl": 55.4375, "learning_rate": 2.557968428962295e-06, "loss": 1.1355, "num_tokens": 265075683.0, "reward": 0.21500001847743988, "reward_std": 0.22326652705669403, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.2768608033657074, "step": 1397 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1260.0, "completions/mean_length": 2423.175048828125, "completions/mean_terminated_length": 272.3999938964844, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.5422808378588053, "grad_norm": 9.55141830444336, "kl": 57.625, "learning_rate": 2.5551514645153452e-06, "loss": 1.1388, "num_tokens": 265321305.0, "reward": 0.17499999701976776, "reward_std": 0.1514541655778885, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.22360680997371674, "step": 1398 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1669.0, "completions/mean_length": 2172.3125, "completions/mean_terminated_length": 342.4634094238281, "completions/min_length": 47.0, "completions/min_terminated_length": 47.0, "epoch": 0.5426687354538402, "grad_norm": 14.536443710327148, "kl": 60.25, "learning_rate": 2.5523348077608286e-06, "loss": 1.0224, "num_tokens": 265543378.0, "reward": 0.19500000774860382, "reward_std": 0.14742226898670197, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.25051844120025635, "step": 1399 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 1953.0, "completions/mean_length": 2581.72509765625, "completions/mean_terminated_length": 310.3125, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.5430566330488751, "grad_norm": 14.099510192871094, "kl": 73.375, "learning_rate": 2.549518463146631e-06, "loss": 1.2135, "num_tokens": 265798220.0, "reward": 0.16375000774860382, "reward_std": 0.19581837952136993, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.12250000238418579, "rewards/unicoder_reward_fn/std": 0.21698452532291412, "step": 1400 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3154.0, "completions/mean_length": 2730.91259765625, "completions/mean_terminated_length": 455.7666931152344, "completions/min_length": 50.0, "completions/min_terminated_length": 50.0, "epoch": 0.54344453064391, "grad_norm": 17.071985244750977, "kl": 74.625, "learning_rate": 2.5467024351201446e-06, "loss": 1.2091, "num_tokens": 266071309.0, "reward": 0.1237499937415123, "reward_std": 0.149052694439888, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.170461043715477, "step": 1401 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1193.0, "completions/mean_length": 1960.9749755859375, "completions/mean_terminated_length": 214.13636779785156, "completions/min_length": 35.0, "completions/min_terminated_length": 35.0, "epoch": 0.543832428238945, "grad_norm": 10.720776557922363, "kl": 50.25, "learning_rate": 2.543886728128264e-06, "loss": 1.0766, "num_tokens": 266266923.0, "reward": 0.4475000500679016, "reward_std": 0.3051479756832123, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.39000001549720764, "rewards/unicoder_reward_fn/std": 0.3922024965286255, "step": 1402 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1192.0, "completions/mean_length": 2248.087646484375, "completions/mean_terminated_length": 305.4102478027344, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.5442203258339798, "grad_norm": 5.292662620544434, "kl": 29.21875, "learning_rate": 2.5410713466173732e-06, "loss": 0.5571, "num_tokens": 266497522.0, "reward": 0.20875000953674316, "reward_std": 0.09307878464460373, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.49025312066078186, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.23328964412212372, "step": 1403 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1072.0, "completions/mean_length": 1461.1500244140625, "completions/mean_terminated_length": 263.49090576171875, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.5446082234290147, "grad_norm": 16.92027473449707, "kl": 16.765625, "learning_rate": 2.538256295033343e-06, "loss": 0.5418, "num_tokens": 266666062.0, "reward": 0.2550000250339508, "reward_std": 0.14713867008686066, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.2567567825317383, "step": 1404 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1473.0, "completions/mean_length": 1491.375, "completions/mean_terminated_length": 237.29629516601562, "completions/min_length": 75.0, "completions/min_terminated_length": 75.0, "epoch": 0.5449961210240496, "grad_norm": 7.502524375915527, "kl": 33.78125, "learning_rate": 2.5354415778215257e-06, "loss": 0.8285, "num_tokens": 266834700.0, "reward": 0.23250000178813934, "reward_std": 0.15097270905971527, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.19119875133037567, "step": 1405 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 556.0, "completions/mean_length": 864.1875, "completions/mean_terminated_length": 178.65151977539062, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.5453840186190846, "grad_norm": 19.93394660949707, "kl": 22.3671875, "learning_rate": 2.532627199426743e-06, "loss": 0.7465, "num_tokens": 266952651.0, "reward": 0.4162500500679016, "reward_std": 0.15008561313152313, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.3375000059604645, "rewards/unicoder_reward_fn/std": 0.28920450806617737, "step": 1406 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1965.0, "completions/mean_length": 794.5125122070312, "completions/mean_terminated_length": 268.18841552734375, "completions/min_length": 75.0, "completions/min_terminated_length": 75.0, "epoch": 0.5457719162141195, "grad_norm": 6.38470983505249, "kl": 20.90625, "learning_rate": 2.5298131642932845e-06, "loss": 0.4216, "num_tokens": 267062340.0, "reward": 0.4074999988079071, "reward_std": 0.22834233939647675, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.32249999046325684, "rewards/unicoder_reward_fn/std": 0.3607218563556671, "step": 1407 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 726.0, "completions/mean_length": 1682.862548828125, "completions/mean_terminated_length": 234.97999572753906, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.5461598138091543, "grad_norm": 4.158061981201172, "kl": 59.15625, "learning_rate": 2.526999476864894e-06, "loss": 1.1545, "num_tokens": 267246393.0, "reward": 0.23375001549720764, "reward_std": 0.20526523888111115, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.17249999940395355, "rewards/unicoder_reward_fn/std": 0.25406816601753235, "step": 1408 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 532.0, "completions/mean_length": 1213.6375732421875, "completions/mean_terminated_length": 252.85000610351562, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.5465477114041893, "grad_norm": 8.474355697631836, "kl": 57.625, "learning_rate": 2.524186141584766e-06, "loss": 1.189, "num_tokens": 267392692.0, "reward": 0.4150000512599945, "reward_std": 0.18345202505588531, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.33750003576278687, "rewards/unicoder_reward_fn/std": 0.3494570851325989, "step": 1409 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 619.0, "completions/mean_length": 1931.050048828125, "completions/mean_terminated_length": 247.20001220703125, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.5469356089992242, "grad_norm": 10.868085861206055, "kl": 83.125, "learning_rate": 2.521373162895546e-06, "loss": 1.4762, "num_tokens": 267596432.0, "reward": 0.36500000953674316, "reward_std": 0.23877504467964172, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.3100000023841858, "rewards/unicoder_reward_fn/std": 0.3696046471595764, "step": 1410 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 437.0, "completions/mean_length": 1679.6375732421875, "completions/mean_terminated_length": 229.8199920654297, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.5473235065942591, "grad_norm": 13.653889656066895, "kl": 86.25, "learning_rate": 2.518560545239307e-06, "loss": 1.428, "num_tokens": 267778195.0, "reward": 0.35374999046325684, "reward_std": 0.23953519761562347, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.28434914350509644, "step": 1411 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 470.0, "completions/mean_length": 2072.03759765625, "completions/mean_terminated_length": 240.83334350585938, "completions/min_length": 67.0, "completions/min_terminated_length": 67.0, "epoch": 0.547711404189294, "grad_norm": 17.621566772460938, "kl": 82.5, "learning_rate": 2.5157482930575585e-06, "loss": 1.2609, "num_tokens": 267991614.0, "reward": 0.2225000113248825, "reward_std": 0.1838037222623825, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.24118642508983612, "step": 1412 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 425.0, "completions/mean_length": 1714.4625244140625, "completions/mean_terminated_length": 207.7755126953125, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.548099301784329, "grad_norm": 15.200885772705078, "kl": 80.875, "learning_rate": 2.512936410791229e-06, "loss": 1.3226, "num_tokens": 268181267.0, "reward": 0.23875002562999725, "reward_std": 0.2061721533536911, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.17249998450279236, "rewards/unicoder_reward_fn/std": 0.27695223689079285, "step": 1413 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2005.0, "completions/mean_length": 1352.4375, "completions/mean_terminated_length": 245.38597106933594, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.5484871993793639, "grad_norm": 4.977359771728516, "kl": 45.5625, "learning_rate": 2.5101249028806655e-06, "loss": 0.9151, "num_tokens": 268343998.0, "reward": 0.34375, "reward_std": 0.16874969005584717, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.27000001072883606, "rewards/unicoder_reward_fn/std": 0.2986891269683838, "step": 1414 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2256.0, "completions/mean_length": 1398.8250732421875, "completions/mean_terminated_length": 242.8928680419922, "completions/min_length": 76.0, "completions/min_terminated_length": 76.0, "epoch": 0.5488750969743987, "grad_norm": 25.442140579223633, "kl": 38.875, "learning_rate": 2.507313773765623e-06, "loss": 1.2213, "num_tokens": 268504520.0, "reward": 0.23375001549720764, "reward_std": 0.18563173711299896, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.24462580680847168, "step": 1415 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2000.0, "completions/mean_length": 1458.5, "completions/mean_terminated_length": 259.6363525390625, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.5492629945694336, "grad_norm": 16.618846893310547, "kl": 26.4375, "learning_rate": 2.504503027885256e-06, "loss": 0.7945, "num_tokens": 268665216.0, "reward": 0.26750001311302185, "reward_std": 0.1259322464466095, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.1925000101327896, "rewards/unicoder_reward_fn/std": 0.25743380188941956, "step": 1416 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1622.0, "completions/mean_length": 1084.25, "completions/mean_terminated_length": 209.87095642089844, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.5496508921644686, "grad_norm": 29.784547805786133, "kl": 23.21875, "learning_rate": 2.501692669678119e-06, "loss": 0.8288, "num_tokens": 268802484.0, "reward": 0.22750000655651093, "reward_std": 0.12397072464227676, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.1786021739244461, "step": 1417 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1679.0, "completions/mean_length": 1908.925048828125, "completions/mean_terminated_length": 292.39129638671875, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.5500387897595035, "grad_norm": 9.020005226135254, "kl": 44.375, "learning_rate": 2.4988827035821496e-06, "loss": 1.0179, "num_tokens": 269004094.0, "reward": 0.35625001788139343, "reward_std": 0.2700204849243164, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.29749998450279236, "rewards/unicoder_reward_fn/std": 0.36836108565330505, "step": 1418 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 3711.0, "completions/mean_length": 2231.449951171875, "completions/mean_terminated_length": 366.8999938964844, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.5504266873545384, "grad_norm": 6.859643936157227, "kl": 59.1875, "learning_rate": 2.496073134034667e-06, "loss": 1.1506, "num_tokens": 269227970.0, "reward": 0.21125002205371857, "reward_std": 0.2089776247739792, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.23620149493217468, "step": 1419 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1777.0, "completions/mean_length": 2140.5625, "completions/mean_terminated_length": 280.5121765136719, "completions/min_length": 53.0, "completions/min_terminated_length": 53.0, "epoch": 0.5508145849495734, "grad_norm": 9.85229778289795, "kl": 66.75, "learning_rate": 2.493263965472366e-06, "loss": 1.2936, "num_tokens": 269447847.0, "reward": 0.2237500250339508, "reward_std": 0.19041834771633148, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.18000000715255737, "rewards/unicoder_reward_fn/std": 0.25969797372817993, "step": 1420 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 630.0, "completions/mean_length": 2217.675048828125, "completions/mean_terminated_length": 243.02565002441406, "completions/min_length": 53.0, "completions/min_terminated_length": 53.0, "epoch": 0.5512024825446082, "grad_norm": 8.548258781433105, "kl": 67.25, "learning_rate": 2.490455202331307e-06, "loss": 1.3087, "num_tokens": 269671989.0, "reward": 0.21125002205371857, "reward_std": 0.22890983521938324, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.29126524925231934, "step": 1421 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 475.0, "completions/mean_length": 3609.375, "completions/mean_terminated_length": 203.0, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.5515903801396431, "grad_norm": 34.48482894897461, "kl": 69.0, "learning_rate": 2.48764684904691e-06, "loss": 0.9294, "num_tokens": 270005459.0, "reward": 0.046250004321336746, "reward_std": 0.10231158882379532, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.03500000014901161, "rewards/unicoder_reward_fn/std": 0.13416408002376556, "step": 1422 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 215.0, "completions/mean_length": 3451.862548828125, "completions/mean_terminated_length": 132.07693481445312, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.551978277734678, "grad_norm": 19.133695602416992, "kl": 53.875, "learning_rate": 2.48483891005395e-06, "loss": 0.8341, "num_tokens": 270332296.0, "reward": 0.07500000298023224, "reward_std": 0.11980140209197998, "rewards/format_reward/mean": 0.17499999701976776, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.15652476251125336, "step": 1423 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 267.0, "completions/mean_length": 3900.775146484375, "completions/mean_terminated_length": 191.5, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.552366175329713, "grad_norm": 18.88619041442871, "kl": 35.5, "learning_rate": 2.4820313897865433e-06, "loss": 0.4792, "num_tokens": 270695110.0, "reward": 0.030000001192092896, "reward_std": 0.08485281467437744, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.12877219915390015, "step": 1424 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9125, "completions/max_length": 4096.0, "completions/max_terminated_length": 375.0, "completions/mean_length": 3751.949951171875, "completions/mean_terminated_length": 164.0, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.5527540729247479, "grad_norm": 4.243536949157715, "kl": 17.71875, "learning_rate": 2.4792242926781498e-06, "loss": 0.3424, "num_tokens": 271036666.0, "reward": 0.03125, "reward_std": 0.06199761852622032, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.022499999031424522, "rewards/unicoder_reward_fn/std": 0.07790776342153549, "step": 1425 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 315.0, "completions/mean_length": 3949.137451171875, "completions/mean_terminated_length": 179.6666717529297, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.5531419705197828, "grad_norm": 2.628861665725708, "kl": 7.1953125, "learning_rate": 2.4764176231615578e-06, "loss": 0.1457, "num_tokens": 271400581.0, "reward": 0.01875000074505806, "reward_std": 0.045707106590270996, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.09426598995923996, "step": 1426 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 244.0, "completions/mean_length": 3898.47509765625, "completions/mean_terminated_length": 145.5, "completions/min_length": 38.0, "completions/min_terminated_length": 38.0, "epoch": 0.5535298681148176, "grad_norm": 3.4772305488586426, "kl": 2.98828125, "learning_rate": 2.4736113856688808e-06, "loss": 0.1183, "num_tokens": 271761819.0, "reward": 0.008750000037252903, "reward_std": 0.02474873699247837, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.044721364974975586, "step": 1427 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 221.0, "completions/mean_length": 4047.5625, "completions/mean_terminated_length": 221.0, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.5539177657098526, "grad_norm": 0.9297699332237244, "kl": 1.64453125, "learning_rate": 2.4708055846315533e-06, "loss": 0.0479, "num_tokens": 272138432.0, "reward": 0.008750000968575478, "reward_std": 0.02474874071776867, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 1428 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.5543056633048875, "grad_norm": 1.2463856935501099, "kl": 1.1953125, "learning_rate": 2.4680002244803154e-06, "loss": 0.0119, "num_tokens": 272517320.0, "reward": 0.010000000707805157, "reward_std": 0.02828427031636238, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944272994995117, "step": 1429 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 81.0, "completions/mean_length": 4045.8125, "completions/mean_terminated_length": 81.0, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "epoch": 0.5546935608999224, "grad_norm": 0.8169512748718262, "kl": 0.796875, "learning_rate": 2.4651953096452163e-06, "loss": 0.0395, "num_tokens": 272882417.0, "reward": 0.003750000149011612, "reward_std": 0.01060660183429718, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1430 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 325.0, "completions/mean_length": 4048.862548828125, "completions/mean_terminated_length": 325.0, "completions/min_length": 325.0, "completions/min_terminated_length": 325.0, "epoch": 0.5550814584949574, "grad_norm": 1.2212711572647095, "kl": 0.5556640625, "learning_rate": 2.462390844555598e-06, "loss": 0.0346, "num_tokens": 273262630.0, "reward": 0.016249999403953552, "reward_std": 0.04596194252371788, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.0994923859834671, "step": 1431 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 189.0, "completions/mean_length": 3898.775146484375, "completions/mean_terminated_length": 151.5, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.5554693560899923, "grad_norm": 3.8148276805877686, "kl": 0.9365234375, "learning_rate": 2.4595868336400923e-06, "loss": 0.1175, "num_tokens": 273624236.0, "reward": 0.015000000596046448, "reward_std": 0.035100504755973816, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.04386400431394577, "step": 1432 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 255.0, "completions/mean_length": 3899.875, "completions/mean_terminated_length": 173.5, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.5558572536850271, "grad_norm": 3.7593994140625, "kl": 2.4296875, "learning_rate": 2.4567832813266178e-06, "loss": 0.1381, "num_tokens": 273977738.0, "reward": 0.020000001415610313, "reward_std": 0.049242641776800156, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.07647287845611572, "step": 1433 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.556245151280062, "grad_norm": 2.2659716606140137, "kl": 3.21875, "learning_rate": 2.4539801920423646e-06, "loss": 0.0322, "num_tokens": 274364650.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1434 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 95.0, "completions/mean_length": 4045.987548828125, "completions/mean_terminated_length": 95.0, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.556633048875097, "grad_norm": 1.202398657798767, "kl": 3.57421875, "learning_rate": 2.451177570213794e-06, "loss": 0.0686, "num_tokens": 274725665.0, "reward": 0.0062500000931322575, "reward_std": 0.01767767034471035, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.044721364974975586, "step": 1435 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 195.0, "completions/mean_length": 4047.237548828125, "completions/mean_terminated_length": 195.0, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.5570209464701319, "grad_norm": 1.7217556238174438, "kl": 4.01171875, "learning_rate": 2.448375420266629e-06, "loss": 0.0704, "num_tokens": 275111716.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1436 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.5574088440651668, "grad_norm": 1.586068868637085, "kl": 2.1328125, "learning_rate": 2.445573746625845e-06, "loss": 0.0213, "num_tokens": 275493844.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1437 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 107.0, "completions/mean_length": 4046.137451171875, "completions/mean_terminated_length": 107.0, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.5577967416602017, "grad_norm": 0.7995090484619141, "kl": 1.603515625, "learning_rate": 2.4427725537156683e-06, "loss": 0.0487, "num_tokens": 275877903.0, "reward": 0.003750000149011612, "reward_std": 0.01060660183429718, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1438 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.5581846392552366, "grad_norm": 0.8977259993553162, "kl": 1.228515625, "learning_rate": 2.439971845959565e-06, "loss": 0.0123, "num_tokens": 276245655.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1439 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 675.0, "completions/mean_length": 4004.400146484375, "completions/mean_terminated_length": 432.0, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.5585725368502715, "grad_norm": 1.278554916381836, "kl": 0.8701171875, "learning_rate": 2.4371716277802345e-06, "loss": 0.0507, "num_tokens": 276611919.0, "reward": 0.017500001937150955, "reward_std": 0.033743150532245636, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.09426598995923996, "step": 1440 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 227.0, "completions/mean_length": 3801.16259765625, "completions/mean_terminated_length": 164.83334350585938, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.5589604344453064, "grad_norm": 5.658029079437256, "kl": 0.791015625, "learning_rate": 2.4343719035996034e-06, "loss": 0.1628, "num_tokens": 276968388.0, "reward": 0.033750005066394806, "reward_std": 0.066155806183815, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.09209240972995758, "step": 1441 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 211.0, "completions/mean_length": 3849.10009765625, "completions/mean_terminated_length": 145.60000610351562, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.5593483320403414, "grad_norm": 4.4965291023254395, "kl": 0.6435546875, "learning_rate": 2.431572677838819e-06, "loss": 0.1266, "num_tokens": 277320628.0, "reward": 0.012500000186264515, "reward_std": 0.0297470074146986, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.04974619299173355, "step": 1442 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 117.0, "completions/mean_length": 4046.262451171875, "completions/mean_terminated_length": 117.0, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.5597362296353763, "grad_norm": 1.4049103260040283, "kl": 0.8232421875, "learning_rate": 2.4287739549182406e-06, "loss": 0.0422, "num_tokens": 277685529.0, "reward": 0.007500000298023224, "reward_std": 0.02121320366859436, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472135752439499, "step": 1443 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 225.0, "completions/mean_length": 3998.625, "completions/mean_terminated_length": 201.0, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.5601241272304112, "grad_norm": 2.223289966583252, "kl": 1.298828125, "learning_rate": 2.4259757392574327e-06, "loss": 0.0655, "num_tokens": 278050435.0, "reward": 0.03375000134110451, "reward_std": 0.07724738121032715, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.030000004917383194, "rewards/unicoder_reward_fn/std": 0.12768632173538208, "step": 1444 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 239.0, "completions/mean_length": 3948.737548828125, "completions/mean_terminated_length": 169.0, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.560512024825446, "grad_norm": 2.7930235862731934, "kl": 2.09375, "learning_rate": 2.4231780352751616e-06, "loss": 0.114, "num_tokens": 278408814.0, "reward": 0.0062500000931322575, "reward_std": 0.0176776684820652, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360678762197495, "step": 1445 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 225.0, "completions/mean_length": 4047.612548828125, "completions/mean_terminated_length": 225.0, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "epoch": 0.560899922420481, "grad_norm": 1.1172118186950684, "kl": 3.38671875, "learning_rate": 2.420380847389383e-06, "loss": 0.0638, "num_tokens": 278790191.0, "reward": 0.003750000149011612, "reward_std": 0.01060660183429718, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1446 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 325.0, "completions/mean_length": 4048.862548828125, "completions/mean_terminated_length": 325.0, "completions/min_length": 325.0, "completions/min_terminated_length": 325.0, "epoch": 0.5612878200155159, "grad_norm": 1.0406296253204346, "kl": 3.69921875, "learning_rate": 2.4175841800172373e-06, "loss": 0.0659, "num_tokens": 279152364.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1447 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.5616757176105508, "grad_norm": 1.1516505479812622, "kl": 1.994140625, "learning_rate": 2.4147880375750454e-06, "loss": 0.0199, "num_tokens": 279529500.0, "reward": 0.005000000353902578, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.044721364974975586, "step": 1448 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 173.0, "completions/mean_length": 3997.824951171875, "completions/mean_terminated_length": 169.0, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.5620636152055857, "grad_norm": 1.0558521747589111, "kl": 3.3125, "learning_rate": 2.4119924244782966e-06, "loss": 0.0653, "num_tokens": 279888166.0, "reward": 0.015000000596046448, "reward_std": 0.03854496404528618, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 1449 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 183.0, "completions/mean_length": 3996.8125, "completions/mean_terminated_length": 128.5, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.5624515128006207, "grad_norm": 1.9146497249603271, "kl": 1.45703125, "learning_rate": 2.4091973451416474e-06, "loss": 0.078, "num_tokens": 280259311.0, "reward": 0.09375, "reward_std": 0.03889086842536926, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0924999937415123, "rewards/unicoder_reward_fn/std": 0.2544664442539215, "step": 1450 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 378.0, "completions/mean_length": 3951.387451171875, "completions/mean_terminated_length": 239.6666717529297, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.5628394103956555, "grad_norm": 3.4928297996520996, "kl": 1.626953125, "learning_rate": 2.4064028039789084e-06, "loss": 0.1128, "num_tokens": 280618150.0, "reward": 0.020000001415610313, "reward_std": 0.05656854063272476, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.017500000074505806, "rewards/unicoder_reward_fn/std": 0.11988918483257294, "step": 1451 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.5632273079906904, "grad_norm": 0.9967352151870728, "kl": 1.333984375, "learning_rate": 2.403608805403041e-06, "loss": 0.0133, "num_tokens": 280998622.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1452 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 252.0, "completions/mean_length": 3949.387451171875, "completions/mean_terminated_length": 186.33334350585938, "completions/min_length": 57.0, "completions/min_terminated_length": 57.0, "epoch": 0.5636152055857254, "grad_norm": 3.188056230545044, "kl": 1.72265625, "learning_rate": 2.400815353826152e-06, "loss": 0.1125, "num_tokens": 281360805.0, "reward": 0.005000000353902578, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360680624842644, "step": 1453 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.5640031031807603, "grad_norm": 1.8744856119155884, "kl": 2.453125, "learning_rate": 2.3980224536594806e-06, "loss": 0.0245, "num_tokens": 281731181.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1454 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 648.0, "completions/mean_length": 4052.900146484375, "completions/mean_terminated_length": 648.0, "completions/min_length": 648.0, "completions/min_terminated_length": 648.0, "epoch": 0.5643910007757952, "grad_norm": 0.7146034836769104, "kl": 1.64453125, "learning_rate": 2.3952301093133995e-06, "loss": 0.0447, "num_tokens": 282109645.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1455 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 635.0, "completions/mean_length": 4004.375, "completions/mean_terminated_length": 431.0, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.5647788983708301, "grad_norm": 1.5299909114837646, "kl": 1.779296875, "learning_rate": 2.3924383251974004e-06, "loss": 0.0752, "num_tokens": 282477203.0, "reward": 0.012500000186264515, "reward_std": 0.0353553406894207, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944271504878998, "step": 1456 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 510.0, "completions/mean_length": 4004.462646484375, "completions/mean_terminated_length": 434.5, "completions/min_length": 359.0, "completions/min_terminated_length": 359.0, "epoch": 0.565166795965865, "grad_norm": 0.7173221111297607, "kl": 1.64453125, "learning_rate": 2.389647105720091e-06, "loss": 0.0474, "num_tokens": 282846984.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1457 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1161.0, "completions/mean_length": 4059.3125, "completions/mean_terminated_length": 1161.0, "completions/min_length": 1161.0, "completions/min_terminated_length": 1161.0, "epoch": 0.5655546935608999, "grad_norm": 2.1533119678497314, "kl": 2.43359375, "learning_rate": 2.3868564552891894e-06, "loss": 0.0244, "num_tokens": 283217537.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1458 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 294.0, "completions/mean_length": 3999.41259765625, "completions/mean_terminated_length": 232.5, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.5659425911559348, "grad_norm": 0.8813987970352173, "kl": 2.9296875, "learning_rate": 2.384066378311512e-06, "loss": 0.0891, "num_tokens": 283579946.0, "reward": 0.010000000707805157, "reward_std": 0.02828427217900753, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 1459 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 1266.0, "completions/mean_length": 3737.75, "completions/mean_terminated_length": 513.5, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.5663304887509697, "grad_norm": 3.691190242767334, "kl": 2.62109375, "learning_rate": 2.3812768791929724e-06, "loss": 0.1435, "num_tokens": 283921654.0, "reward": 0.02250000089406967, "reward_std": 0.05096125602722168, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 1460 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 265.0, "completions/mean_length": 4048.112548828125, "completions/mean_terminated_length": 265.0, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "epoch": 0.5667183863460047, "grad_norm": 2.4518909454345703, "kl": 3.37890625, "learning_rate": 2.37848796233857e-06, "loss": 0.0338, "num_tokens": 284286903.0, "reward": 0.0, "reward_std": 0.0, "rewards/format_reward/mean": 0.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1461 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 1635.0, "completions/mean_length": 3928.425048828125, "completions/mean_terminated_length": 744.5, "completions/min_length": 380.0, "completions/min_terminated_length": 380.0, "epoch": 0.5671062839410396, "grad_norm": 1.5633727312088013, "kl": 4.34375, "learning_rate": 2.3756996321523845e-06, "loss": 0.1004, "num_tokens": 284656713.0, "reward": 0.0062500000931322575, "reward_std": 0.014511304907500744, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1462 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1733.0, "completions/mean_length": 3621.300048828125, "completions/mean_terminated_length": 643.6363525390625, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.5674941815360744, "grad_norm": 7.483672142028809, "kl": 5.65625, "learning_rate": 2.372911893037571e-06, "loss": 0.247, "num_tokens": 284991337.0, "reward": 0.020000001415610313, "reward_std": 0.03240213543176651, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1463 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1165.0, "completions/mean_length": 3887.487548828125, "completions/mean_terminated_length": 759.7999877929688, "completions/min_length": 340.0, "completions/min_terminated_length": 340.0, "epoch": 0.5678820791311094, "grad_norm": 3.8498189449310303, "kl": 5.3515625, "learning_rate": 2.370124749396349e-06, "loss": 0.1425, "num_tokens": 285348808.0, "reward": 0.015000000596046448, "reward_std": 0.026680251583456993, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1464 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3753.0, "completions/mean_length": 3633.3125, "completions/mean_terminated_length": 1248.6923828125, "completions/min_length": 159.0, "completions/min_terminated_length": 159.0, "epoch": 0.5682699767261443, "grad_norm": 4.787471771240234, "kl": 5.6796875, "learning_rate": 2.36733820563e-06, "loss": 0.1881, "num_tokens": 285687529.0, "reward": 0.016249999403953552, "reward_std": 0.030215784907341003, "rewards/format_reward/mean": 0.16249999403953552, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1465 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3872.0, "completions/mean_length": 3476.85009765625, "completions/mean_terminated_length": 1182.3529052734375, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "epoch": 0.5686578743211792, "grad_norm": 6.993010520935059, "kl": 8.8828125, "learning_rate": 2.364552266138856e-06, "loss": 0.3042, "num_tokens": 286010573.0, "reward": 0.027500001713633537, "reward_std": 0.035644859075546265, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1466 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3388.0, "completions/mean_length": 3847.625, "completions/mean_terminated_length": 1257.4285888671875, "completions/min_length": 508.0, "completions/min_terminated_length": 508.0, "epoch": 0.5690457719162141, "grad_norm": 2.4038174152374268, "kl": 10.40625, "learning_rate": 2.3617669353222956e-06, "loss": 0.2007, "num_tokens": 286365023.0, "reward": 0.012500000186264515, "reward_std": 0.024116460233926773, "rewards/format_reward/mean": 0.125, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1467 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1063.0, "completions/mean_length": 3462.10009765625, "completions/mean_terminated_length": 715.2000122070312, "completions/min_length": 273.0, "completions/min_terminated_length": 273.0, "epoch": 0.5694336695112491, "grad_norm": 6.282442569732666, "kl": 12.203125, "learning_rate": 2.3589822175787373e-06, "loss": 0.2466, "num_tokens": 286694695.0, "reward": 0.027500001713633537, "reward_std": 0.035644859075546265, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1468 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3365.0, "completions/mean_length": 3424.587646484375, "completions/mean_terminated_length": 1269.0, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "epoch": 0.569821567106284, "grad_norm": 13.570630073547363, "kl": 14.265625, "learning_rate": 2.3561981173056292e-06, "loss": 0.2912, "num_tokens": 287016286.0, "reward": 0.03500000014901161, "reward_std": 0.048268865793943405, "rewards/format_reward/mean": 0.3499999940395355, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1469 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3098.0, "completions/mean_length": 3002.112548828125, "completions/mean_terminated_length": 854.8518676757812, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "epoch": 0.5702094647013188, "grad_norm": 5.706627368927002, "kl": 23.625, "learning_rate": 2.353414638899447e-06, "loss": 0.4896, "num_tokens": 287294487.0, "reward": 0.04000000283122063, "reward_std": 0.048268865793943405, "rewards/format_reward/mean": 0.4000000059604645, "rewards/format_reward/std": 0.49298882484436035, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1470 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.775, "completions/max_length": 4096.0, "completions/max_terminated_length": 3950.0, "completions/mean_length": 3597.75, "completions/mean_terminated_length": 1881.5555419921875, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.5705973622963537, "grad_norm": 15.924654006958008, "kl": 28.78125, "learning_rate": 2.350631786755683e-06, "loss": 0.4186, "num_tokens": 287627067.0, "reward": 0.030000001192092896, "reward_std": 0.040273960679769516, "rewards/format_reward/mean": 0.30000001192092896, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1471 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 2552.0, "completions/mean_length": 3517.375, "completions/mean_terminated_length": 1202.875, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.5709852598913887, "grad_norm": 21.167945861816406, "kl": 35.9375, "learning_rate": 2.3478495652688384e-06, "loss": 0.5889, "num_tokens": 287949825.0, "reward": 0.023749999701976776, "reward_std": 0.039473988115787506, "rewards/format_reward/mean": 0.23749999701976776, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1472 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 2529.0, "completions/mean_length": 3867.3125, "completions/mean_terminated_length": 1046.8333740234375, "completions/min_length": 220.0, "completions/min_terminated_length": 220.0, "epoch": 0.5713731574864236, "grad_norm": 43.9486198425293, "kl": 55.125, "learning_rate": 2.3450679788324236e-06, "loss": 0.6741, "num_tokens": 288307434.0, "reward": 0.007500000298023224, "reward_std": 0.018771236762404442, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1473 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3433.0, "completions/mean_length": 3688.824951171875, "completions/mean_terminated_length": 1134.727294921875, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.5717610550814585, "grad_norm": 28.65070152282715, "kl": 65.5, "learning_rate": 2.342287031838941e-06, "loss": 0.7655, "num_tokens": 288646204.0, "reward": 0.021250002086162567, "reward_std": 0.03726455941796303, "rewards/format_reward/mean": 0.13750000298023224, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 1474 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 2312.0, "completions/mean_length": 4033.574951171875, "completions/mean_terminated_length": 1599.0, "completions/min_length": 886.0, "completions/min_terminated_length": 886.0, "epoch": 0.5721489526764935, "grad_norm": 52.282012939453125, "kl": 66.375, "learning_rate": 2.339506728679887e-06, "loss": 0.6759, "num_tokens": 289012138.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1475 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1869.0, "completions/mean_length": 3915.712646484375, "completions/mean_terminated_length": 1211.4000244140625, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "epoch": 0.5725368502715283, "grad_norm": 49.99222183227539, "kl": 58.1875, "learning_rate": 2.33672707374574e-06, "loss": 0.5809, "num_tokens": 289375987.0, "reward": 0.0012500000884756446, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.012500000186264515, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1476 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 3469.0, "completions/mean_length": 3523.5, "completions/mean_terminated_length": 1233.5, "completions/min_length": 118.0, "completions/min_terminated_length": 118.0, "epoch": 0.5729247478665632, "grad_norm": 26.874052047729492, "kl": 33.0, "learning_rate": 2.3339480714259528e-06, "loss": 0.4964, "num_tokens": 289692251.0, "reward": 0.015000000596046448, "reward_std": 0.03265853598713875, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1477 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3611.0, "completions/mean_length": 3340.962646484375, "completions/mean_terminated_length": 1219.666748046875, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.5733126454615981, "grad_norm": 65.2752685546875, "kl": 28.5625, "learning_rate": 2.331169726108951e-06, "loss": 0.5061, "num_tokens": 290011232.0, "reward": 0.02500000037252903, "reward_std": 0.04035984352231026, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1478 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 3695.0, "completions/mean_length": 3120.949951171875, "completions/mean_terminated_length": 1095.84619140625, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.5737005430566331, "grad_norm": 25.763702392578125, "kl": 29.375, "learning_rate": 2.3283920421821194e-06, "loss": 0.4699, "num_tokens": 290322092.0, "reward": 0.0312500037252903, "reward_std": 0.06390555948019028, "rewards/format_reward/mean": 0.1875, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 1479 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1573.0, "completions/mean_length": 3969.337646484375, "completions/mean_terminated_length": 718.3333740234375, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.574088440651668, "grad_norm": 31.151620864868164, "kl": 42.3125, "learning_rate": 2.325615024031803e-06, "loss": 0.4393, "num_tokens": 290686647.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1480 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.825, "completions/max_length": 4096.0, "completions/max_terminated_length": 2634.0, "completions/mean_length": 3546.650146484375, "completions/mean_terminated_length": 956.857177734375, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.5744763382467029, "grad_norm": 25.644817352294922, "kl": 34.9375, "learning_rate": 2.322838676043289e-06, "loss": 0.3718, "num_tokens": 291011779.0, "reward": 0.0062500000931322575, "reward_std": 0.015235701575875282, "rewards/format_reward/mean": 0.0625, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1481 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3807.0, "completions/mean_length": 3203.337646484375, "completions/mean_terminated_length": 991.0869750976562, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "epoch": 0.5748642358417377, "grad_norm": 11.573705673217773, "kl": 28.25, "learning_rate": 2.3200630026008118e-06, "loss": 0.4328, "num_tokens": 291319862.0, "reward": 0.011250000447034836, "reward_std": 0.024493901059031487, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1482 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 4045.0, "completions/mean_length": 3062.012451171875, "completions/mean_terminated_length": 1511.03125, "completions/min_length": 238.0, "completions/min_terminated_length": 238.0, "epoch": 0.5752521334367727, "grad_norm": 14.27853012084961, "kl": 16.265625, "learning_rate": 2.3172880080875387e-06, "loss": 0.1841, "num_tokens": 291603487.0, "reward": 0.011250000447034836, "reward_std": 0.0209575854241848, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1483 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4002.0, "completions/mean_length": 3300.41259765625, "completions/mean_terminated_length": 1550.1199951171875, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.5756400310318076, "grad_norm": 53.95165252685547, "kl": 6.2734375, "learning_rate": 2.3145136968855646e-06, "loss": 0.2669, "num_tokens": 291904648.0, "reward": 0.016249999403953552, "reward_std": 0.029839128255844116, "rewards/format_reward/mean": 0.16249999403953552, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1484 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 4030.0, "completions/mean_length": 3378.175048828125, "completions/mean_terminated_length": 1703.25, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.5760279286268425, "grad_norm": 43.960811614990234, "kl": 4.04296875, "learning_rate": 2.3117400733759054e-06, "loss": 0.1835, "num_tokens": 292225238.0, "reward": 0.012500000186264515, "reward_std": 0.022598328068852425, "rewards/format_reward/mean": 0.125, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1485 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3608.0, "completions/mean_length": 3042.16259765625, "completions/mean_terminated_length": 1376.4193115234375, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.5764158262218774, "grad_norm": 23.598112106323242, "kl": 5.2421875, "learning_rate": 2.3089671419384913e-06, "loss": 0.3019, "num_tokens": 292512771.0, "reward": 0.015000000596046448, "reward_std": 0.032111361622810364, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1486 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3391.0, "completions/mean_length": 3566.699951171875, "completions/mean_terminated_length": 1867.368408203125, "completions/min_length": 350.0, "completions/min_terminated_length": 350.0, "epoch": 0.5768037238169124, "grad_norm": 2.3559577465057373, "kl": 3.84375, "learning_rate": 2.306194906952159e-06, "loss": 0.0974, "num_tokens": 292841891.0, "reward": 0.005000000353902578, "reward_std": 0.011700167320668697, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1487 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3472.0, "completions/mean_length": 2791.425048828125, "completions/mean_terminated_length": 1827.1739501953125, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.5771916214119472, "grad_norm": 14.121095657348633, "kl": 5.9375, "learning_rate": 2.303423372794647e-06, "loss": 0.2399, "num_tokens": 293104541.0, "reward": 0.026250002905726433, "reward_std": 0.055134255439043045, "rewards/format_reward/mean": 0.16249999403953552, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944272994995117, "step": 1488 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4090.0, "completions/mean_length": 3530.712646484375, "completions/mean_terminated_length": 2129.78271484375, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "epoch": 0.5775795190069821, "grad_norm": 6.080544471740723, "kl": 4.65625, "learning_rate": 2.300652543842585e-06, "loss": 0.1836, "num_tokens": 293433670.0, "reward": 0.011249999515712261, "reward_std": 0.024493902921676636, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1489 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4072.0, "completions/mean_length": 3844.875, "completions/mean_terminated_length": 1863.77783203125, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.5779674166020171, "grad_norm": 3.9858295917510986, "kl": 6.6796875, "learning_rate": 2.2978824244714906e-06, "loss": 0.1461, "num_tokens": 293794460.0, "reward": 0.012500000186264515, "reward_std": 0.024116460233926773, "rewards/format_reward/mean": 0.125, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1490 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 1628.0, "completions/mean_length": 3832.375, "completions/mean_terminated_length": 581.0, "completions/min_length": 98.0, "completions/min_terminated_length": 98.0, "epoch": 0.578355314197052, "grad_norm": 4.0088934898376465, "kl": 7.078125, "learning_rate": 2.2951130190557603e-06, "loss": 0.209, "num_tokens": 294143810.0, "reward": 0.02250000275671482, "reward_std": 0.058755673468112946, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 1491 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2512.0, "completions/mean_length": 3908.837646484375, "completions/mean_terminated_length": 1101.4000244140625, "completions/min_length": 284.0, "completions/min_terminated_length": 284.0, "epoch": 0.5787432117920869, "grad_norm": 1.9135760068893433, "kl": 7.90625, "learning_rate": 2.2923443319686627e-06, "loss": 0.1457, "num_tokens": 294505717.0, "reward": 0.007500000298023224, "reward_std": 0.01578209362924099, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1492 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3135.0, "completions/mean_length": 3631.837646484375, "completions/mean_terminated_length": 720.2727661132812, "completions/min_length": 110.0, "completions/min_terminated_length": 110.0, "epoch": 0.5791311093871218, "grad_norm": 4.421146869659424, "kl": 11.53125, "learning_rate": 2.2895763675823347e-06, "loss": 0.3123, "num_tokens": 294849008.0, "reward": 0.0325000025331974, "reward_std": 0.07587238401174545, "rewards/format_reward/mean": 0.17499999701976776, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.07647287845611572, "step": 1493 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1658.0, "completions/mean_length": 3638.175048828125, "completions/mean_terminated_length": 433.3999938964844, "completions/min_length": 123.0, "completions/min_terminated_length": 123.0, "epoch": 0.5795190069821567, "grad_norm": 4.900472164154053, "kl": 10.8125, "learning_rate": 2.286809130267768e-06, "loss": 0.3178, "num_tokens": 295183750.0, "reward": 0.015000000596046448, "reward_std": 0.029669394716620445, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1494 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1071.0, "completions/mean_length": 3312.75, "completions/mean_terminated_length": 410.1176452636719, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.5799069045771916, "grad_norm": 7.772348880767822, "kl": 12.859375, "learning_rate": 2.2840426243948077e-06, "loss": 0.3948, "num_tokens": 295496610.0, "reward": 0.027500001713633537, "reward_std": 0.039904993027448654, "rewards/format_reward/mean": 0.22499999403953552, "rewards/format_reward/std": 0.4202169179916382, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.044721364974975586, "step": 1495 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.825, "completions/max_length": 4096.0, "completions/max_terminated_length": 1961.0, "completions/mean_length": 3491.650146484375, "completions/mean_terminated_length": 642.5714721679688, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.5802948021722265, "grad_norm": 4.99086856842041, "kl": 13.484375, "learning_rate": 2.2812768543321455e-06, "loss": 0.3419, "num_tokens": 295810982.0, "reward": 0.02250000089406967, "reward_std": 0.03673219308257103, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1496 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2744.0, "completions/mean_length": 2867.525146484375, "completions/mean_terminated_length": 707.1034545898438, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.5806826997672614, "grad_norm": 31.663433074951172, "kl": 15.890625, "learning_rate": 2.2785118244473088e-06, "loss": 0.5585, "num_tokens": 296078104.0, "reward": 0.04500000178813934, "reward_std": 0.0666201040148735, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 1497 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2288.0, "completions/mean_length": 3070.675048828125, "completions/mean_terminated_length": 529.6521606445312, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.5810705973622964, "grad_norm": 26.379173278808594, "kl": 13.234375, "learning_rate": 2.2757475391066587e-06, "loss": 0.4969, "num_tokens": 296377942.0, "reward": 0.027500001713633537, "reward_std": 0.04166033864021301, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1498 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 2332.0, "completions/mean_length": 2943.3125, "completions/mean_terminated_length": 549.2692260742188, "completions/min_length": 193.0, "completions/min_terminated_length": 193.0, "epoch": 0.5814584949573313, "grad_norm": 8.494921684265137, "kl": 22.34375, "learning_rate": 2.272984002675378e-06, "loss": 0.6331, "num_tokens": 296665543.0, "reward": 0.04125000163912773, "reward_std": 0.0704910010099411, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944272994995117, "step": 1499 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3142.0, "completions/mean_length": 3678.112548828125, "completions/mean_terminated_length": 752.9000244140625, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "epoch": 0.5818463925523661, "grad_norm": 9.302452087402344, "kl": 29.1875, "learning_rate": 2.2702212195174667e-06, "loss": 0.4683, "num_tokens": 297009472.0, "reward": 0.015000000596046448, "reward_std": 0.03265853598713875, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1500 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 832.0, "completions/mean_length": 3076.8125, "completions/mean_terminated_length": 389.8636474609375, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.5822342901474011, "grad_norm": 3.5495216846466064, "kl": 27.9375, "learning_rate": 2.2674591939957393e-06, "loss": 0.5728, "num_tokens": 297303577.0, "reward": 0.023749999701976776, "reward_std": 0.041915953159332275, "rewards/format_reward/mean": 0.21250000596046448, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360678762197495, "step": 1501 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 2268.0, "completions/mean_length": 3112.28759765625, "completions/mean_terminated_length": 518.8636474609375, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.582622187742436, "grad_norm": 3.408179998397827, "kl": 29.375, "learning_rate": 2.264697930471809e-06, "loss": 0.6142, "num_tokens": 297603088.0, "reward": 0.03125, "reward_std": 0.043895378708839417, "rewards/format_reward/mean": 0.2874999940395355, "rewards/format_reward/std": 0.45545217394828796, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1502 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2893.0, "completions/mean_length": 3079.03759765625, "completions/mean_terminated_length": 558.7391357421875, "completions/min_length": 217.0, "completions/min_terminated_length": 217.0, "epoch": 0.5830100853374709, "grad_norm": 3.1438417434692383, "kl": 25.9375, "learning_rate": 2.2619374333060885e-06, "loss": 0.4853, "num_tokens": 297892211.0, "reward": 0.055000003427267075, "reward_std": 0.09255288541316986, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.126265749335289, "step": 1503 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3801.0, "completions/mean_length": 2684.012451171875, "completions/mean_terminated_length": 673.0, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.5833979829325058, "grad_norm": 2.3101658821105957, "kl": 36.1875, "learning_rate": 2.2591777068577802e-06, "loss": 0.7982, "num_tokens": 298150900.0, "reward": 0.08624999970197678, "reward_std": 0.12029910087585449, "rewards/format_reward/mean": 0.36250001192092896, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.20500075817108154, "step": 1504 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3129.0, "completions/mean_length": 2851.775146484375, "completions/mean_terminated_length": 663.6551513671875, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.5837858805275408, "grad_norm": 2.9271345138549805, "kl": 30.4375, "learning_rate": 2.2564187554848667e-06, "loss": 0.6424, "num_tokens": 298424850.0, "reward": 0.04750000312924385, "reward_std": 0.07999961823225021, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.08282969892024994, "step": 1505 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1974.0, "completions/mean_length": 2796.625, "completions/mean_terminated_length": 511.5172424316406, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.5841737781225756, "grad_norm": 3.260845422744751, "kl": 31.125, "learning_rate": 2.2536605835441115e-06, "loss": 0.6468, "num_tokens": 298692596.0, "reward": 0.042500000447034836, "reward_std": 0.06174199655652046, "rewards/format_reward/mean": 0.3499999940395355, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.04974619299173355, "step": 1506 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1783.0, "completions/mean_length": 2818.25, "completions/mean_terminated_length": 571.1724243164062, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.5845616757176105, "grad_norm": 9.711002349853516, "kl": 22.28125, "learning_rate": 2.250903195391043e-06, "loss": 0.6578, "num_tokens": 298963696.0, "reward": 0.032499998807907104, "reward_std": 0.04565637931227684, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1507 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 2272.0, "completions/mean_length": 2371.262451171875, "completions/mean_terminated_length": 464.97369384765625, "completions/min_length": 110.0, "completions/min_terminated_length": 110.0, "epoch": 0.5849495733126454, "grad_norm": 11.18257999420166, "kl": 24.875, "learning_rate": 2.2481465953799547e-06, "loss": 0.7224, "num_tokens": 299200925.0, "reward": 0.048750001937150955, "reward_std": 0.052151549607515335, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472136124968529, "step": 1508 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 652.0, "completions/mean_length": 1782.8375244140625, "completions/mean_terminated_length": 394.94000244140625, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.5853374709076804, "grad_norm": 18.808752059936523, "kl": 24.5625, "learning_rate": 2.245390787863893e-06, "loss": 0.7337, "num_tokens": 299390696.0, "reward": 0.06000000238418579, "reward_std": 0.053066909313201904, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360678762197495, "step": 1509 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2866.0, "completions/mean_length": 2046.7249755859375, "completions/mean_terminated_length": 607.872314453125, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.5857253685027153, "grad_norm": 8.102198600769043, "kl": 24.875, "learning_rate": 2.2426357771946545e-06, "loss": 0.6536, "num_tokens": 299596170.0, "reward": 0.05375000461935997, "reward_std": 0.04700635001063347, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1510 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3259.0, "completions/mean_length": 2187.712646484375, "completions/mean_terminated_length": 545.6976928710938, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.5861132660977502, "grad_norm": 4.691996097564697, "kl": 33.625, "learning_rate": 2.239881567722777e-06, "loss": 0.7761, "num_tokens": 299826059.0, "reward": 0.048750001937150955, "reward_std": 0.05188746377825737, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360680624842644, "step": 1511 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1711.0, "completions/mean_length": 1658.0999755859375, "completions/mean_terminated_length": 484.2962951660156, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.5865011636927852, "grad_norm": 8.206478118896484, "kl": 36.34375, "learning_rate": 2.2371281637975327e-06, "loss": 0.8941, "num_tokens": 300012875.0, "reward": 0.057500001043081284, "reward_std": 0.04970111325383186, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1512 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3089.0, "completions/mean_length": 1534.0999755859375, "completions/mean_terminated_length": 500.35089111328125, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.58688906128782, "grad_norm": 5.565590858459473, "kl": 47.375, "learning_rate": 2.2343755697669243e-06, "loss": 1.039, "num_tokens": 300178563.0, "reward": 0.06499999761581421, "reward_std": 0.04738222062587738, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1513 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1075.0, "completions/mean_length": 1393.0, "completions/mean_terminated_length": 430.9152526855469, "completions/min_length": 131.0, "completions/min_terminated_length": 131.0, "epoch": 0.5872769588828549, "grad_norm": 5.584439754486084, "kl": 40.875, "learning_rate": 2.2316237899776723e-06, "loss": 0.7091, "num_tokens": 300343211.0, "reward": 0.06000000238418579, "reward_std": 0.04494025930762291, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1514 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 816.0, "completions/mean_length": 1140.5125732421875, "completions/mean_terminated_length": 458.4769287109375, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.5876648564778898, "grad_norm": 4.598094940185547, "kl": 44.25, "learning_rate": 2.228872828775211e-06, "loss": 0.9341, "num_tokens": 300482348.0, "reward": 0.07250000536441803, "reward_std": 0.0373707041144371, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1515 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1410.0, "completions/mean_length": 1007.6875, "completions/mean_terminated_length": 462.6911926269531, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.5880527540729248, "grad_norm": 5.032867431640625, "kl": 35.76025390625, "learning_rate": 2.2261226905036857e-06, "loss": 0.7279, "num_tokens": 300606571.0, "reward": 0.07000000774860382, "reward_std": 0.04591278359293938, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1516 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 1038.0, "completions/mean_length": 1468.3375244140625, "completions/mean_terminated_length": 471.637939453125, "completions/min_length": 98.0, "completions/min_terminated_length": 98.0, "epoch": 0.5884406516679597, "grad_norm": 20.988733291625977, "kl": 57.0, "learning_rate": 2.2233733795059394e-06, "loss": 0.8141, "num_tokens": 300758182.0, "reward": 0.07375001162290573, "reward_std": 0.0631822943687439, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944271504878998, "step": 1517 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 754.0, "completions/mean_length": 1083.3499755859375, "completions/mean_terminated_length": 444.30303955078125, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.5888285492629945, "grad_norm": 4.439300060272217, "kl": 40.8125, "learning_rate": 2.2206249001235093e-06, "loss": 0.8808, "num_tokens": 300896290.0, "reward": 0.08000000566244125, "reward_std": 0.035014621913433075, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1518 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 981.0, "completions/mean_length": 1219.6624755859375, "completions/mean_terminated_length": 500.578125, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.5892164468580294, "grad_norm": 7.441005229949951, "kl": 41.125, "learning_rate": 2.21787725669662e-06, "loss": 0.7272, "num_tokens": 301042351.0, "reward": 0.07750000059604645, "reward_std": 0.03964371979236603, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360680624842644, "step": 1519 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 958.0, "completions/mean_length": 1068.3499755859375, "completions/mean_terminated_length": 480.8955078125, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.5896043444530644, "grad_norm": 4.091614723205566, "kl": 24.890625, "learning_rate": 2.2151304535641748e-06, "loss": 0.475, "num_tokens": 301183227.0, "reward": 0.08000000566244125, "reward_std": 0.03257265314459801, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1520 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 692.0, "completions/mean_length": 1255.175048828125, "completions/mean_terminated_length": 430.4193420410156, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.5899922420480993, "grad_norm": 3.963103771209717, "kl": 39.0625, "learning_rate": 2.212384495063752e-06, "loss": 0.7654, "num_tokens": 301328945.0, "reward": 0.07625000923871994, "reward_std": 0.049669429659843445, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472135752439499, "step": 1521 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1190.0, "completions/mean_length": 1153.4000244140625, "completions/mean_terminated_length": 474.3384704589844, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.5903801396431342, "grad_norm": 11.485233306884766, "kl": 32.46875, "learning_rate": 2.209639385531593e-06, "loss": 0.7814, "num_tokens": 301467401.0, "reward": 0.07625000923871994, "reward_std": 0.040190111845731735, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1522 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 866.0, "completions/mean_length": 1398.7625732421875, "completions/mean_terminated_length": 499.683349609375, "completions/min_length": 243.0, "completions/min_terminated_length": 243.0, "epoch": 0.5907680372381692, "grad_norm": 6.0459089279174805, "kl": 38.46875, "learning_rate": 2.206895129302602e-06, "loss": 0.6734, "num_tokens": 301632638.0, "reward": 0.09000000357627869, "reward_std": 0.07581327110528946, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.1111220121383667, "step": 1523 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 802.0, "completions/mean_length": 1576.4124755859375, "completions/mean_terminated_length": 363.27777099609375, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.591155934833204, "grad_norm": 7.444712162017822, "kl": 53.875, "learning_rate": 2.2041517307103337e-06, "loss": 1.1719, "num_tokens": 301807247.0, "reward": 0.061250001192092896, "reward_std": 0.05024750158190727, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1524 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 755.0, "completions/mean_length": 1042.3125, "completions/mean_terminated_length": 449.80596923828125, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.5915438324282389, "grad_norm": 8.146140098571777, "kl": 30.5625, "learning_rate": 2.2014091940869885e-06, "loss": 0.6768, "num_tokens": 301943920.0, "reward": 0.07375001162290573, "reward_std": 0.03720096871256828, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1525 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1129.0, "completions/mean_length": 1213.9625244140625, "completions/mean_terminated_length": 436.2698669433594, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.5919317300232738, "grad_norm": 9.51911735534668, "kl": 38.0, "learning_rate": 2.198667523763407e-06, "loss": 0.8626, "num_tokens": 302084253.0, "reward": 0.07250000536441803, "reward_std": 0.04254698008298874, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1526 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 653.0, "completions/mean_length": 867.9249877929688, "completions/mean_terminated_length": 406.77142333984375, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.5923196276183088, "grad_norm": 11.697940826416016, "kl": 26.609375, "learning_rate": 2.1959267240690603e-06, "loss": 0.6618, "num_tokens": 302206311.0, "reward": 0.08250000327825546, "reward_std": 0.03429849445819855, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1527 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 927.0, "completions/mean_length": 994.7250366210938, "completions/mean_terminated_length": 500.31884765625, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.5927075252133437, "grad_norm": 3.597548723220825, "kl": 29.796875, "learning_rate": 2.193186799332045e-06, "loss": 0.6347, "num_tokens": 302341297.0, "reward": 0.08125000447034836, "reward_std": 0.031479086726903915, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1528 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 783.0, "completions/mean_length": 803.2999877929688, "completions/mean_terminated_length": 437.4444580078125, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.5930954228083786, "grad_norm": 4.395090103149414, "kl": 27.40625, "learning_rate": 2.1904477538790776e-06, "loss": 0.6578, "num_tokens": 302449865.0, "reward": 0.08375000208616257, "reward_std": 0.030762961134314537, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1529 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3057.0, "completions/mean_length": 1042.5625, "completions/mean_terminated_length": 450.1044616699219, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.5934833204034135, "grad_norm": 12.706833839416504, "kl": 45.78125, "learning_rate": 2.1877095920354817e-06, "loss": 0.8871, "num_tokens": 302569150.0, "reward": 0.08500000089406967, "reward_std": 0.061397042125463486, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944272249937057, "step": 1530 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 789.0, "completions/mean_length": 1119.5250244140625, "completions/mean_terminated_length": 488.1515197753906, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "epoch": 0.5938712179984484, "grad_norm": 8.301674842834473, "kl": 44.40625, "learning_rate": 2.1849723181251923e-06, "loss": 0.8769, "num_tokens": 302708408.0, "reward": 0.07375000417232513, "reward_std": 0.0426320806145668, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1531 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 967.0, "completions/mean_length": 1139.8125, "completions/mean_terminated_length": 457.6153869628906, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.5942591155934833, "grad_norm": 17.829025268554688, "kl": 59.0, "learning_rate": 2.1822359364707364e-06, "loss": 1.0582, "num_tokens": 302849273.0, "reward": 0.07500001043081284, "reward_std": 0.042346369475126266, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360678762197495, "step": 1532 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 772.0, "completions/mean_length": 907.5, "completions/mean_terminated_length": 399.18841552734375, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.5946470131885182, "grad_norm": 9.238690376281738, "kl": 43.59375, "learning_rate": 2.1795004513932343e-06, "loss": 0.8438, "num_tokens": 302970329.0, "reward": 0.07875000685453415, "reward_std": 0.03257187083363533, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1533 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 814.0, "completions/mean_length": 964.0250244140625, "completions/mean_terminated_length": 411.3235168457031, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.5950349107835532, "grad_norm": 5.436985969543457, "kl": 41.6875, "learning_rate": 2.176765867212391e-06, "loss": 0.8261, "num_tokens": 303095707.0, "reward": 0.07750000804662704, "reward_std": 0.03964371979236603, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1534 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 903.0, "completions/mean_length": 1514.2125244140625, "completions/mean_terminated_length": 407.7321472167969, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.5954228083785881, "grad_norm": 12.529389381408691, "kl": 64.1875, "learning_rate": 2.174032188246487e-06, "loss": 1.122, "num_tokens": 303261124.0, "reward": 0.061250001192092896, "reward_std": 0.046165578067302704, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1535 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 717.0, "completions/mean_length": 1130.4625244140625, "completions/mean_terminated_length": 330.2381286621094, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.595810705973623, "grad_norm": 3.5876917839050293, "kl": 45.59375, "learning_rate": 2.171299418812376e-06, "loss": 0.9137, "num_tokens": 303404665.0, "reward": 0.08875000476837158, "reward_std": 0.06752756983041763, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 1536 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 783.0, "completions/mean_length": 1334.9375, "completions/mean_terminated_length": 352.1864318847656, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.5961986035686578, "grad_norm": 7.428382873535156, "kl": 46.15625, "learning_rate": 2.168567563225472e-06, "loss": 0.8262, "num_tokens": 303555572.0, "reward": 0.06750000268220901, "reward_std": 0.04145262762904167, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1537 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 720.0, "completions/mean_length": 1839.4124755859375, "completions/mean_terminated_length": 411.7755126953125, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.5965865011636928, "grad_norm": 4.712552070617676, "kl": 46.9375, "learning_rate": 2.1658366257997506e-06, "loss": 1.0203, "num_tokens": 303762861.0, "reward": 0.061250001192092896, "reward_std": 0.03884093090891838, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1538 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 597.0, "completions/mean_length": 1722.362548828125, "completions/mean_terminated_length": 372.6470642089844, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.5969743987587277, "grad_norm": 10.694636344909668, "kl": 40.9375, "learning_rate": 2.1631066108477327e-06, "loss": 0.802, "num_tokens": 303944482.0, "reward": 0.07250000536441803, "reward_std": 0.05803673341870308, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.07044164091348648, "step": 1539 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 775.0, "completions/mean_length": 1423.9375, "completions/mean_terminated_length": 410.39654541015625, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.5973622963537626, "grad_norm": 36.4913215637207, "kl": 19.609375, "learning_rate": 2.1603775226804848e-06, "loss": 0.7115, "num_tokens": 304105245.0, "reward": 0.07125000655651093, "reward_std": 0.03892681002616882, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1540 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 607.0, "completions/mean_length": 1807.9375, "completions/mean_terminated_length": 360.38775634765625, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.5977501939487975, "grad_norm": 27.376285552978516, "kl": 26.75, "learning_rate": 2.15764936560761e-06, "loss": 0.8359, "num_tokens": 304291920.0, "reward": 0.06625000387430191, "reward_std": 0.045136917382478714, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.03142199665307999, "step": 1541 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 603.0, "completions/mean_length": 1849.737548828125, "completions/mean_terminated_length": 352.22918701171875, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.5981380915438325, "grad_norm": 32.15016555786133, "kl": 25.125, "learning_rate": 2.15492214393724e-06, "loss": 0.8431, "num_tokens": 304489739.0, "reward": 0.0625, "reward_std": 0.04128367826342583, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1542 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 556.0, "completions/mean_length": 2502.262451171875, "completions/mean_terminated_length": 346.0294189453125, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.5985259891388673, "grad_norm": 3.127028703689575, "kl": 38.375, "learning_rate": 2.152195861976032e-06, "loss": 0.8076, "num_tokens": 304728936.0, "reward": 0.042500000447034836, "reward_std": 0.03682352602481842, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1543 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 558.0, "completions/mean_length": 2075.387451171875, "completions/mean_terminated_length": 336.7209167480469, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.5989138867339022, "grad_norm": 12.89581298828125, "kl": 49.1875, "learning_rate": 2.149470524029156e-06, "loss": 1.2358, "num_tokens": 304933815.0, "reward": 0.07250000536441803, "reward_std": 0.044649481773376465, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.0603785514831543, "step": 1544 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 536.0, "completions/mean_length": 1291.0374755859375, "completions/mean_terminated_length": 356.0500183105469, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.5993017843289372, "grad_norm": 7.498045921325684, "kl": 34.4375, "learning_rate": 2.146746134400291e-06, "loss": 0.7834, "num_tokens": 305096018.0, "reward": 0.07500000298023224, "reward_std": 0.03067629411816597, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1545 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 709.0, "completions/mean_length": 1386.0250244140625, "completions/mean_terminated_length": 358.10345458984375, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.5996896819239721, "grad_norm": 4.987734794616699, "kl": 52.6875, "learning_rate": 2.1440226973916233e-06, "loss": 1.1528, "num_tokens": 305261340.0, "reward": 0.07000000774860382, "reward_std": 0.04002038389444351, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1546 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1047.0, "completions/mean_length": 983.7750244140625, "completions/mean_terminated_length": 379.9104309082031, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.600077579519007, "grad_norm": 5.727750778198242, "kl": 36.74267578125, "learning_rate": 2.1413002173038294e-06, "loss": 0.6847, "num_tokens": 305383946.0, "reward": 0.08375000208616257, "reward_std": 0.02630280889570713, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1547 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 629.0, "completions/mean_length": 1466.5625, "completions/mean_terminated_length": 339.6607360839844, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.6004654771140419, "grad_norm": 20.92728614807129, "kl": 73.3125, "learning_rate": 2.1385786984360777e-06, "loss": 1.3831, "num_tokens": 305557159.0, "reward": 0.07000000774860382, "reward_std": 0.03972678631544113, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1548 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 581.0, "completions/mean_length": 721.4874877929688, "completions/mean_terminated_length": 346.5416564941406, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.6008533747090768, "grad_norm": 7.241044521331787, "kl": 39.3125, "learning_rate": 2.135858145086019e-06, "loss": 0.851, "num_tokens": 305653926.0, "reward": 0.08749999850988388, "reward_std": 0.027482260018587112, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1549 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 676.0, "completions/mean_length": 1174.1875, "completions/mean_terminated_length": 325.9193420410156, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.6012412723041117, "grad_norm": 11.960846900939941, "kl": 69.875, "learning_rate": 2.1331385615497756e-06, "loss": 1.4785, "num_tokens": 305801229.0, "reward": 0.07625000923871994, "reward_std": 0.04111394286155701, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1550 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 508.0, "completions/mean_length": 1056.425048828125, "completions/mean_terminated_length": 296.53125, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.6016291698991466, "grad_norm": 12.01406192779541, "kl": 60.0, "learning_rate": 2.1304199521219443e-06, "loss": 1.2305, "num_tokens": 305935335.0, "reward": 0.07750000804662704, "reward_std": 0.035014618188142776, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1551 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 513.0, "completions/mean_length": 920.0250244140625, "completions/mean_terminated_length": 303.7910461425781, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.6020170674941815, "grad_norm": 10.838774681091309, "kl": 40.0, "learning_rate": 2.127702321095579e-06, "loss": 0.9853, "num_tokens": 306050529.0, "reward": 0.08250000327825546, "reward_std": 0.032111361622810364, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1552 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 457.0, "completions/mean_length": 1159.7750244140625, "completions/mean_terminated_length": 307.32257080078125, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.6024049650892165, "grad_norm": 7.867642879486084, "kl": 49.65625, "learning_rate": 2.1249856727621887e-06, "loss": 1.1006, "num_tokens": 306188999.0, "reward": 0.07500000298023224, "reward_std": 0.042801812291145325, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1553 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 481.0, "completions/mean_length": 972.2625122070312, "completions/mean_terminated_length": 309.6515197753906, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.6027928626842514, "grad_norm": 6.564205169677734, "kl": 50.77734375, "learning_rate": 2.1222700114117344e-06, "loss": 1.1418, "num_tokens": 306318708.0, "reward": 0.08250000327825546, "reward_std": 0.033374663442373276, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1554 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 588.0, "completions/mean_length": 741.5250244140625, "completions/mean_terminated_length": 262.3143005371094, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.6031807602792862, "grad_norm": 3.9014217853546143, "kl": 33.90625, "learning_rate": 2.1195553413326147e-06, "loss": 0.6907, "num_tokens": 306425022.0, "reward": 0.08375000208616257, "reward_std": 0.03265775367617607, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1555 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 577.0, "completions/mean_length": 1154.625, "completions/mean_terminated_length": 300.6773986816406, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.6035686578743211, "grad_norm": 14.149297714233398, "kl": 57.5625, "learning_rate": 2.1168416668116646e-06, "loss": 1.0184, "num_tokens": 306567400.0, "reward": 0.07124999910593033, "reward_std": 0.044272035360336304, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1556 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 438.0, "completions/mean_length": 1001.7625122070312, "completions/mean_terminated_length": 287.70770263671875, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.6039565554693561, "grad_norm": 16.84815216064453, "kl": 60.9375, "learning_rate": 2.1141289921341462e-06, "loss": 1.1044, "num_tokens": 306691597.0, "reward": 0.07750000804662704, "reward_std": 0.03909654542803764, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1557 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1732.0, "completions/mean_length": 1077.0, "completions/mean_terminated_length": 322.25, "completions/min_length": 118.0, "completions/min_terminated_length": 118.0, "epoch": 0.604344453064391, "grad_norm": 16.90736961364746, "kl": 63.625, "learning_rate": 2.1114173215837436e-06, "loss": 1.1846, "num_tokens": 306835325.0, "reward": 0.07874999940395355, "reward_std": 0.033119045197963715, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1558 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 574.0, "completions/mean_length": 973.7750244140625, "completions/mean_terminated_length": 311.48486328125, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.6047323506594259, "grad_norm": 9.992039680480957, "kl": 54.625, "learning_rate": 2.1087066594425543e-06, "loss": 1.0536, "num_tokens": 306965803.0, "reward": 0.08000000566244125, "reward_std": 0.03375132009387016, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1559 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 451.0, "completions/mean_length": 742.6375122070312, "completions/mean_terminated_length": 263.5857238769531, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.6051202482544609, "grad_norm": 8.934617042541504, "kl": 42.0, "learning_rate": 2.1059970099910835e-06, "loss": 0.9732, "num_tokens": 307071870.0, "reward": 0.08625000715255737, "reward_std": 0.029123002663254738, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1560 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 469.0, "completions/mean_length": 1019.4500122070312, "completions/mean_terminated_length": 250.3125, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.6055081458494957, "grad_norm": 5.208062171936035, "kl": 59.8125, "learning_rate": 2.1032883775082387e-06, "loss": 1.2373, "num_tokens": 307197818.0, "reward": 0.07625000923871994, "reward_std": 0.03892681002616882, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1561 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 588.0, "completions/mean_length": 1561.7625732421875, "completions/mean_terminated_length": 270.73583984375, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.6058960434445306, "grad_norm": 10.015835762023926, "kl": 68.0625, "learning_rate": 2.1005807662713192e-06, "loss": 1.3035, "num_tokens": 307371807.0, "reward": 0.06750000268220901, "reward_std": 0.04309258982539177, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1562 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 478.0, "completions/mean_length": 1257.4375, "completions/mean_terminated_length": 247.10169982910156, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.6062839410395655, "grad_norm": 2.519930124282837, "kl": 43.0, "learning_rate": 2.0978741805560134e-06, "loss": 0.8628, "num_tokens": 307529818.0, "reward": 0.07000000029802322, "reward_std": 0.04309258610010147, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1563 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 385.0, "completions/mean_length": 997.1500244140625, "completions/mean_terminated_length": 222.4375, "completions/min_length": 98.0, "completions/min_terminated_length": 98.0, "epoch": 0.6066718386346005, "grad_norm": 21.30571174621582, "kl": 44.0625, "learning_rate": 2.095168624636389e-06, "loss": 1.1525, "num_tokens": 307656718.0, "reward": 0.07874999940395355, "reward_std": 0.0334957055747509, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1564 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 522.0, "completions/mean_length": 1115.3125, "completions/mean_terminated_length": 249.95159912109375, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.6070597362296354, "grad_norm": 14.148331642150879, "kl": 43.984375, "learning_rate": 2.092464102784888e-06, "loss": 1.0118, "num_tokens": 307790279.0, "reward": 0.07500000298023224, "reward_std": 0.036654576659202576, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1565 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 438.0, "completions/mean_length": 1249.0875244140625, "completions/mean_terminated_length": 235.7796630859375, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.6074476338246703, "grad_norm": 12.522868156433105, "kl": 57.4375, "learning_rate": 2.089760619272321e-06, "loss": 1.3706, "num_tokens": 307941950.0, "reward": 0.07000000774860382, "reward_std": 0.04145262762904167, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1566 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 613.0, "completions/mean_length": 1654.75, "completions/mean_terminated_length": 266.5882568359375, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.6078355314197051, "grad_norm": 16.43212127685547, "kl": 79.0, "learning_rate": 2.0870581783678563e-06, "loss": 1.5448, "num_tokens": 308124074.0, "reward": 0.0625, "reward_std": 0.04473254829645157, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1567 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 860.0, "completions/mean_length": 1162.6624755859375, "completions/mean_terminated_length": 248.99998474121094, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.6082234290147401, "grad_norm": 4.607955455780029, "kl": 55.5, "learning_rate": 2.0843567843390163e-06, "loss": 1.1886, "num_tokens": 308270439.0, "reward": 0.07375001162290573, "reward_std": 0.03964293748140335, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1568 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 1312.0, "completions/mean_length": 1306.3875732421875, "completions/mean_terminated_length": 248.2586212158203, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.608611326609775, "grad_norm": 12.733635902404785, "kl": 76.625, "learning_rate": 2.081656441451672e-06, "loss": 1.5109, "num_tokens": 308426878.0, "reward": 0.06625000387430191, "reward_std": 0.0475519560277462, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1569 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1405.0, "completions/mean_length": 1264.1624755859375, "completions/mean_terminated_length": 256.2203369140625, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.6089992242048099, "grad_norm": 13.090742111206055, "kl": 76.8125, "learning_rate": 2.0789571539700325e-06, "loss": 1.5032, "num_tokens": 308576403.0, "reward": 0.07000000774860382, "reward_std": 0.04090702161192894, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1570 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 3947.0, "completions/mean_length": 1345.375, "completions/mean_terminated_length": 302.03448486328125, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.6093871217998449, "grad_norm": 7.2566351890563965, "kl": 65.5625, "learning_rate": 2.076258926156643e-06, "loss": 1.2477, "num_tokens": 308727241.0, "reward": 0.07375000417232513, "reward_std": 0.04035906121134758, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1571 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 398.0, "completions/mean_length": 1429.2000732421875, "completions/mean_terminated_length": 217.01817321777344, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.6097750193948798, "grad_norm": 6.3884382247924805, "kl": 68.5625, "learning_rate": 2.0735617622723727e-06, "loss": 1.357, "num_tokens": 308888193.0, "reward": 0.06750000268220901, "reward_std": 0.044903066009283066, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1572 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 455.0, "completions/mean_length": 1087.4375, "completions/mean_terminated_length": 213.98387145996094, "completions/min_length": 92.0, "completions/min_terminated_length": 92.0, "epoch": 0.6101629169899146, "grad_norm": 2.820256471633911, "kl": 49.0, "learning_rate": 2.0708656665764094e-06, "loss": 1.0476, "num_tokens": 309024140.0, "reward": 0.07500000298023224, "reward_std": 0.03429849445819855, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1573 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 382.0, "completions/mean_length": 1087.0250244140625, "completions/mean_terminated_length": 213.45159912109375, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.6105508145849495, "grad_norm": 25.098329544067383, "kl": 44.90625, "learning_rate": 2.0681706433262596e-06, "loss": 1.1561, "num_tokens": 309164574.0, "reward": 0.07500000298023224, "reward_std": 0.04427282139658928, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1574 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 332.0, "completions/mean_length": 1227.300048828125, "completions/mean_terminated_length": 206.23728942871094, "completions/min_length": 69.0, "completions/min_terminated_length": 69.0, "epoch": 0.6109387121799845, "grad_norm": 11.25967788696289, "kl": 44.9375, "learning_rate": 2.065476696777729e-06, "loss": 0.9896, "num_tokens": 309315318.0, "reward": 0.07250000536441803, "reward_std": 0.0389275960624218, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1575 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1000.0, "completions/mean_length": 993.7999877929688, "completions/mean_terminated_length": 218.25, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "epoch": 0.6113266097750194, "grad_norm": 15.018499374389648, "kl": 49.0, "learning_rate": 2.0627838311849313e-06, "loss": 1.2085, "num_tokens": 309449478.0, "reward": 0.07750000804662704, "reward_std": 0.040020376443862915, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1576 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 618.0, "completions/mean_length": 845.5, "completions/mean_terminated_length": 214.80596923828125, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.6117145073700543, "grad_norm": 4.9941725730896, "kl": 48.875, "learning_rate": 2.0600920508002663e-06, "loss": 1.0795, "num_tokens": 309570454.0, "reward": 0.08125000447034836, "reward_std": 0.034844886511564255, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1577 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1109.0, "completions/mean_length": 594.0499877929688, "completions/mean_terminated_length": 258.24658203125, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.6121024049650892, "grad_norm": 2.960097312927246, "kl": 41.5625, "learning_rate": 2.057401359874422e-06, "loss": 0.8596, "num_tokens": 309674706.0, "reward": 0.08500000089406967, "reward_std": 0.029669394716620445, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1578 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2166.0, "completions/mean_length": 708.5125122070312, "completions/mean_terminated_length": 279.1126708984375, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.6124903025601242, "grad_norm": 13.12483024597168, "kl": 65.0, "learning_rate": 2.054711762656369e-06, "loss": 1.2119, "num_tokens": 309778851.0, "reward": 0.08375000208616257, "reward_std": 0.029839128255844116, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1579 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2344.0, "completions/mean_length": 667.2000122070312, "completions/mean_terminated_length": 232.56336975097656, "completions/min_length": 78.0, "completions/min_terminated_length": 78.0, "epoch": 0.612878200155159, "grad_norm": 3.8463518619537354, "kl": 55.90625, "learning_rate": 2.0520232633933462e-06, "loss": 1.1181, "num_tokens": 309884587.0, "reward": 0.08874999731779099, "reward_std": 0.026935869827866554, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1580 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 801.0, "completions/mean_length": 632.0375366210938, "completions/mean_terminated_length": 192.94366455078125, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.6132660977501939, "grad_norm": 6.437378883361816, "kl": 41.125, "learning_rate": 2.0493358663308615e-06, "loss": 0.7822, "num_tokens": 309992910.0, "reward": 0.08624999970197678, "reward_std": 0.027651995420455933, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1581 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3556.0, "completions/mean_length": 279.26251220703125, "completions/mean_terminated_length": 230.94937133789062, "completions/min_length": 67.0, "completions/min_terminated_length": 67.0, "epoch": 0.6136539953452289, "grad_norm": 10.12276554107666, "kl": 16.828125, "learning_rate": 2.046649575712683e-06, "loss": 0.2089, "num_tokens": 310056235.0, "reward": 0.0950000062584877, "reward_std": 0.008711026050150394, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1582 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 2112.0, "completions/mean_length": 382.8125, "completions/mean_terminated_length": 187.38157653808594, "completions/min_length": 77.0, "completions/min_terminated_length": 77.0, "epoch": 0.6140418929402638, "grad_norm": 2.3057408332824707, "kl": 28.919921875, "learning_rate": 2.043964395780829e-06, "loss": 0.6006, "num_tokens": 310135804.0, "reward": 0.0950000062584877, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1583 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 273.0, "completions/mean_length": 305.01251220703125, "completions/mean_terminated_length": 157.3116912841797, "completions/min_length": 80.0, "completions/min_terminated_length": 80.0, "epoch": 0.6144297905352987, "grad_norm": 8.31735610961914, "kl": 18.4765625, "learning_rate": 2.0412803307755652e-06, "loss": 0.4449, "num_tokens": 310202445.0, "reward": 0.0937500074505806, "reward_std": 0.0176776684820652, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1584 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 382.0, "completions/mean_length": 405.5625, "completions/mean_terminated_length": 159.53334045410156, "completions/min_length": 64.0, "completions/min_terminated_length": 64.0, "epoch": 0.6148176881303336, "grad_norm": 3.97418475151062, "kl": 32.384765625, "learning_rate": 2.0385973849353955e-06, "loss": 0.645, "num_tokens": 310285210.0, "reward": 0.0937500074505806, "reward_std": 0.012246559374034405, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1585 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1018.0, "completions/max_terminated_length": 1018.0, "completions/mean_length": 178.91250610351562, "completions/mean_terminated_length": 178.91250610351562, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "epoch": 0.6152055857253685, "grad_norm": 6.202755451202393, "kl": 2.4921875, "learning_rate": 2.0359155624970607e-06, "loss": 0.0986, "num_tokens": 310347283.0, "reward": 0.0950000062584877, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1586 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 485.0, "completions/mean_length": 299.75, "completions/mean_terminated_length": 151.8441619873047, "completions/min_length": 68.0, "completions/min_terminated_length": 68.0, "epoch": 0.6155934833204034, "grad_norm": 8.94528865814209, "kl": 21.52099609375, "learning_rate": 2.0332348676955223e-06, "loss": 0.4761, "num_tokens": 310420799.0, "reward": 0.09375, "reward_std": 0.015235701575875282, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1587 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1371.0, "completions/mean_length": 415.38751220703125, "completions/mean_terminated_length": 170.01333618164062, "completions/min_length": 58.0, "completions/min_terminated_length": 58.0, "epoch": 0.6159813809154383, "grad_norm": 6.2959675788879395, "kl": 38.908203125, "learning_rate": 2.0305553047639625e-06, "loss": 0.7591, "num_tokens": 310506030.0, "reward": 0.08875000476837158, "reward_std": 0.024493902921676636, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1588 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 1040.0, "completions/mean_length": 371.8374938964844, "completions/mean_terminated_length": 175.82894897460938, "completions/min_length": 69.0, "completions/min_terminated_length": 69.0, "epoch": 0.6163692785104732, "grad_norm": 4.622107028961182, "kl": 28.6640625, "learning_rate": 2.0278768779337793e-06, "loss": 0.6342, "num_tokens": 310580241.0, "reward": 0.0925000011920929, "reward_std": 0.02121320180594921, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1589 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 898.0, "completions/mean_length": 427.3999938964844, "completions/mean_terminated_length": 182.82667541503906, "completions/min_length": 58.0, "completions/min_terminated_length": 58.0, "epoch": 0.6167571761055082, "grad_norm": 3.258878231048584, "kl": 32.40625, "learning_rate": 2.0251995914345724e-06, "loss": 0.6845, "num_tokens": 310670521.0, "reward": 0.08874999731779099, "reward_std": 0.024493901059031487, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1590 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 603.0, "completions/mean_length": 563.4625244140625, "completions/mean_terminated_length": 170.9583282470703, "completions/min_length": 53.0, "completions/min_terminated_length": 53.0, "epoch": 0.6171450737005431, "grad_norm": 8.514192581176758, "kl": 37.53125, "learning_rate": 2.0225234494941433e-06, "loss": 0.6826, "num_tokens": 310751750.0, "reward": 0.08125000447034836, "reward_std": 0.03429771587252617, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1591 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1362.0, "completions/mean_length": 890.1000366210938, "completions/mean_terminated_length": 210.06060791015625, "completions/min_length": 62.0, "completions/min_terminated_length": 62.0, "epoch": 0.6175329712955779, "grad_norm": 50.8883056640625, "kl": 69.6875, "learning_rate": 2.0198484563384855e-06, "loss": 1.0469, "num_tokens": 310861998.0, "reward": 0.08375000208616257, "reward_std": 0.030215784907341003, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1592 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 232.0, "completions/mean_length": 387.25, "completions/mean_terminated_length": 140.0, "completions/min_length": 75.0, "completions/min_terminated_length": 75.0, "epoch": 0.6179208688906129, "grad_norm": 6.969679355621338, "kl": 33.9052734375, "learning_rate": 2.0171746161917778e-06, "loss": 0.797, "num_tokens": 310944426.0, "reward": 0.09125000238418579, "reward_std": 0.022306768223643303, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1593 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 235.0, "completions/mean_length": 431.45001220703125, "completions/mean_terminated_length": 134.32432556152344, "completions/min_length": 58.0, "completions/min_terminated_length": 58.0, "epoch": 0.6183087664856478, "grad_norm": 2.263883590698242, "kl": 41.06640625, "learning_rate": 2.0145019332763797e-06, "loss": 0.7945, "num_tokens": 311023702.0, "reward": 0.08875000476837158, "reward_std": 0.024493901059031487, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1594 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 869.0, "completions/mean_length": 351.3999938964844, "completions/mean_terminated_length": 154.3157958984375, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.6186966640806827, "grad_norm": 5.040775775909424, "kl": 35.8046875, "learning_rate": 2.011830411812821e-06, "loss": 0.6747, "num_tokens": 311101566.0, "reward": 0.09250000864267349, "reward_std": 0.016329267993569374, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1595 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 299.0, "completions/mean_length": 643.1875, "completions/mean_terminated_length": 149.92857360839844, "completions/min_length": 62.0, "completions/min_terminated_length": 62.0, "epoch": 0.6190845616757176, "grad_norm": 12.177059173583984, "kl": 45.0361328125, "learning_rate": 2.0091600560198012e-06, "loss": 0.7317, "num_tokens": 311208093.0, "reward": 0.08874999731779099, "reward_std": 0.01759178563952446, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1596 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 258.0, "completions/mean_length": 248.41250610351562, "completions/mean_terminated_length": 149.75640869140625, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.6194724592707526, "grad_norm": 2.058959484100342, "kl": 14.25537109375, "learning_rate": 2.006490870114174e-06, "loss": 0.3298, "num_tokens": 311282694.0, "reward": 0.0950000062584877, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1597 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 282.0, "completions/mean_length": 207.375, "completions/mean_terminated_length": 158.1519012451172, "completions/min_length": 76.0, "completions/min_terminated_length": 76.0, "epoch": 0.6198603568657874, "grad_norm": 1.1608037948608398, "kl": 7.25927734375, "learning_rate": 2.0038228583109482e-06, "loss": 0.1596, "num_tokens": 311352428.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1598 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 662.0, "completions/mean_length": 272.1625061035156, "completions/mean_terminated_length": 174.11538696289062, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.6202482544608223, "grad_norm": 8.061070442199707, "kl": 16.38037109375, "learning_rate": 2.00115602482328e-06, "loss": 0.3415, "num_tokens": 311421521.0, "reward": 0.0949999988079071, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1599 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 385.0, "completions/mean_length": 322.8500061035156, "completions/mean_terminated_length": 175.8441619873047, "completions/min_length": 99.0, "completions/min_terminated_length": 99.0, "epoch": 0.6206361520558572, "grad_norm": 5.454685688018799, "kl": 19.1083984375, "learning_rate": 1.9984903738624635e-06, "loss": 0.4441, "num_tokens": 311491085.0, "reward": 0.0950000062584877, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1600 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 423.0, "completions/mean_length": 858.6875, "completions/mean_terminated_length": 171.98484802246094, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.6210240496508922, "grad_norm": 51.57035446166992, "kl": 60.4375, "learning_rate": 1.995825909637925e-06, "loss": 0.8398, "num_tokens": 311604476.0, "reward": 0.0950000062584877, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1601 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 269.0, "completions/mean_length": 772.9249877929688, "completions/mean_terminated_length": 186.5, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.6214119472459271, "grad_norm": 57.752742767333984, "kl": 57.1875, "learning_rate": 1.9931626363572177e-06, "loss": 0.6167, "num_tokens": 311722254.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1602 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 298.0, "completions/mean_length": 367.4875183105469, "completions/mean_terminated_length": 171.25, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.621799844840962, "grad_norm": 9.580121040344238, "kl": 19.8310546875, "learning_rate": 1.9905005582260125e-06, "loss": 0.375, "num_tokens": 311810781.0, "reward": 0.0962500050663948, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1603 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 337.0, "completions/mean_length": 321.0375061035156, "completions/mean_terminated_length": 173.96104431152344, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.622187742435997, "grad_norm": 2.86392879486084, "kl": 21.76416015625, "learning_rate": 1.9878396794480947e-06, "loss": 0.4713, "num_tokens": 311884464.0, "reward": 0.0950000062584877, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1604 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 302.0, "completions/mean_length": 420.7124938964844, "completions/mean_terminated_length": 175.69334411621094, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.6225756400310318, "grad_norm": 2.138658285140991, "kl": 31.1953125, "learning_rate": 1.9851800042253556e-06, "loss": 0.6008, "num_tokens": 311965129.0, "reward": 0.09375, "reward_std": 0.015235701575875282, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1605 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 312.0, "completions/mean_length": 228.1750030517578, "completions/mean_terminated_length": 179.21519470214844, "completions/min_length": 89.0, "completions/min_terminated_length": 89.0, "epoch": 0.6229635376260667, "grad_norm": 0.597082793712616, "kl": 7.66357421875, "learning_rate": 1.9825215367577847e-06, "loss": 0.1635, "num_tokens": 312026039.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1606 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 256.0, "completions/mean_length": 412.5500183105469, "completions/mean_terminated_length": 166.98666381835938, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.6233514352211016, "grad_norm": 5.599918365478516, "kl": 32.2314453125, "learning_rate": 1.979864281243465e-06, "loss": 0.6103, "num_tokens": 312112595.0, "reward": 0.0950000062584877, "reward_std": 0.011700168251991272, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1607 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 337.0, "completions/mean_length": 241.5124969482422, "completions/mean_terminated_length": 192.72152709960938, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.6237393328161366, "grad_norm": 1.1711432933807373, "kl": 6.396484375, "learning_rate": 1.9772082418785663e-06, "loss": 0.1261, "num_tokens": 312188364.0, "reward": 0.0962500050663948, "reward_std": 0.008164633996784687, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1608 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 564.0, "completions/mean_length": 257.2749938964844, "completions/mean_terminated_length": 208.6835479736328, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.6241272304111715, "grad_norm": 0.8955493569374084, "kl": 8.96435546875, "learning_rate": 1.974553422857337e-06, "loss": 0.1744, "num_tokens": 312255898.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1609 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 306.0, "completions/mean_length": 335.0249938964844, "completions/mean_terminated_length": 188.49349975585938, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.6245151280062063, "grad_norm": 5.765621185302734, "kl": 19.703125, "learning_rate": 1.971899828372098e-06, "loss": 0.4588, "num_tokens": 312336660.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1610 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 340.0, "completions/mean_length": 581.2999877929688, "completions/mean_terminated_length": 190.7777862548828, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.6249030256012412, "grad_norm": 6.36816930770874, "kl": 46.59375, "learning_rate": 1.969247462613239e-06, "loss": 0.8194, "num_tokens": 312432012.0, "reward": 0.08999999612569809, "reward_std": 0.020958369597792625, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1611 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 356.0, "completions/max_terminated_length": 356.0, "completions/mean_length": 201.2375030517578, "completions/mean_terminated_length": 201.2375030517578, "completions/min_length": 99.0, "completions/min_terminated_length": 99.0, "epoch": 0.6252909231962762, "grad_norm": 0.2937581539154053, "kl": 0.5703125, "learning_rate": 1.966596329769206e-06, "loss": 0.0057, "num_tokens": 312501383.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1612 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 370.0, "completions/mean_length": 349.07501220703125, "completions/mean_terminated_length": 203.09091186523438, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "epoch": 0.6256788207913111, "grad_norm": 2.49267578125, "kl": 22.32275390625, "learning_rate": 1.9639464340264995e-06, "loss": 0.4873, "num_tokens": 312575645.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1613 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 393.0, "completions/mean_length": 244.83750915527344, "completions/mean_terminated_length": 196.08860778808594, "completions/min_length": 58.0, "completions/min_terminated_length": 58.0, "epoch": 0.626066718386346, "grad_norm": 0.6440484523773193, "kl": 8.1083984375, "learning_rate": 1.961297779569668e-06, "loss": 0.1654, "num_tokens": 312640728.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1614 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 454.0, "completions/mean_length": 254.6875, "completions/mean_terminated_length": 206.06329345703125, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.626454615981381, "grad_norm": 0.6257917284965515, "kl": 8.17041015625, "learning_rate": 1.958650370581297e-06, "loss": 0.1727, "num_tokens": 312714167.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1615 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 317.0, "completions/mean_length": 250.8249969482422, "completions/mean_terminated_length": 202.1519012451172, "completions/min_length": 92.0, "completions/min_terminated_length": 92.0, "epoch": 0.6268425135764158, "grad_norm": 0.6713263988494873, "kl": 0.7109375, "learning_rate": 1.9560042112420068e-06, "loss": -0.0123, "num_tokens": 312781481.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1616 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 374.0, "completions/max_terminated_length": 374.0, "completions/mean_length": 221.4250030517578, "completions/mean_terminated_length": 221.4250030517578, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.6272304111714507, "grad_norm": 0.45247721672058105, "kl": 0.38818359375, "learning_rate": 1.953359305730445e-06, "loss": 0.0028, "num_tokens": 312848195.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1617 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 369.0, "completions/mean_length": 270.1625061035156, "completions/mean_terminated_length": 221.7341766357422, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.6276183087664856, "grad_norm": 1.299689531326294, "kl": 7.92822265625, "learning_rate": 1.950715658223278e-06, "loss": 0.164, "num_tokens": 312919528.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1618 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 460.0, "completions/max_terminated_length": 460.0, "completions/mean_length": 224.0749969482422, "completions/mean_terminated_length": 224.0749969482422, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.6280062063615206, "grad_norm": 0.14166377484798431, "kl": 0.3876953125, "learning_rate": 1.9480732728951864e-06, "loss": 0.0039, "num_tokens": 312984038.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1619 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 433.0, "completions/max_terminated_length": 433.0, "completions/mean_length": 222.21250915527344, "completions/mean_terminated_length": 222.21250915527344, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.6283941039565555, "grad_norm": 0.33772632479667664, "kl": 0.45068359375, "learning_rate": 1.9454321539188568e-06, "loss": 0.0045, "num_tokens": 313053735.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1620 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 426.0, "completions/mean_length": 278.1000061035156, "completions/mean_terminated_length": 229.77215576171875, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.6287820015515904, "grad_norm": 0.5357539057731628, "kl": 0.9267578125, "learning_rate": 1.942792305464976e-06, "loss": 0.0093, "num_tokens": 313127471.0, "reward": 0.12000000476837158, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.0603785514831543, "step": 1621 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3210.0, "completions/mean_length": 501.4375, "completions/mean_terminated_length": 312.25, "completions/min_length": 77.0, "completions/min_terminated_length": 77.0, "epoch": 0.6291698991466252, "grad_norm": 1.9791405200958252, "kl": 7.29345703125, "learning_rate": 1.9401537317022272e-06, "loss": 0.1449, "num_tokens": 313212682.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1622 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 499.0, "completions/max_terminated_length": 499.0, "completions/mean_length": 267.2250061035156, "completions/mean_terminated_length": 267.2250061035156, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.6295577967416602, "grad_norm": 1.1855552196502686, "kl": 0.63037109375, "learning_rate": 1.9375164367972753e-06, "loss": 0.0081, "num_tokens": 313298036.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1623 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 417.0, "completions/mean_length": 282.3000183105469, "completions/mean_terminated_length": 234.02532958984375, "completions/min_length": 98.0, "completions/min_terminated_length": 98.0, "epoch": 0.6299456943366951, "grad_norm": 0.9159151911735535, "kl": 0.51904296875, "learning_rate": 1.9348804249147723e-06, "loss": -0.0022, "num_tokens": 313368828.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1624 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 560.0, "completions/mean_length": 327.8500061035156, "completions/mean_terminated_length": 231.23077392578125, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.63033359193173, "grad_norm": 2.07940673828125, "kl": 17.5, "learning_rate": 1.9322457002173394e-06, "loss": 0.3518, "num_tokens": 313440832.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1625 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 390.0, "completions/mean_length": 328.38751220703125, "completions/mean_terminated_length": 231.7820587158203, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.630721489526765, "grad_norm": 1.9987000226974487, "kl": 16.9921875, "learning_rate": 1.9296122668655655e-06, "loss": 0.3029, "num_tokens": 313510615.0, "reward": 0.09624999761581421, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1626 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 434.0, "completions/mean_length": 333.63751220703125, "completions/mean_terminated_length": 237.1666717529297, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.6311093871217999, "grad_norm": 9.998502731323242, "kl": 14.0048828125, "learning_rate": 1.926980129018002e-06, "loss": 0.2243, "num_tokens": 313585826.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1627 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 551.0, "completions/mean_length": 527.8875122070312, "completions/mean_terminated_length": 238.58108520507812, "completions/min_length": 90.0, "completions/min_terminated_length": 90.0, "epoch": 0.6314972847168348, "grad_norm": 7.2893571853637695, "kl": 39.28125, "learning_rate": 1.9243492908311547e-06, "loss": 0.6992, "num_tokens": 313661161.0, "reward": 0.08875000476837158, "reward_std": 0.024493902921676636, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1628 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 404.0, "completions/mean_length": 529.4249877929688, "completions/mean_terminated_length": 240.24325561523438, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.6318851823118696, "grad_norm": 6.687590599060059, "kl": 36.41796875, "learning_rate": 1.9217197564594746e-06, "loss": 0.642, "num_tokens": 313755443.0, "reward": 0.0937500074505806, "reward_std": 0.015235701575875282, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1629 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 476.0, "completions/mean_length": 336.0375061035156, "completions/mean_terminated_length": 239.62820434570312, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.6322730799069046, "grad_norm": 2.301506280899048, "kl": 14.7021484375, "learning_rate": 1.9190915300553574e-06, "loss": 0.3212, "num_tokens": 313821662.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1630 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 456.0, "completions/mean_length": 319.5, "completions/mean_terminated_length": 271.6961975097656, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.6326609775019395, "grad_norm": 0.8904873728752136, "kl": 6.0087890625, "learning_rate": 1.9164646157691295e-06, "loss": 0.1036, "num_tokens": 313912710.0, "reward": 0.09750000387430191, "reward_std": 0.004629100672900677, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1631 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2770.0, "completions/mean_length": 343.1750183105469, "completions/mean_terminated_length": 295.6708984375, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.6330488750969744, "grad_norm": 10.255084991455078, "kl": 8.28076171875, "learning_rate": 1.9138390177490484e-06, "loss": 0.2438, "num_tokens": 313990180.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1632 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 589.0, "completions/mean_length": 407.8500061035156, "completions/mean_terminated_length": 313.28204345703125, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.6334367726920093, "grad_norm": 2.395143508911133, "kl": 5.26708984375, "learning_rate": 1.91121474014129e-06, "loss": 0.0966, "num_tokens": 314066536.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1633 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 607.0, "completions/max_terminated_length": 607.0, "completions/mean_length": 338.2250061035156, "completions/mean_terminated_length": 338.2250061035156, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.6338246702870443, "grad_norm": 0.10045450180768967, "kl": 0.32568359375, "learning_rate": 1.9085917870899496e-06, "loss": 0.0033, "num_tokens": 314150714.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1634 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 699.0, "completions/max_terminated_length": 699.0, "completions/mean_length": 352.1125183105469, "completions/mean_terminated_length": 352.1125183105469, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.6342125678820791, "grad_norm": 0.0960577055811882, "kl": 0.30078125, "learning_rate": 1.9059701627370265e-06, "loss": 0.003, "num_tokens": 314231267.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1635 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 726.0, "completions/mean_length": 403.8125, "completions/mean_terminated_length": 357.0759582519531, "completions/min_length": 135.0, "completions/min_terminated_length": 135.0, "epoch": 0.634600465477114, "grad_norm": 1.8518617153167725, "kl": 4.6259765625, "learning_rate": 1.903349871222422e-06, "loss": 0.1128, "num_tokens": 314320108.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1636 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 617.0, "completions/mean_length": 475.7124938964844, "completions/mean_terminated_length": 334.6623229980469, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.6349883630721489, "grad_norm": 4.726003170013428, "kl": 15.14453125, "learning_rate": 1.9007309166839365e-06, "loss": 0.3598, "num_tokens": 314410189.0, "reward": 0.0962500050663948, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1637 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 1455.0, "completions/mean_length": 539.6375122070312, "completions/mean_terminated_length": 352.4605407714844, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.6353762606671839, "grad_norm": 3.8100504875183105, "kl": 16.9365234375, "learning_rate": 1.898113303257255e-06, "loss": 0.3987, "num_tokens": 314500144.0, "reward": 0.0962500050663948, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1638 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 909.0, "completions/mean_length": 430.51251220703125, "completions/mean_terminated_length": 384.1139221191406, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.6357641582622188, "grad_norm": 0.7212842106819153, "kl": 6.37744140625, "learning_rate": 1.8954970350759478e-06, "loss": 0.1312, "num_tokens": 314584009.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1639 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 642.0, "completions/mean_length": 451.1125183105469, "completions/mean_terminated_length": 404.9747009277344, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.6361520558572537, "grad_norm": 0.5095143914222717, "kl": 6.6630859375, "learning_rate": 1.8928821162714589e-06, "loss": 0.1359, "num_tokens": 314665434.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1640 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 993.0, "completions/max_terminated_length": 993.0, "completions/mean_length": 381.7749938964844, "completions/mean_terminated_length": 381.7749938964844, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.6365399534522886, "grad_norm": 0.19407372176647186, "kl": 0.42578125, "learning_rate": 1.890268550973101e-06, "loss": 0.0043, "num_tokens": 314737808.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1641 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 1208.0, "completions/mean_length": 545.125, "completions/mean_terminated_length": 454.0769348144531, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.6369278510473235, "grad_norm": 1.8790109157562256, "kl": 13.39453125, "learning_rate": 1.8876563433080524e-06, "loss": 0.2499, "num_tokens": 314831330.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1642 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1086.0, "completions/mean_length": 462.2124938964844, "completions/mean_terminated_length": 416.2152099609375, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.6373157486423584, "grad_norm": 1.0188664197921753, "kl": 7.1123046875, "learning_rate": 1.8850454974013432e-06, "loss": 0.1356, "num_tokens": 314930171.0, "reward": 0.09749999642372131, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1643 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 753.0, "completions/mean_length": 452.1125183105469, "completions/mean_terminated_length": 405.98736572265625, "completions/min_length": 193.0, "completions/min_terminated_length": 193.0, "epoch": 0.6377036462373933, "grad_norm": 1.452317476272583, "kl": 7.93505859375, "learning_rate": 1.8824360173758582e-06, "loss": 0.153, "num_tokens": 315016308.0, "reward": 0.09874999523162842, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1644 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 884.0, "completions/mean_length": 491.1000061035156, "completions/mean_terminated_length": 398.66668701171875, "completions/min_length": 135.0, "completions/min_terminated_length": 135.0, "epoch": 0.6380915438324283, "grad_norm": 2.5287842750549316, "kl": 16.09130859375, "learning_rate": 1.8798279073523207e-06, "loss": 0.3005, "num_tokens": 315099308.0, "reward": 0.09750000387430191, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1645 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3919.0, "completions/max_terminated_length": 3919.0, "completions/mean_length": 549.4750366210938, "completions/mean_terminated_length": 549.4750366210938, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.6384794414274632, "grad_norm": 1.6993200778961182, "kl": 2.6103515625, "learning_rate": 1.8772211714492921e-06, "loss": 0.0263, "num_tokens": 315191546.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1646 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1842.0, "completions/max_terminated_length": 1842.0, "completions/mean_length": 464.5874938964844, "completions/mean_terminated_length": 464.5874938964844, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.638867339022498, "grad_norm": 0.06913130730390549, "kl": 0.3212890625, "learning_rate": 1.8746158137831655e-06, "loss": 0.0032, "num_tokens": 315276505.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1647 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 977.0, "completions/max_terminated_length": 977.0, "completions/mean_length": 450.6000061035156, "completions/mean_terminated_length": 450.6000061035156, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.6392552366175329, "grad_norm": 0.06755506247282028, "kl": 0.35205078125, "learning_rate": 1.872011838468155e-06, "loss": 0.0035, "num_tokens": 315353457.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1648 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1926.0, "completions/max_terminated_length": 1926.0, "completions/mean_length": 471.4250183105469, "completions/mean_terminated_length": 471.4250183105469, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.6396431342125679, "grad_norm": 0.3065289258956909, "kl": 0.37255859375, "learning_rate": 1.8694092496162946e-06, "loss": 0.0037, "num_tokens": 315429627.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1649 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1712.0, "completions/mean_length": 489.2124938964844, "completions/mean_terminated_length": 443.5569763183594, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.6400310318076028, "grad_norm": 0.06823644042015076, "kl": 0.33447265625, "learning_rate": 1.866808051337427e-06, "loss": 0.0033, "num_tokens": 315518460.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1650 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1322.0, "completions/mean_length": 471.9750061035156, "completions/mean_terminated_length": 426.1012878417969, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.6404189294026377, "grad_norm": 0.07871992141008377, "kl": 0.36181640625, "learning_rate": 1.864208247739197e-06, "loss": 0.0036, "num_tokens": 315605266.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1651 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2670.0, "completions/mean_length": 620.8624877929688, "completions/mean_terminated_length": 531.7564086914062, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.6408068269976727, "grad_norm": 1.0181828737258911, "kl": 7.38720703125, "learning_rate": 1.861609842927053e-06, "loss": 0.1375, "num_tokens": 315697023.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1652 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3303.0, "completions/max_terminated_length": 3303.0, "completions/mean_length": 523.4249877929688, "completions/mean_terminated_length": 523.4249877929688, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.6411947245927075, "grad_norm": 0.06863528490066528, "kl": 0.33154296875, "learning_rate": 1.859012841004228e-06, "loss": 0.0033, "num_tokens": 315792985.0, "reward": 0.10000000149011612, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1653 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3556.0, "completions/mean_length": 869.0375366210938, "completions/mean_terminated_length": 743.3117065429688, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "epoch": 0.6415826221877424, "grad_norm": 1.2163811922073364, "kl": 3.5634765625, "learning_rate": 1.856417246071744e-06, "loss": 0.0683, "num_tokens": 315911628.0, "reward": 0.09749999642372131, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1654 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3307.0, "completions/mean_length": 861.1875, "completions/mean_terminated_length": 645.5333251953125, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.6419705197827773, "grad_norm": 2.5044960975646973, "kl": 18.12890625, "learning_rate": 1.8538230622283992e-06, "loss": 0.3719, "num_tokens": 316030323.0, "reward": 0.11625000089406967, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.0603785514831543, "step": 1655 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2908.0, "completions/mean_length": 652.9375, "completions/mean_terminated_length": 609.3544311523438, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "epoch": 0.6423584173778123, "grad_norm": 0.8570151925086975, "kl": 6.38720703125, "learning_rate": 1.8512302935707639e-06, "loss": 0.1388, "num_tokens": 316138110.0, "reward": 0.09875001013278961, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1656 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3909.0, "completions/mean_length": 884.4750366210938, "completions/mean_terminated_length": 670.3733520507812, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.6427463149728472, "grad_norm": 3.049962043762207, "kl": 24.657958984375, "learning_rate": 1.8486389441931741e-06, "loss": 0.4403, "num_tokens": 316256508.0, "reward": 0.1537500023841858, "reward_std": 0.012793734669685364, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.18113566935062408, "step": 1657 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2671.0, "completions/mean_length": 838.8125, "completions/mean_terminated_length": 621.6666870117188, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.6431342125678821, "grad_norm": 2.595970392227173, "kl": 25.5859375, "learning_rate": 1.8460490181877234e-06, "loss": 0.5291, "num_tokens": 316371981.0, "reward": 0.0925000011920929, "reward_std": 0.02121320180594921, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1658 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2890.0, "completions/mean_length": 615.7875366210938, "completions/mean_terminated_length": 526.55126953125, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.6435221101629169, "grad_norm": 0.4700644314289093, "kl": 6.119140625, "learning_rate": 1.8434605196442594e-06, "loss": 0.1134, "num_tokens": 316466716.0, "reward": 0.09875000268220901, "reward_std": 0.0035355337895452976, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339753627777, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1659 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 2834.0, "completions/mean_length": 752.625, "completions/mean_terminated_length": 576.6578979492188, "completions/min_length": 245.0, "completions/min_terminated_length": 245.0, "epoch": 0.6439100077579519, "grad_norm": 1.2080081701278687, "kl": 15.673828125, "learning_rate": 1.8408734526503746e-06, "loss": 0.2912, "num_tokens": 316572638.0, "reward": 0.0950000062584877, "reward_std": 0.011700168251991272, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1660 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3303.0, "completions/mean_length": 1013.9000244140625, "completions/mean_terminated_length": 718.3561401367188, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.6442979053529868, "grad_norm": 1.2687854766845703, "kl": 20.1611328125, "learning_rate": 1.8382878212913997e-06, "loss": 0.3918, "num_tokens": 316709510.0, "reward": 0.0950000062584877, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1661 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3399.0, "completions/mean_length": 779.3250122070312, "completions/mean_terminated_length": 650.1038818359375, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.6446858029480217, "grad_norm": 1.4350708723068237, "kl": 17.662109375, "learning_rate": 1.835703629650401e-06, "loss": 0.3516, "num_tokens": 316820624.0, "reward": 0.0962500050663948, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1662 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3391.0, "completions/mean_length": 1004.7875366210938, "completions/mean_terminated_length": 708.369873046875, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "epoch": 0.6450737005430567, "grad_norm": 1.5780149698257446, "kl": 12.740234375, "learning_rate": 1.833120881808168e-06, "loss": 0.2794, "num_tokens": 316948415.0, "reward": 0.09750000387430191, "reward_std": 0.018771236762404442, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1663 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1885.0, "completions/mean_length": 859.5875244140625, "completions/mean_terminated_length": 499.9861145019531, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.6454615981380916, "grad_norm": 1.5530447959899902, "kl": 7.875, "learning_rate": 1.8305395818432154e-06, "loss": 0.1241, "num_tokens": 317074678.0, "reward": 0.0950000062584877, "reward_std": 0.011700167320668697, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1664 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1720.0, "completions/mean_length": 637.0375366210938, "completions/mean_terminated_length": 406.44000244140625, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.6458494957331264, "grad_norm": 1.5415289402008057, "kl": 18.6015625, "learning_rate": 1.827959733831765e-06, "loss": 0.3664, "num_tokens": 317166513.0, "reward": 0.10125000774860382, "reward_std": 0.02474873699247837, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472136124968529, "step": 1665 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2847.0, "completions/mean_length": 516.3875122070312, "completions/mean_terminated_length": 471.0759582519531, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.6462373933281613, "grad_norm": 0.5172843337059021, "kl": 0.44970703125, "learning_rate": 1.8253813418477494e-06, "loss": 0.0036, "num_tokens": 317241440.0, "reward": 0.10250001400709152, "reward_std": 0.007071068976074457, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360682487487793, "step": 1666 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3554.0, "completions/mean_length": 1283.9749755859375, "completions/mean_terminated_length": 580.96875, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.6466252909231963, "grad_norm": 13.314061164855957, "kl": 59.9892578125, "learning_rate": 1.8228044099628028e-06, "loss": 0.9676, "num_tokens": 317398174.0, "reward": 0.0962500050663948, "reward_std": 0.0655711218714714, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 1667 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 842.0, "completions/mean_length": 976.8624877929688, "completions/mean_terminated_length": 426.4264831542969, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.6470131885182312, "grad_norm": 5.808863162994385, "kl": 35.1220703125, "learning_rate": 1.820228942246251e-06, "loss": 0.5799, "num_tokens": 317528563.0, "reward": 0.11749999970197678, "reward_std": 0.03447909280657768, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.06930943578481674, "step": 1668 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 2543.0, "completions/mean_length": 1464.9749755859375, "completions/mean_terminated_length": 467.0, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.6474010861132661, "grad_norm": 11.015401840209961, "kl": 66.75, "learning_rate": 1.8176549427651107e-06, "loss": 1.1496, "num_tokens": 317697673.0, "reward": 0.08125000447034836, "reward_std": 0.033921051770448685, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1669 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2213.0, "completions/mean_length": 1617.7625732421875, "completions/mean_terminated_length": 555.6607666015625, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "epoch": 0.647788983708301, "grad_norm": 6.356179714202881, "kl": 60.0625, "learning_rate": 1.815082415584078e-06, "loss": 1.1032, "num_tokens": 317880198.0, "reward": 0.08125000447034836, "reward_std": 0.03783402964472771, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1670 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 994.0, "completions/mean_length": 1315.0999755859375, "completions/mean_terminated_length": 448.9179992675781, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.648176881303336, "grad_norm": 8.382823944091797, "kl": 48.46875, "learning_rate": 1.8125113647655223e-06, "loss": 0.8314, "num_tokens": 318035126.0, "reward": 0.09750000387430191, "reward_std": 0.06168710067868233, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.061828672885894775, "step": 1671 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2263.0, "completions/mean_length": 1363.7249755859375, "completions/mean_terminated_length": 512.6884765625, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.6485647788983708, "grad_norm": 5.0330071449279785, "kl": 50.25, "learning_rate": 1.8099417943694864e-06, "loss": 1.0193, "num_tokens": 318193352.0, "reward": 0.08125000447034836, "reward_std": 0.03593923896551132, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1672 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3848.0, "completions/mean_length": 1033.4124755859375, "completions/mean_terminated_length": 545.1739501953125, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "epoch": 0.6489526764934057, "grad_norm": 11.123847961425781, "kl": 20.59375, "learning_rate": 1.8073737084536719e-06, "loss": 0.5021, "num_tokens": 318332561.0, "reward": 0.09000000357627869, "reward_std": 0.023400336503982544, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1673 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1776.0, "completions/mean_length": 1248.925048828125, "completions/mean_terminated_length": 537.15625, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.6493405740884407, "grad_norm": 5.892149448394775, "kl": 14.9375, "learning_rate": 1.8048071110734362e-06, "loss": 0.3231, "num_tokens": 318493667.0, "reward": 0.10000000149011612, "reward_std": 0.037542473524808884, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.04974619299173355, "step": 1674 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2966.0, "completions/mean_length": 1695.862548828125, "completions/mean_terminated_length": 604.890869140625, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.6497284716834756, "grad_norm": 11.696320533752441, "kl": 24.609375, "learning_rate": 1.802242006281788e-06, "loss": 0.5501, "num_tokens": 318667256.0, "reward": 0.09875001013278961, "reward_std": 0.07087641209363937, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.11373966187238693, "step": 1675 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3627.0, "completions/mean_length": 1642.2750244140625, "completions/mean_terminated_length": 590.6785888671875, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.6501163692785105, "grad_norm": 4.440528869628906, "kl": 28.96875, "learning_rate": 1.7996783981293763e-06, "loss": 0.5785, "num_tokens": 318848846.0, "reward": 0.10375000536441803, "reward_std": 0.0683412179350853, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.017500000074505806, "rewards/unicoder_reward_fn/std": 0.11558875441551208, "step": 1676 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1062.0, "completions/mean_length": 1444.550048828125, "completions/mean_terminated_length": 500.8135681152344, "completions/min_length": 211.0, "completions/min_terminated_length": 211.0, "epoch": 0.6505042668735453, "grad_norm": 5.298960208892822, "kl": 8.112548828125, "learning_rate": 1.7971162906644911e-06, "loss": 0.187, "num_tokens": 319013242.0, "reward": 0.12000000476837158, "reward_std": 0.05605887994170189, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.02250000089406967, "rewards/unicoder_reward_fn/std": 0.10551262646913528, "step": 1677 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2827.0, "completions/mean_length": 1612.1500244140625, "completions/mean_terminated_length": 547.6428833007812, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.6508921644685803, "grad_norm": 5.850111961364746, "kl": 37.1875, "learning_rate": 1.7945556879330473e-06, "loss": 0.7148, "num_tokens": 319184646.0, "reward": 0.10250001400709152, "reward_std": 0.06889895349740982, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932424664497375, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.08535110205411911, "step": 1678 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2965.0, "completions/mean_length": 1983.2000732421875, "completions/mean_terminated_length": 646.5305786132812, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.6512800620636152, "grad_norm": 10.754670143127441, "kl": 37.875, "learning_rate": 1.7919965939785867e-06, "loss": 0.5338, "num_tokens": 319390934.0, "reward": 0.11750000715255737, "reward_std": 0.088487409055233, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.0325000025331974, "rewards/unicoder_reward_fn/std": 0.10765039920806885, "step": 1679 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2754.0, "completions/mean_length": 2004.7750244140625, "completions/mean_terminated_length": 681.7550659179688, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "epoch": 0.6516679596586501, "grad_norm": 9.604680061340332, "kl": 16.017578125, "learning_rate": 1.789439012842269e-06, "loss": 0.3297, "num_tokens": 319601012.0, "reward": 0.10125000774860382, "reward_std": 0.03956083208322525, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.054191410541534424, "step": 1680 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 1851.0, "completions/mean_length": 1842.625, "completions/mean_terminated_length": 629.2692260742188, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.652055857253685, "grad_norm": 2.1412322521209717, "kl": 20.796875, "learning_rate": 1.7868829485628614e-06, "loss": 0.3585, "num_tokens": 319791534.0, "reward": 0.11625000089406967, "reward_std": 0.08041234314441681, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.10732654482126236, "step": 1681 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3512.0, "completions/mean_length": 2073.087646484375, "completions/mean_terminated_length": 577.8912963867188, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "epoch": 0.65244375484872, "grad_norm": 3.052741527557373, "kl": 28.65625, "learning_rate": 1.7843284051767417e-06, "loss": 0.476, "num_tokens": 320011429.0, "reward": 0.0925000011920929, "reward_std": 0.03206084296107292, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.03142199665307999, "step": 1682 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3244.0, "completions/mean_length": 1705.7750244140625, "completions/mean_terminated_length": 681.3928833007812, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.6528316524437549, "grad_norm": 1.0835375785827637, "kl": 8.648681640625, "learning_rate": 1.7817753867178804e-06, "loss": 0.103, "num_tokens": 320187355.0, "reward": 0.11124999821186066, "reward_std": 0.04679989069700241, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.017500000074505806, "rewards/unicoder_reward_fn/std": 0.06516774743795395, "step": 1683 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 3581.0, "completions/mean_length": 1849.4124755859375, "completions/mean_terminated_length": 639.7115478515625, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.6532195500387897, "grad_norm": 5.014141082763672, "kl": 19.1875, "learning_rate": 1.779223897217842e-06, "loss": 0.1857, "num_tokens": 320382204.0, "reward": 0.14374999701976776, "reward_std": 0.1376245617866516, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.16146089136600494, "step": 1684 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 3020.0, "completions/mean_length": 1555.5, "completions/mean_terminated_length": 591.862060546875, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.6536074476338247, "grad_norm": 2.557659864425659, "kl": 9.27099609375, "learning_rate": 1.7766739407057781e-06, "loss": 0.1852, "num_tokens": 320553540.0, "reward": 0.10875000804662704, "reward_std": 0.03509972617030144, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.04871773719787598, "step": 1685 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 3114.0, "completions/mean_length": 1988.8499755859375, "completions/mean_terminated_length": 584.0833740234375, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "epoch": 0.6539953452288596, "grad_norm": 4.080078125, "kl": 24.21875, "learning_rate": 1.774125521208417e-06, "loss": 0.4081, "num_tokens": 320761440.0, "reward": 0.12250000238418579, "reward_std": 0.06990926712751389, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.0325000025331974, "rewards/unicoder_reward_fn/std": 0.08682151883840561, "step": 1686 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1669.0, "completions/mean_length": 1523.925048828125, "completions/mean_terminated_length": 608.440673828125, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.6543832428238945, "grad_norm": 1.9096888303756714, "kl": 8.98486328125, "learning_rate": 1.7715786427500614e-06, "loss": 0.1969, "num_tokens": 320928546.0, "reward": 0.15625, "reward_std": 0.10362248867750168, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.16958625614643097, "step": 1687 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4031.0, "completions/mean_length": 1416.9749755859375, "completions/mean_terminated_length": 798.7384643554688, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.6547711404189294, "grad_norm": 2.5447936058044434, "kl": 25.28125, "learning_rate": 1.7690333093525802e-06, "loss": 0.5763, "num_tokens": 321098184.0, "reward": 0.1274999976158142, "reward_std": 0.08780962973833084, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.12231936305761337, "step": 1688 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 2350.0, "completions/mean_length": 1537.9500732421875, "completions/mean_terminated_length": 685.2667236328125, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.6551590380139644, "grad_norm": 0.7601783275604248, "kl": 3.37109375, "learning_rate": 1.766489525035402e-06, "loss": 0.0896, "num_tokens": 321272124.0, "reward": 0.1587500125169754, "reward_std": 0.09591598808765411, "rewards/format_reward/mean": 0.987500011920929, "rewards/format_reward/std": 0.11180339008569717, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.14372971653938293, "step": 1689 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3189.0, "completions/mean_length": 1723.3125, "completions/mean_terminated_length": 706.4464721679688, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "epoch": 0.6555469356089992, "grad_norm": 5.0019731521606445, "kl": 28.375, "learning_rate": 1.7639472938155077e-06, "loss": 0.4869, "num_tokens": 321457405.0, "reward": 0.14250001311302185, "reward_std": 0.11357436329126358, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.05250000208616257, "rewards/unicoder_reward_fn/std": 0.126265749335289, "step": 1690 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3703.0, "completions/mean_length": 1383.0374755859375, "completions/mean_terminated_length": 704.796875, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.6559348332040341, "grad_norm": 0.7496294975280762, "kl": 6.91162109375, "learning_rate": 1.7614066197074309e-06, "loss": 0.103, "num_tokens": 321619144.0, "reward": 0.18375001847743988, "reward_std": 0.14430049061775208, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.19051894545555115, "step": 1691 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3968.0, "completions/mean_length": 1179.6875, "completions/mean_terminated_length": 714.7681274414062, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.656322730799069, "grad_norm": 8.839340209960938, "kl": 21.69384765625, "learning_rate": 1.7588675067232402e-06, "loss": 0.3881, "num_tokens": 321764895.0, "reward": 0.1550000011920929, "reward_std": 0.10698225349187851, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.14086082577705383, "step": 1692 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2470.0, "completions/mean_length": 1468.1624755859375, "completions/mean_terminated_length": 705.241943359375, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.656710628394104, "grad_norm": 1.9078792333602905, "kl": 15.296875, "learning_rate": 1.7563299588725425e-06, "loss": 0.2752, "num_tokens": 321930260.0, "reward": 0.16250000894069672, "reward_std": 0.1375834196805954, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.17166349291801453, "step": 1693 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2342.0, "completions/mean_length": 1258.2249755859375, "completions/mean_terminated_length": 656.2727661132812, "completions/min_length": 231.0, "completions/min_terminated_length": 231.0, "epoch": 0.6570985259891389, "grad_norm": 4.022554397583008, "kl": 26.15625, "learning_rate": 1.753793980162472e-06, "loss": 0.5044, "num_tokens": 322082870.0, "reward": 0.11750000715255737, "reward_std": 0.057416882365942, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.0885537788271904, "step": 1694 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1744.0, "completions/mean_length": 1106.1875, "completions/mean_terminated_length": 629.5507202148438, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.6574864235841738, "grad_norm": 3.601515293121338, "kl": 18.671875, "learning_rate": 1.7512595745976857e-06, "loss": 0.3533, "num_tokens": 322220181.0, "reward": 0.15000000596046448, "reward_std": 0.09587617963552475, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.14300128817558289, "step": 1695 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3825.0, "completions/mean_length": 1049.5999755859375, "completions/mean_terminated_length": 711.1111450195312, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "epoch": 0.6578743211792087, "grad_norm": 1.2949178218841553, "kl": 14.78125, "learning_rate": 1.7487267461803548e-06, "loss": 0.2705, "num_tokens": 322352349.0, "reward": 0.14124999940395355, "reward_std": 0.08815698325634003, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.15342091023921967, "step": 1696 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1776.0, "completions/mean_length": 1239.0750732421875, "completions/mean_terminated_length": 579.7846069335938, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.6582622187742436, "grad_norm": 1.8340611457824707, "kl": 17.984375, "learning_rate": 1.7461954989101626e-06, "loss": 0.3615, "num_tokens": 322500435.0, "reward": 0.16375000774860382, "reward_std": 0.13967184722423553, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.07249999791383743, "rewards/unicoder_reward_fn/std": 0.1889176368713379, "step": 1697 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3324.0, "completions/mean_length": 1686.7000732421875, "completions/mean_terminated_length": 654.1428833007812, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "epoch": 0.6586501163692785, "grad_norm": 5.283705711364746, "kl": 12.671875, "learning_rate": 1.7436658367842935e-06, "loss": 0.2545, "num_tokens": 322683795.0, "reward": 0.16500000655651093, "reward_std": 0.1293555498123169, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.07249999791383743, "rewards/unicoder_reward_fn/std": 0.15008437633514404, "step": 1698 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.65, "completions/max_length": 4096.0, "completions/max_terminated_length": 2722.0, "completions/mean_length": 2858.16259765625, "completions/mean_terminated_length": 559.3214721679688, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "epoch": 0.6590380139643134, "grad_norm": 32.01725769042969, "kl": 15.5625, "learning_rate": 1.741137763797428e-06, "loss": 0.2242, "num_tokens": 322954288.0, "reward": 0.10750000923871994, "reward_std": 0.05984846502542496, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.017500000074505806, "rewards/unicoder_reward_fn/std": 0.07919692248106003, "step": 1699 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2406.0, "completions/mean_length": 3018.862548828125, "completions/mean_terminated_length": 649.1599731445312, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.6594259115593484, "grad_norm": 61.9713134765625, "kl": 23.5625, "learning_rate": 1.7386112839417403e-06, "loss": 0.3374, "num_tokens": 323244725.0, "reward": 0.10125000774860382, "reward_std": 0.06354853510856628, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.07647288590669632, "step": 1700 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2418.0, "completions/mean_length": 2809.324951171875, "completions/mean_terminated_length": 546.5516967773438, "completions/min_length": 242.0, "completions/min_terminated_length": 242.0, "epoch": 0.6598138091543833, "grad_norm": 50.54338836669922, "kl": 13.65625, "learning_rate": 1.7360864012068851e-06, "loss": 0.1205, "num_tokens": 323514215.0, "reward": 0.13625000417232513, "reward_std": 0.09684091061353683, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.04500000178813934, "rewards/unicoder_reward_fn/std": 0.1621844619512558, "step": 1701 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3608.0, "completions/mean_length": 2144.91259765625, "completions/mean_terminated_length": 548.5681762695312, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.6602017067494181, "grad_norm": 4.776397228240967, "kl": 17.1171875, "learning_rate": 1.7335631195799992e-06, "loss": 0.3953, "num_tokens": 323731624.0, "reward": 0.20875000953674316, "reward_std": 0.16192813217639923, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.2517784833908081, "step": 1702 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3626.0, "completions/mean_length": 2806.212646484375, "completions/mean_terminated_length": 656.5667114257812, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.660589604344453, "grad_norm": 2.352006435394287, "kl": 15.515625, "learning_rate": 1.7310414430456874e-06, "loss": 0.2487, "num_tokens": 323999505.0, "reward": 0.19375000894069672, "reward_std": 0.19650961458683014, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.21333695948123932, "step": 1703 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 1964.0, "completions/mean_length": 2254.712646484375, "completions/mean_terminated_length": 588.7857055664062, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.660977501939488, "grad_norm": 4.626423358917236, "kl": 25.0625, "learning_rate": 1.72852137558602e-06, "loss": 0.4252, "num_tokens": 324222650.0, "reward": 0.20500002801418304, "reward_std": 0.18184909224510193, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.24772383272647858, "step": 1704 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 861.0, "completions/mean_length": 3054.112548828125, "completions/mean_terminated_length": 472.0434875488281, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "epoch": 0.6613653995345229, "grad_norm": 6.9042510986328125, "kl": 25.515625, "learning_rate": 1.7260029211805298e-06, "loss": 0.4559, "num_tokens": 324513771.0, "reward": 0.10375000536441803, "reward_std": 0.07283319532871246, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.11638001352548599, "step": 1705 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2717.0, "completions/mean_length": 2223.487548828125, "completions/mean_terminated_length": 612.2557983398438, "completions/min_length": 144.0, "completions/min_terminated_length": 144.0, "epoch": 0.6617532971295578, "grad_norm": 3.335026502609253, "kl": 10.372314453125, "learning_rate": 1.7234860838061978e-06, "loss": 0.2194, "num_tokens": 324745978.0, "reward": 0.14500001072883606, "reward_std": 0.0965721532702446, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.05250000208616257, "rewards/unicoder_reward_fn/std": 0.13404609262943268, "step": 1706 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 3962.0, "completions/mean_length": 2430.0, "completions/mean_terminated_length": 588.631591796875, "completions/min_length": 144.0, "completions/min_terminated_length": 144.0, "epoch": 0.6621411947245927, "grad_norm": 1.8523914813995361, "kl": 25.5, "learning_rate": 1.7209708674374565e-06, "loss": 0.4611, "num_tokens": 324986442.0, "reward": 0.12125001102685928, "reward_std": 0.0890558511018753, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.11946611851453781, "step": 1707 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 919.0, "completions/mean_length": 2382.875, "completions/mean_terminated_length": 489.4210510253906, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.6625290923196276, "grad_norm": 5.570544719696045, "kl": 37.46875, "learning_rate": 1.7184572760461734e-06, "loss": 0.566, "num_tokens": 325227272.0, "reward": 0.12124999612569809, "reward_std": 0.10863528400659561, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.1207571029663086, "step": 1708 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1397.0, "completions/mean_length": 2296.699951171875, "completions/mean_terminated_length": 585.1707153320312, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.6629169899146625, "grad_norm": 9.813475608825684, "kl": 39.8125, "learning_rate": 1.7159453136016527e-06, "loss": 0.6729, "num_tokens": 325457856.0, "reward": 0.17250001430511475, "reward_std": 0.18226709961891174, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.0924999937415123, "rewards/unicoder_reward_fn/std": 0.2011203020811081, "step": 1709 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1216.0, "completions/mean_length": 2158.025146484375, "completions/mean_terminated_length": 490.4651184082031, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.6633048875096974, "grad_norm": 9.599997520446777, "kl": 31.34375, "learning_rate": 1.7134349840706271e-06, "loss": 0.5635, "num_tokens": 325682002.0, "reward": 0.1287500113248825, "reward_std": 0.1071070209145546, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.04249999672174454, "rewards/unicoder_reward_fn/std": 0.11337737739086151, "step": 1710 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2324.0, "completions/mean_length": 2198.75, "completions/mean_terminated_length": 566.2325439453125, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "epoch": 0.6636927851047324, "grad_norm": 17.011707305908203, "kl": 55.8125, "learning_rate": 1.7109262914172486e-06, "loss": 0.9569, "num_tokens": 325904534.0, "reward": 0.12124999612569809, "reward_std": 0.11673889309167862, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.04500000178813934, "rewards/unicoder_reward_fn/std": 0.13861887156963348, "step": 1711 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 1036.0, "completions/mean_length": 2457.525146484375, "completions/mean_terminated_length": 454.9444580078125, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.6640806826997673, "grad_norm": 2.493668556213379, "kl": 35.375, "learning_rate": 1.7084192396030862e-06, "loss": 0.5592, "num_tokens": 326148536.0, "reward": 0.10375000536441803, "reward_std": 0.090838223695755, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.12065224349498749, "step": 1712 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1017.0, "completions/mean_length": 1788.675048828125, "completions/mean_terminated_length": 476.66668701171875, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.6644685802948022, "grad_norm": 4.869513988494873, "kl": 22.109375, "learning_rate": 1.705913832587117e-06, "loss": 0.4769, "num_tokens": 326344806.0, "reward": 0.13250000774860382, "reward_std": 0.10468011349439621, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.11462652683258057, "step": 1713 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1080.0, "completions/mean_length": 2006.1624755859375, "completions/mean_terminated_length": 461.5, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.664856477889837, "grad_norm": 7.657306671142578, "kl": 23.625, "learning_rate": 1.7034100743257187e-06, "loss": 0.4686, "num_tokens": 326543347.0, "reward": 0.15625, "reward_std": 0.1333610564470291, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.16257424652576447, "step": 1714 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 764.0, "completions/mean_length": 2416.237548828125, "completions/mean_terminated_length": 464.0810852050781, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.665244375484872, "grad_norm": 6.615090370178223, "kl": 26.375, "learning_rate": 1.7009079687726702e-06, "loss": 0.4555, "num_tokens": 326780190.0, "reward": 0.11125000566244125, "reward_std": 0.09088089317083359, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.030000001192092896, "rewards/unicoder_reward_fn/std": 0.12365734577178955, "step": 1715 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 1542.0, "completions/mean_length": 2219.712646484375, "completions/mean_terminated_length": 522.1190795898438, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.6656322730799069, "grad_norm": 5.935461044311523, "kl": 31.484375, "learning_rate": 1.698407519879135e-06, "loss": 0.5226, "num_tokens": 327011519.0, "reward": 0.09125000238418579, "reward_std": 0.06999819725751877, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.07356216758489609, "step": 1716 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1409.0, "completions/mean_length": 2051.675048828125, "completions/mean_terminated_length": 540.6521606445312, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "epoch": 0.6660201706749418, "grad_norm": 5.506263256072998, "kl": 49.8125, "learning_rate": 1.695908731593665e-06, "loss": 0.8953, "num_tokens": 327228261.0, "reward": 0.125, "reward_std": 0.1285284012556076, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.19025298953056335, "step": 1717 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 1207.0, "completions/mean_length": 1950.6375732421875, "completions/mean_terminated_length": 520.3958740234375, "completions/min_length": 201.0, "completions/min_terminated_length": 201.0, "epoch": 0.6664080682699767, "grad_norm": 3.7970995903015137, "kl": 40.375, "learning_rate": 1.693411607862188e-06, "loss": 0.7281, "num_tokens": 327429152.0, "reward": 0.10375001281499863, "reward_std": 0.06835062801837921, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.12065224349498749, "step": 1718 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1073.0, "completions/mean_length": 1994.3250732421875, "completions/mean_terminated_length": 518.6808471679688, "completions/min_length": 224.0, "completions/min_terminated_length": 224.0, "epoch": 0.6667959658650117, "grad_norm": 15.810676574707031, "kl": 37.75, "learning_rate": 1.6909161526280018e-06, "loss": 0.7201, "num_tokens": 327637154.0, "reward": 0.1275000125169754, "reward_std": 0.11633341759443283, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.1395970731973648, "step": 1719 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1045.0, "completions/mean_length": 2015.3375244140625, "completions/mean_terminated_length": 554.44677734375, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "epoch": 0.6671838634600465, "grad_norm": 6.134160995483398, "kl": 35.28125, "learning_rate": 1.6884223698317723e-06, "loss": 0.6687, "num_tokens": 327846893.0, "reward": 0.08624999970197678, "reward_std": 0.04490509629249573, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.04472135752439499, "step": 1720 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1291.0, "completions/mean_length": 2086.324951171875, "completions/mean_terminated_length": 442.04547119140625, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.6675717610550814, "grad_norm": 5.271361351013184, "kl": 53.1875, "learning_rate": 1.6859302634115219e-06, "loss": 0.887, "num_tokens": 328061927.0, "reward": 0.09750000387430191, "reward_std": 0.08447324484586716, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.07626569271087646, "step": 1721 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1106.0, "completions/mean_length": 2002.4500732421875, "completions/mean_terminated_length": 532.5106201171875, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.6679596586501164, "grad_norm": 6.216116428375244, "kl": 42.1875, "learning_rate": 1.6834398373026275e-06, "loss": 0.708, "num_tokens": 328261787.0, "reward": 0.1600000113248825, "reward_std": 0.1317330300807953, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.15813390910625458, "step": 1722 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1170.0, "completions/mean_length": 1975.0875244140625, "completions/mean_terminated_length": 485.9361572265625, "completions/min_length": 211.0, "completions/min_terminated_length": 211.0, "epoch": 0.6683475562451513, "grad_norm": 8.498207092285156, "kl": 37.5, "learning_rate": 1.6809510954378119e-06, "loss": 0.6066, "num_tokens": 328460506.0, "reward": 0.13625000417232513, "reward_std": 0.13824878633022308, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.1635058969259262, "step": 1723 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 2042.0, "completions/mean_length": 2233.625, "completions/mean_terminated_length": 548.6190795898438, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.6687354538401862, "grad_norm": 9.7827787399292, "kl": 52.75, "learning_rate": 1.678464041747137e-06, "loss": 0.8417, "num_tokens": 328682228.0, "reward": 0.11500000953674316, "reward_std": 0.111603282392025, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.12528449296951294, "step": 1724 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 964.0, "completions/mean_length": 2372.237548828125, "completions/mean_terminated_length": 467.02630615234375, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.6691233514352211, "grad_norm": 4.124914646148682, "kl": 33.453125, "learning_rate": 1.675978680158002e-06, "loss": 0.6342, "num_tokens": 328919255.0, "reward": 0.13250000774860382, "reward_std": 0.1422296166419983, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.15061058104038239, "step": 1725 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1177.0, "completions/mean_length": 2910.237548828125, "completions/mean_terminated_length": 582.629638671875, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.669511249030256, "grad_norm": 5.518553733825684, "kl": 42.25, "learning_rate": 1.6734950145951299e-06, "loss": 0.6997, "num_tokens": 329198522.0, "reward": 0.12000000476837158, "reward_std": 0.14674688875675201, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.18264926970005035, "step": 1726 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 1238.0, "completions/mean_length": 2513.60009765625, "completions/mean_terminated_length": 579.5555419921875, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "epoch": 0.6698991466252909, "grad_norm": 21.450321197509766, "kl": 29.53125, "learning_rate": 1.6710130489805693e-06, "loss": 0.6122, "num_tokens": 329439842.0, "reward": 0.07500000298023224, "reward_std": 0.07146799564361572, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.04871773719787598, "step": 1727 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1238.0, "completions/mean_length": 2493.22509765625, "completions/mean_terminated_length": 432.5142822265625, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.6702870442203258, "grad_norm": 20.4842586517334, "kl": 18.9375, "learning_rate": 1.6685327872336826e-06, "loss": 0.4788, "num_tokens": 329688180.0, "reward": 0.09375, "reward_std": 0.08359101414680481, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545217394828796, "rewards/unicoder_reward_fn/mean": 0.022499999031424522, "rewards/unicoder_reward_fn/std": 0.0899718627333641, "step": 1728 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 892.0, "completions/mean_length": 2172.900146484375, "completions/mean_terminated_length": 432.952392578125, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "epoch": 0.6706749418153607, "grad_norm": 16.896717071533203, "kl": 20.90625, "learning_rate": 1.6660542332711405e-06, "loss": 0.5641, "num_tokens": 329908132.0, "reward": 0.15000000596046448, "reward_std": 0.1539614498615265, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.08000000566244125, "rewards/unicoder_reward_fn/std": 0.19770841300487518, "step": 1729 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1431.0, "completions/mean_length": 2510.637451171875, "completions/mean_terminated_length": 472.3143005371094, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.6710628394103957, "grad_norm": 19.594682693481445, "kl": 24.0625, "learning_rate": 1.663577391006919e-06, "loss": 0.6445, "num_tokens": 330157287.0, "reward": 0.0962500050663948, "reward_std": 0.11401630938053131, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.030000001192092896, "rewards/unicoder_reward_fn/std": 0.11517735570669174, "step": 1730 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 817.0, "completions/mean_length": 2630.91259765625, "completions/mean_terminated_length": 433.28125, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.6714507370054306, "grad_norm": 18.819793701171875, "kl": 28.28125, "learning_rate": 1.6611022643522895e-06, "loss": 0.6413, "num_tokens": 330414944.0, "reward": 0.125, "reward_std": 0.17611566185951233, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.0650000050663948, "rewards/unicoder_reward_fn/std": 0.18765711784362793, "step": 1731 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1011.0, "completions/mean_length": 2603.60009765625, "completions/mean_terminated_length": 478.06060791015625, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.6718386346004654, "grad_norm": 8.443586349487305, "kl": 36.09375, "learning_rate": 1.658628857215816e-06, "loss": 0.7724, "num_tokens": 330662232.0, "reward": 0.11000000685453415, "reward_std": 0.14064009487628937, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.16761544346809387, "step": 1732 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1113.0, "completions/mean_length": 2876.074951171875, "completions/mean_terminated_length": 481.40740966796875, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.6722265321955004, "grad_norm": 2.826646566390991, "kl": 41.4375, "learning_rate": 1.6561571735033452e-06, "loss": 0.6578, "num_tokens": 330931054.0, "reward": 0.0937500074505806, "reward_std": 0.11460476368665695, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.04000000283122063, "rewards/unicoder_reward_fn/std": 0.14372970163822174, "step": 1733 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 1283.0, "completions/mean_length": 2395.050048828125, "completions/mean_terminated_length": 515.0526123046875, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.6726144297905353, "grad_norm": 3.7246007919311523, "kl": 49.4375, "learning_rate": 1.653687217118002e-06, "loss": 0.8015, "num_tokens": 331167506.0, "reward": 0.14500001072883606, "reward_std": 0.1739877462387085, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.17603223025798798, "step": 1734 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1176.0, "completions/mean_length": 1998.3250732421875, "completions/mean_terminated_length": 447.86956787109375, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.6730023273855702, "grad_norm": 4.796194076538086, "kl": 44.3125, "learning_rate": 1.6512189919601865e-06, "loss": 0.8234, "num_tokens": 331381372.0, "reward": 0.16750000417232513, "reward_std": 0.1489756554365158, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.1827358901500702, "step": 1735 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2130.0, "completions/mean_length": 2123.699951171875, "completions/mean_terminated_length": 589.6889038085938, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.6733902249806051, "grad_norm": 5.403931140899658, "kl": 30.6875, "learning_rate": 1.6487525019275624e-06, "loss": 0.6345, "num_tokens": 331600332.0, "reward": 0.14124999940395355, "reward_std": 0.12772496044635773, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.144410640001297, "step": 1736 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2094.0, "completions/mean_length": 1925.050048828125, "completions/mean_terminated_length": 622.47998046875, "completions/min_length": 245.0, "completions/min_terminated_length": 245.0, "epoch": 0.6737781225756401, "grad_norm": 5.765691757202148, "kl": 30.71875, "learning_rate": 1.646287750915054e-06, "loss": 0.3996, "num_tokens": 331794880.0, "reward": 0.16250000894069672, "reward_std": 0.11204905807971954, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932424664497375, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.13309842348098755, "step": 1737 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2001.0, "completions/mean_length": 2044.9625244140625, "completions/mean_terminated_length": 604.872314453125, "completions/min_length": 239.0, "completions/min_terminated_length": 239.0, "epoch": 0.674166020170675, "grad_norm": 4.103687763214111, "kl": 36.40625, "learning_rate": 1.6438247428148391e-06, "loss": 0.6763, "num_tokens": 332007021.0, "reward": 0.1262500137090683, "reward_std": 0.13221004605293274, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.14408154785633087, "step": 1738 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3704.0, "completions/mean_length": 2208.0625, "completions/mean_terminated_length": 583.5581665039062, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.6745539177657098, "grad_norm": 4.316061973571777, "kl": 20.796875, "learning_rate": 1.6413634815163444e-06, "loss": 0.5095, "num_tokens": 332230106.0, "reward": 0.1875000149011612, "reward_std": 0.16128328442573547, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.10249999910593033, "rewards/unicoder_reward_fn/std": 0.2110702097415924, "step": 1739 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 1080.0, "completions/mean_length": 1777.9375, "completions/mean_terminated_length": 529.75, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.6749418153607447, "grad_norm": 3.317728281021118, "kl": 32.1875, "learning_rate": 1.6389039709062377e-06, "loss": 0.6263, "num_tokens": 332418261.0, "reward": 0.1287500113248825, "reward_std": 0.11640427261590958, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.12117568403482437, "step": 1740 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3264.0, "completions/mean_length": 1741.2750244140625, "completions/mean_terminated_length": 670.9454345703125, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.6753297129557797, "grad_norm": 3.031064033508301, "kl": 15.83642578125, "learning_rate": 1.6364462148684208e-06, "loss": 0.3574, "num_tokens": 332614243.0, "reward": 0.1612500250339508, "reward_std": 0.11506819725036621, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.1660982072353363, "step": 1741 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2825.0, "completions/mean_length": 1513.4000244140625, "completions/mean_terminated_length": 594.1694946289062, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.6757176105508146, "grad_norm": 15.072419166564941, "kl": 17.1796875, "learning_rate": 1.6339902172840268e-06, "loss": 0.4281, "num_tokens": 332780083.0, "reward": 0.14249999821186066, "reward_std": 0.08617151528596878, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.09409797936677933, "step": 1742 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1288.0, "completions/mean_length": 1373.3499755859375, "completions/mean_terminated_length": 525.3114624023438, "completions/min_length": 90.0, "completions/min_terminated_length": 90.0, "epoch": 0.6761055081458495, "grad_norm": 6.691806793212891, "kl": 17.39404296875, "learning_rate": 1.6315359820314116e-06, "loss": 0.4132, "num_tokens": 332935135.0, "reward": 0.16625000536441803, "reward_std": 0.1205512285232544, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.16067500412464142, "step": 1743 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 1569.0, "completions/mean_length": 1446.8499755859375, "completions/mean_terminated_length": 563.800048828125, "completions/min_length": 245.0, "completions/min_terminated_length": 245.0, "epoch": 0.6764934057408845, "grad_norm": 2.0998480319976807, "kl": 24.265625, "learning_rate": 1.629083512986146e-06, "loss": 0.5559, "num_tokens": 333100235.0, "reward": 0.21625001728534698, "reward_std": 0.14102615416049957, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.1274999976158142, "rewards/unicoder_reward_fn/std": 0.17787423729896545, "step": 1744 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2038.0, "completions/mean_length": 1211.6375732421875, "completions/mean_terminated_length": 599.8030395507812, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.6768813033359193, "grad_norm": 1.419338345527649, "kl": 9.64990234375, "learning_rate": 1.6266328140210151e-06, "loss": 0.1481, "num_tokens": 333244334.0, "reward": 0.2537500560283661, "reward_std": 0.17465655505657196, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.2287970930337906, "step": 1745 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2255.0, "completions/mean_length": 1467.0875244140625, "completions/mean_terminated_length": 648.245849609375, "completions/min_length": 202.0, "completions/min_terminated_length": 202.0, "epoch": 0.6772692009309542, "grad_norm": 19.9470272064209, "kl": 35.625, "learning_rate": 1.624183889006007e-06, "loss": 0.6258, "num_tokens": 333405629.0, "reward": 0.26125001907348633, "reward_std": 0.21714237332344055, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.23029977083206177, "step": 1746 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 1461.0, "completions/mean_length": 1573.5999755859375, "completions/mean_terminated_length": 616.8275756835938, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.6776570985259891, "grad_norm": 14.988546371459961, "kl": 33.3828125, "learning_rate": 1.6217367418083067e-06, "loss": 0.6337, "num_tokens": 333586125.0, "reward": 0.1875000149011612, "reward_std": 0.141906276345253, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.16226248443126678, "step": 1747 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 698.0, "completions/mean_length": 1918.7000732421875, "completions/mean_terminated_length": 389.9574279785156, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.6780449961210241, "grad_norm": 22.02706527709961, "kl": 43.875, "learning_rate": 1.6192913762922958e-06, "loss": 0.7457, "num_tokens": 333785717.0, "reward": 0.18250000476837158, "reward_std": 0.16629302501678467, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.2134111076593399, "step": 1748 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.775, "completions/max_length": 4096.0, "completions/max_terminated_length": 990.0, "completions/mean_length": 3277.637451171875, "completions/mean_terminated_length": 458.8333435058594, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.678432893716059, "grad_norm": 149.04644775390625, "kl": 70.1875, "learning_rate": 1.6168477963195384e-06, "loss": 0.7941, "num_tokens": 334090832.0, "reward": 0.09875001013278961, "reward_std": 0.08568398654460907, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.022499999031424522, "rewards/unicoder_reward_fn/std": 0.0899718627333641, "step": 1749 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8125, "completions/max_length": 4096.0, "completions/max_terminated_length": 785.0, "completions/mean_length": 3430.3125, "completions/mean_terminated_length": 545.6666870117188, "completions/min_length": 401.0, "completions/min_terminated_length": 401.0, "epoch": 0.6788207913110939, "grad_norm": 114.8969955444336, "kl": 59.875, "learning_rate": 1.6144060057487827e-06, "loss": 0.7021, "num_tokens": 334412257.0, "reward": 0.14125001430511475, "reward_std": 0.10458069294691086, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.1706094741821289, "step": 1750 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 986.0, "completions/mean_length": 2878.387451171875, "completions/mean_terminated_length": 488.25927734375, "completions/min_length": 118.0, "completions/min_terminated_length": 118.0, "epoch": 0.6792086889061287, "grad_norm": 67.52017974853516, "kl": 56.75, "learning_rate": 1.6119660084359485e-06, "loss": 0.7698, "num_tokens": 334692176.0, "reward": 0.13625000417232513, "reward_std": 0.1261049509048462, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.14222463965415955, "step": 1751 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 1089.0, "completions/mean_length": 1927.0999755859375, "completions/mean_terminated_length": 481.16668701171875, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.6795965865011637, "grad_norm": 26.263986587524414, "kl": 51.8125, "learning_rate": 1.6095278082341234e-06, "loss": 0.7561, "num_tokens": 334886016.0, "reward": 0.20000000298023224, "reward_std": 0.1809391975402832, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932424664497375, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.20320221781730652, "step": 1752 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1725.0, "completions/mean_length": 1595.2000732421875, "completions/mean_terminated_length": 586.1052856445312, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.6799844840961986, "grad_norm": 5.915400981903076, "kl": 30.515625, "learning_rate": 1.6070914089935608e-06, "loss": 0.5826, "num_tokens": 335060216.0, "reward": 0.257500022649765, "reward_std": 0.18854060769081116, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.26897603273391724, "step": 1753 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2064.0, "completions/mean_length": 1483.3375244140625, "completions/mean_terminated_length": 669.5573120117188, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "epoch": 0.6803723816912335, "grad_norm": 3.750664472579956, "kl": 28.625, "learning_rate": 1.6046568145616647e-06, "loss": 0.5214, "num_tokens": 335233283.0, "reward": 0.1862500011920929, "reward_std": 0.14966242015361786, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.182805135846138, "step": 1754 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 2207.0, "completions/mean_length": 1586.8250732421875, "completions/mean_terminated_length": 635.0689697265625, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "epoch": 0.6807602792862685, "grad_norm": 8.075926780700684, "kl": 18.234375, "learning_rate": 1.602224028782995e-06, "loss": 0.4174, "num_tokens": 335407725.0, "reward": 0.23250003159046173, "reward_std": 0.18122820556163788, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.14000001549720764, "rewards/unicoder_reward_fn/std": 0.2707701027393341, "step": 1755 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1673.0, "completions/mean_length": 1437.3875732421875, "completions/mean_terminated_length": 665.5322265625, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.6811481768813034, "grad_norm": 15.375327110290527, "kl": 27.40625, "learning_rate": 1.5997930554992508e-06, "loss": 0.4545, "num_tokens": 335570204.0, "reward": 0.23499999940395355, "reward_std": 0.19612815976142883, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.2709570527076721, "step": 1756 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1529.0, "completions/mean_length": 1348.050048828125, "completions/mean_terminated_length": 661.0625, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.6815360744763382, "grad_norm": 4.230234622955322, "kl": 19.0439453125, "learning_rate": 1.59736389854927e-06, "loss": 0.3581, "num_tokens": 335735448.0, "reward": 0.23749999701976776, "reward_std": 0.187339648604393, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.20861205458641052, "step": 1757 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1234.0, "completions/mean_length": 1548.75, "completions/mean_terminated_length": 642.1016845703125, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.6819239720713731, "grad_norm": 8.693156242370605, "kl": 13.65625, "learning_rate": 1.5949365617690254e-06, "loss": 0.3179, "num_tokens": 335902252.0, "reward": 0.23500001430511475, "reward_std": 0.17954835295677185, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.27038416266441345, "step": 1758 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1302.0, "completions/mean_length": 1660.237548828125, "completions/mean_terminated_length": 616.3392944335938, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.6823118696664081, "grad_norm": 13.247208595275879, "kl": 17.0859375, "learning_rate": 1.5925110489916107e-06, "loss": 0.3464, "num_tokens": 336087551.0, "reward": 0.2237500250339508, "reward_std": 0.15027673542499542, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.22034209966659546, "step": 1759 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 1035.0, "completions/mean_length": 2240.78759765625, "completions/mean_terminated_length": 562.2619018554688, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.682699767261443, "grad_norm": 8.67684555053711, "kl": 19.75, "learning_rate": 1.590087364047243e-06, "loss": 0.4717, "num_tokens": 336319886.0, "reward": 0.17875002324581146, "reward_std": 0.1259888857603073, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.23350657522678375, "step": 1760 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1148.0, "completions/mean_length": 3199.525146484375, "completions/mean_terminated_length": 680.857177734375, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.6830876648564779, "grad_norm": 32.292232513427734, "kl": 30.21875, "learning_rate": 1.5876655107632522e-06, "loss": 0.3755, "num_tokens": 336628984.0, "reward": 0.0950000062584877, "reward_std": 0.03405144438147545, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.03823643922805786, "step": 1761 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.85, "completions/max_length": 4096.0, "completions/max_terminated_length": 1076.0, "completions/mean_length": 3564.85009765625, "completions/mean_terminated_length": 555.0, "completions/min_length": 329.0, "completions/min_terminated_length": 329.0, "epoch": 0.6834755624515128, "grad_norm": 19.0035400390625, "kl": 31.28125, "learning_rate": 1.5852454929640731e-06, "loss": 0.4714, "num_tokens": 336963076.0, "reward": 0.16124999523162842, "reward_std": 0.16202037036418915, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.19382885098457336, "step": 1762 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2640.0, "completions/mean_length": 3609.650146484375, "completions/mean_terminated_length": 558.9091186523438, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.6838634600465477, "grad_norm": 34.07112121582031, "kl": 38.125, "learning_rate": 1.582827314471246e-06, "loss": 0.4541, "num_tokens": 337302032.0, "reward": 0.10375000536441803, "reward_std": 0.0685933455824852, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.09332127869129181, "step": 1763 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8, "completions/max_length": 4096.0, "completions/max_terminated_length": 913.0, "completions/mean_length": 3373.587646484375, "completions/mean_terminated_length": 483.9375, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.6842513576415826, "grad_norm": 22.16769790649414, "kl": 31.3125, "learning_rate": 1.5804109791034037e-06, "loss": 0.4401, "num_tokens": 337619599.0, "reward": 0.14624999463558197, "reward_std": 0.14912518858909607, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.20609387755393982, "step": 1764 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.825, "completions/max_length": 4096.0, "completions/max_terminated_length": 931.0, "completions/mean_length": 3469.462646484375, "completions/mean_terminated_length": 515.7857666015625, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.6846392552366175, "grad_norm": 23.348804473876953, "kl": 26.5, "learning_rate": 1.5779964906762706e-06, "loss": 0.3273, "num_tokens": 337945708.0, "reward": 0.12125001102685928, "reward_std": 0.12100972980260849, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.15342091023921967, "step": 1765 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 1137.0, "completions/mean_length": 3042.762451171875, "completions/mean_terminated_length": 585.2083740234375, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.6850271528316525, "grad_norm": 19.652963638305664, "kl": 16.8125, "learning_rate": 1.5755838530026529e-06, "loss": 0.1674, "num_tokens": 338239449.0, "reward": 0.14500001072883606, "reward_std": 0.1796748787164688, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.2164003551006317, "step": 1766 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1435.0, "completions/mean_length": 2393.900146484375, "completions/mean_terminated_length": 604.5128173828125, "completions/min_length": 312.0, "completions/min_terminated_length": 312.0, "epoch": 0.6854150504266874, "grad_norm": 6.668430328369141, "kl": 15.4375, "learning_rate": 1.5731730698924344e-06, "loss": 0.2482, "num_tokens": 338480857.0, "reward": 0.11500000953674316, "reward_std": 0.13137395679950714, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.177464559674263, "step": 1767 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 2034.0, "completions/mean_length": 1845.737548828125, "completions/mean_terminated_length": 634.0577392578125, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.6858029480217223, "grad_norm": 6.765340805053711, "kl": 10.8837890625, "learning_rate": 1.5707641451525717e-06, "loss": 0.1938, "num_tokens": 338677684.0, "reward": 0.13750000298023224, "reward_std": 0.138458713889122, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.1671806424856186, "step": 1768 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1513.0, "completions/mean_length": 1769.8125, "completions/mean_terminated_length": 584.7736206054688, "completions/min_length": 311.0, "completions/min_terminated_length": 311.0, "epoch": 0.6861908456167571, "grad_norm": 10.040214538574219, "kl": 15.390625, "learning_rate": 1.5683570825870847e-06, "loss": 0.3616, "num_tokens": 338861325.0, "reward": 0.21375003457069397, "reward_std": 0.19788581132888794, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.24801746010780334, "step": 1769 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2321.0, "completions/mean_length": 1661.2125244140625, "completions/mean_terminated_length": 617.732177734375, "completions/min_length": 261.0, "completions/min_terminated_length": 261.0, "epoch": 0.6865787432117921, "grad_norm": 6.022435188293457, "kl": 6.14990234375, "learning_rate": 1.5659518859970545e-06, "loss": 0.2208, "num_tokens": 339052710.0, "reward": 0.2224999964237213, "reward_std": 0.2222566157579422, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.26470667123794556, "step": 1770 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 1206.0, "completions/mean_length": 1419.5374755859375, "completions/mean_terminated_length": 527.3833618164062, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.686966640806827, "grad_norm": 9.083842277526855, "kl": 4.6689453125, "learning_rate": 1.563548559180615e-06, "loss": 0.1533, "num_tokens": 339202177.0, "reward": 0.24125002324581146, "reward_std": 0.1658533811569214, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.24596747756004333, "step": 1771 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1406.0, "completions/mean_length": 997.5750122070312, "completions/mean_terminated_length": 503.6231994628906, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.6873545384018619, "grad_norm": 6.705094337463379, "kl": 5.56640625, "learning_rate": 1.561147105932946e-06, "loss": 0.1798, "num_tokens": 339327719.0, "reward": 0.20374999940395355, "reward_std": 0.1453009843826294, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.1770181953907013, "step": 1772 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 985.0, "completions/mean_length": 1112.2750244140625, "completions/mean_terminated_length": 479.3636474609375, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.6877424359968968, "grad_norm": 10.724763870239258, "kl": 2.1484375, "learning_rate": 1.5587475300462718e-06, "loss": 0.0651, "num_tokens": 339460341.0, "reward": 0.20999999344348907, "reward_std": 0.16075745224952698, "rewards/format_reward/mean": 0.9750000238418579, "rewards/format_reward/std": 0.15710997581481934, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.2482215315103531, "step": 1773 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1468.0, "completions/mean_length": 1597.2625732421875, "completions/mean_terminated_length": 461.47271728515625, "completions/min_length": 210.0, "completions/min_terminated_length": 210.0, "epoch": 0.6881303335919318, "grad_norm": 11.774740219116211, "kl": 11.703125, "learning_rate": 1.5563498353098481e-06, "loss": 0.3112, "num_tokens": 339651498.0, "reward": 0.17750000953674316, "reward_std": 0.09819843620061874, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.14151082932949066, "step": 1774 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1074.0, "completions/mean_length": 791.6875, "completions/mean_terminated_length": 424.5416564941406, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.6885182311869666, "grad_norm": 1.7096166610717773, "kl": 9.66552734375, "learning_rate": 1.5539540255099645e-06, "loss": 0.1731, "num_tokens": 339760617.0, "reward": 0.22625000774860382, "reward_std": 0.19172362983226776, "rewards/format_reward/mean": 0.9624999761581421, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.22969438135623932, "step": 1775 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1147.0, "completions/mean_length": 1261.4000244140625, "completions/mean_terminated_length": 438.45159912109375, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.6889061287820015, "grad_norm": 4.414111137390137, "kl": 17.960205078125, "learning_rate": 1.5515601044299306e-06, "loss": 0.4661, "num_tokens": 339918025.0, "reward": 0.19500000774860382, "reward_std": 0.15785515308380127, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.10249999910593033, "rewards/unicoder_reward_fn/std": 0.2181481420993805, "step": 1776 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 902.0, "completions/mean_length": 1320.737548828125, "completions/mean_terminated_length": 456.31146240234375, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.6892940263770365, "grad_norm": 2.559438705444336, "kl": 39.625, "learning_rate": 1.5491680758500732e-06, "loss": 0.7901, "num_tokens": 340073724.0, "reward": 0.18250000476837158, "reward_std": 0.13776832818984985, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.2249613106250763, "step": 1777 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 821.0, "completions/mean_length": 854.2374877929688, "completions/mean_terminated_length": 391.1285705566406, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.6896819239720714, "grad_norm": 2.883023977279663, "kl": 31.0, "learning_rate": 1.5467779435477342e-06, "loss": 0.6345, "num_tokens": 340187543.0, "reward": 0.29749998450279236, "reward_std": 0.20018373429775238, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.20499999821186066, "rewards/unicoder_reward_fn/std": 0.26428794860839844, "step": 1778 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 999.0, "completions/mean_length": 1526.925048828125, "completions/mean_terminated_length": 425.89288330078125, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.6900698215671063, "grad_norm": 9.656723022460938, "kl": 47.28125, "learning_rate": 1.5443897112972571e-06, "loss": 0.8338, "num_tokens": 340349289.0, "reward": 0.2587500214576721, "reward_std": 0.2103203386068344, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.17500001192092896, "rewards/unicoder_reward_fn/std": 0.23029977083206177, "step": 1779 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 730.0, "completions/mean_length": 1560.800048828125, "completions/mean_terminated_length": 408.43634033203125, "completions/min_length": 123.0, "completions/min_terminated_length": 123.0, "epoch": 0.6904577191621412, "grad_norm": 8.951403617858887, "kl": 53.625, "learning_rate": 1.5420033828699882e-06, "loss": 0.8704, "num_tokens": 340526297.0, "reward": 0.1912500113248825, "reward_std": 0.14021410048007965, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.10999999195337296, "rewards/unicoder_reward_fn/std": 0.19066838920116425, "step": 1780 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3920.0, "completions/mean_length": 1780.0, "completions/mean_terminated_length": 463.058837890625, "completions/min_length": 159.0, "completions/min_terminated_length": 159.0, "epoch": 0.6908456167571762, "grad_norm": 21.305683135986328, "kl": 63.5, "learning_rate": 1.539618962034265e-06, "loss": 1.0515, "num_tokens": 340713393.0, "reward": 0.19375000894069672, "reward_std": 0.18781499564647675, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.21034933626651764, "step": 1781 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 540.0, "completions/mean_length": 1678.4625244140625, "completions/mean_terminated_length": 376.7115478515625, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.691233514352211, "grad_norm": 10.008017539978027, "kl": 42.75, "learning_rate": 1.5372364525554124e-06, "loss": 0.7739, "num_tokens": 340895870.0, "reward": 0.16875000298023224, "reward_std": 0.15455424785614014, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.08499999344348907, "rewards/unicoder_reward_fn/std": 0.17938010394573212, "step": 1782 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 786.0, "completions/mean_length": 2150.462646484375, "completions/mean_terminated_length": 390.21429443359375, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.6916214119472459, "grad_norm": 33.10399627685547, "kl": 41.15625, "learning_rate": 1.5348558581957397e-06, "loss": 0.7279, "num_tokens": 341118675.0, "reward": 0.11250000447034836, "reward_std": 0.08647697418928146, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.09426598250865936, "step": 1783 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2022.0, "completions/mean_length": 2100.324951171875, "completions/mean_terminated_length": 383.11627197265625, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.6920093095422808, "grad_norm": 20.85453224182129, "kl": 39.59375, "learning_rate": 1.5324771827145296e-06, "loss": 0.6961, "num_tokens": 341339029.0, "reward": 0.16375000774860382, "reward_std": 0.14685426652431488, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.18765711784362793, "step": 1784 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 622.0, "completions/mean_length": 1556.1624755859375, "completions/mean_terminated_length": 401.6908874511719, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.6923972071373158, "grad_norm": 22.309593200683594, "kl": 20.25, "learning_rate": 1.530100429868035e-06, "loss": 0.4418, "num_tokens": 341506554.0, "reward": 0.24000000953674316, "reward_std": 0.177363783121109, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.19420400261878967, "step": 1785 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 759.0, "completions/mean_length": 1958.9000244140625, "completions/mean_terminated_length": 379.3043518066406, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.6927851047323507, "grad_norm": 6.134904384613037, "kl": 40.3125, "learning_rate": 1.527725603409474e-06, "loss": 0.8254, "num_tokens": 341711402.0, "reward": 0.24625001847743988, "reward_std": 0.2318355292081833, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.2969890832901001, "step": 1786 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3302.0, "completions/mean_length": 2817.362548828125, "completions/mean_terminated_length": 568.72412109375, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.6931730023273855, "grad_norm": 13.798765182495117, "kl": 26.78125, "learning_rate": 1.5253527070890218e-06, "loss": 0.5529, "num_tokens": 341993095.0, "reward": 0.15000000596046448, "reward_std": 0.16487465798854828, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.21872761845588684, "step": 1787 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 2311.0, "completions/mean_length": 2642.837646484375, "completions/mean_terminated_length": 463.09375, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.6935608999224204, "grad_norm": 10.179110527038574, "kl": 36.484375, "learning_rate": 1.5229817446538043e-06, "loss": 0.6542, "num_tokens": 342245762.0, "reward": 0.1575000137090683, "reward_std": 0.1810508668422699, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.23780377209186554, "step": 1788 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 813.0, "completions/mean_length": 2659.900146484375, "completions/mean_terminated_length": 389.93548583984375, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.6939487975174554, "grad_norm": 9.920428276062012, "kl": 30.09375, "learning_rate": 1.5206127198478948e-06, "loss": 0.6311, "num_tokens": 342500002.0, "reward": 0.13500000536441803, "reward_std": 0.13135436177253723, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.21250465512275696, "step": 1789 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 611.0, "completions/mean_length": 3128.375, "completions/mean_terminated_length": 409.8095397949219, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.6943366951124903, "grad_norm": 9.088579177856445, "kl": 33.21875, "learning_rate": 1.5182456364123086e-06, "loss": 0.5459, "num_tokens": 342806528.0, "reward": 0.09250000864267349, "reward_std": 0.12265463173389435, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.03500000014901161, "rewards/unicoder_reward_fn/std": 0.1378863900899887, "step": 1790 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 3080.0, "completions/mean_length": 2915.637451171875, "completions/mean_terminated_length": 464.1153869628906, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.6947245927075252, "grad_norm": 7.343377590179443, "kl": 39.3125, "learning_rate": 1.5158804980849925e-06, "loss": 0.7045, "num_tokens": 343082723.0, "reward": 0.10750000923871994, "reward_std": 0.12866799533367157, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.04749999940395355, "rewards/unicoder_reward_fn/std": 0.12825508415699005, "step": 1791 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 959.0, "completions/mean_length": 3122.512451171875, "completions/mean_terminated_length": 387.4761962890625, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.6951124903025602, "grad_norm": 9.569267272949219, "kl": 27.4765625, "learning_rate": 1.5135173086008217e-06, "loss": 0.4179, "num_tokens": 343381676.0, "reward": 0.13375000655651093, "reward_std": 0.15261296927928925, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.1976923793554306, "step": 1792 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 724.0, "completions/mean_length": 3083.525146484375, "completions/mean_terminated_length": 414.2727355957031, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.695500387897595, "grad_norm": 6.755748748779297, "kl": 35.59375, "learning_rate": 1.5111560716915973e-06, "loss": 0.6125, "num_tokens": 343677662.0, "reward": 0.10875000804662704, "reward_std": 0.1512499302625656, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.15829390287399292, "step": 1793 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 627.0, "completions/mean_length": 2851.012451171875, "completions/mean_terminated_length": 407.1481628417969, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.6958882854926299, "grad_norm": 8.752991676330566, "kl": 27.734375, "learning_rate": 1.5087967910860325e-06, "loss": 0.5487, "num_tokens": 343949015.0, "reward": 0.17750000953674316, "reward_std": 0.21398495137691498, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.24833622574806213, "step": 1794 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 691.0, "completions/mean_length": 2537.400146484375, "completions/mean_terminated_length": 428.70587158203125, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.6962761830876648, "grad_norm": 14.624520301818848, "kl": 20.59375, "learning_rate": 1.5064394705097545e-06, "loss": 0.5678, "num_tokens": 344201639.0, "reward": 0.15625, "reward_std": 0.14106445014476776, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.17603223025798798, "step": 1795 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1337.0, "completions/mean_length": 2505.60009765625, "completions/mean_terminated_length": 460.79998779296875, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.6966640806826998, "grad_norm": 10.837366104125977, "kl": 24.5625, "learning_rate": 1.5040841136852945e-06, "loss": 0.4717, "num_tokens": 344440991.0, "reward": 0.1587500125169754, "reward_std": 0.15365783870220184, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.08999999612569809, "rewards/unicoder_reward_fn/std": 0.20842993259429932, "step": 1796 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 885.0, "completions/mean_length": 2230.512451171875, "completions/mean_terminated_length": 456.0243835449219, "completions/min_length": 193.0, "completions/min_terminated_length": 193.0, "epoch": 0.6970519782777347, "grad_norm": 12.168606758117676, "kl": 34.3125, "learning_rate": 1.5017307243320812e-06, "loss": 0.8206, "num_tokens": 344672136.0, "reward": 0.1837500035762787, "reward_std": 0.20308437943458557, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.2199539691209793, "step": 1797 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 649.0, "completions/mean_length": 2620.675048828125, "completions/mean_terminated_length": 407.6875, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.6974398758727696, "grad_norm": 8.194308280944824, "kl": 27.1875, "learning_rate": 1.4993793061664391e-06, "loss": 0.586, "num_tokens": 344935478.0, "reward": 0.125, "reward_std": 0.12855350971221924, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.11896175146102905, "step": 1798 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1155.0, "completions/mean_length": 2311.050048828125, "completions/mean_terminated_length": 434.5641174316406, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.6978277734678044, "grad_norm": 2.6056602001190186, "kl": 33.53125, "learning_rate": 1.497029862901578e-06, "loss": 0.6477, "num_tokens": 345172658.0, "reward": 0.15125000476837158, "reward_std": 0.16332021355628967, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.18169386684894562, "step": 1799 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 704.0, "completions/mean_length": 2532.35009765625, "completions/mean_terminated_length": 416.8235168457031, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.6982156710628394, "grad_norm": 16.989978790283203, "kl": 61.5625, "learning_rate": 1.4946823982475883e-06, "loss": 0.9872, "num_tokens": 345430518.0, "reward": 0.125, "reward_std": 0.17284725606441498, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.17824745178222656, "step": 1800 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1276.0, "completions/mean_length": 3154.925048828125, "completions/mean_terminated_length": 510.952392578125, "completions/min_length": 218.0, "completions/min_terminated_length": 218.0, "epoch": 0.6986035686578743, "grad_norm": 18.255046844482422, "kl": 43.5625, "learning_rate": 1.4923369159114386e-06, "loss": 0.6361, "num_tokens": 345734520.0, "reward": 0.08749999850988388, "reward_std": 0.10575196892023087, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.030000001192092896, "rewards/unicoder_reward_fn/std": 0.09599578380584717, "step": 1801 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 784.0, "completions/mean_length": 1730.6375732421875, "completions/mean_terminated_length": 456.98077392578125, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.6989914662529092, "grad_norm": 18.51280975341797, "kl": 39.0, "learning_rate": 1.4899934195969646e-06, "loss": 0.8024, "num_tokens": 345927323.0, "reward": 0.21124999225139618, "reward_std": 0.15210214257240295, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.2393953502178192, "step": 1802 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 944.0, "completions/mean_length": 1970.300048828125, "completions/mean_terminated_length": 399.13043212890625, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.6993793638479442, "grad_norm": 5.902615070343018, "kl": 34.1875, "learning_rate": 1.4876519130048682e-06, "loss": 0.6008, "num_tokens": 346128619.0, "reward": 0.22874999046325684, "reward_std": 0.21280212700366974, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.26670753955841064, "step": 1803 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1039.0, "completions/mean_length": 1605.0750732421875, "completions/mean_terminated_length": 472.8363342285156, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "epoch": 0.6997672614429791, "grad_norm": 25.581998825073242, "kl": 30.09375, "learning_rate": 1.4853123998327068e-06, "loss": 0.6118, "num_tokens": 346293849.0, "reward": 0.22625000774860382, "reward_std": 0.19130876660346985, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.1935020387172699, "step": 1804 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 769.0, "completions/mean_length": 2199.949951171875, "completions/mean_terminated_length": 396.3902282714844, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.700155159038014, "grad_norm": 7.315665245056152, "kl": 44.3125, "learning_rate": 1.4829748837748908e-06, "loss": 0.8512, "num_tokens": 346517805.0, "reward": 0.21125002205371857, "reward_std": 0.21211543679237366, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.2692582309246063, "step": 1805 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 904.0, "completions/mean_length": 2845.650146484375, "completions/mean_terminated_length": 391.25927734375, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.7005430566330488, "grad_norm": 28.494855880737305, "kl": 58.0625, "learning_rate": 1.4806393685226783e-06, "loss": 0.8703, "num_tokens": 346795353.0, "reward": 0.1275000125169754, "reward_std": 0.17565634846687317, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.07249999791383743, "rewards/unicoder_reward_fn/std": 0.20682960748672485, "step": 1806 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 585.0, "completions/mean_length": 2438.962646484375, "completions/mean_terminated_length": 413.6944580078125, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.7009309542280838, "grad_norm": 14.195747375488281, "kl": 61.0625, "learning_rate": 1.4783058577641657e-06, "loss": 1.0252, "num_tokens": 347041302.0, "reward": 0.12250000238418579, "reward_std": 0.14075437188148499, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.15457168221473694, "step": 1807 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 712.0, "completions/mean_length": 3635.425048828125, "completions/mean_terminated_length": 411.3999938964844, "completions/min_length": 252.0, "completions/min_terminated_length": 252.0, "epoch": 0.7013188518231187, "grad_norm": 58.77248001098633, "kl": 57.5, "learning_rate": 1.4759743551842853e-06, "loss": 0.7039, "num_tokens": 347379360.0, "reward": 0.06000000238418579, "reward_std": 0.0796637311577797, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.0725223645567894, "step": 1808 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1535.0, "completions/mean_length": 3333.28759765625, "completions/mean_terminated_length": 506.76470947265625, "completions/min_length": 268.0, "completions/min_terminated_length": 268.0, "epoch": 0.7017067494181536, "grad_norm": 43.637542724609375, "kl": 55.3125, "learning_rate": 1.4736448644647986e-06, "loss": 0.7552, "num_tokens": 347699703.0, "reward": 0.14750000834465027, "reward_std": 0.19561052322387695, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.26623252034187317, "step": 1809 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 801.0, "completions/mean_length": 2939.887451171875, "completions/mean_terminated_length": 396.44000244140625, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.7020946470131885, "grad_norm": 34.253963470458984, "kl": 48.40625, "learning_rate": 1.4713173892842878e-06, "loss": 0.7449, "num_tokens": 347980158.0, "reward": 0.1237499937415123, "reward_std": 0.1303315907716751, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.0650000050663948, "rewards/unicoder_reward_fn/std": 0.15186603367328644, "step": 1810 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 764.0, "completions/mean_length": 2542.91259765625, "completions/mean_terminated_length": 441.6764831542969, "completions/min_length": 239.0, "completions/min_terminated_length": 239.0, "epoch": 0.7024825446082235, "grad_norm": 6.933495998382568, "kl": 42.8125, "learning_rate": 1.4689919333181556e-06, "loss": 0.8231, "num_tokens": 348238519.0, "reward": 0.1587500125169754, "reward_std": 0.17594674229621887, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.0949999988079071, "rewards/unicoder_reward_fn/std": 0.18819595873355865, "step": 1811 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1039.0, "completions/mean_length": 1629.300048828125, "completions/mean_terminated_length": 441.629638671875, "completions/min_length": 234.0, "completions/min_terminated_length": 234.0, "epoch": 0.7028704422032583, "grad_norm": 24.877805709838867, "kl": 27.78125, "learning_rate": 1.4666685002386127e-06, "loss": 0.7294, "num_tokens": 348422127.0, "reward": 0.2600000202655792, "reward_std": 0.22794778645038605, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.24066101014614105, "step": 1812 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 767.0, "completions/mean_length": 1920.362548828125, "completions/mean_terminated_length": 392.7872314453125, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.7032583397982932, "grad_norm": 16.97941017150879, "kl": 30.65625, "learning_rate": 1.464347093714679e-06, "loss": 0.8137, "num_tokens": 348621988.0, "reward": 0.22875002026557922, "reward_std": 0.17277191579341888, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.442768931388855, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.27737465500831604, "step": 1813 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1026.0, "completions/mean_length": 1186.6375732421875, "completions/mean_terminated_length": 459.296875, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.7036462373933282, "grad_norm": 25.174625396728516, "kl": 17.2578125, "learning_rate": 1.462027717412171e-06, "loss": 0.4534, "num_tokens": 348761943.0, "reward": 0.23875001072883606, "reward_std": 0.14842520654201508, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.2510610520839691, "step": 1814 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 931.0, "completions/mean_length": 1303.862548828125, "completions/mean_terminated_length": 434.1802978515625, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.7040341349883631, "grad_norm": 13.33190631866455, "kl": 23.546875, "learning_rate": 1.4597103749937002e-06, "loss": 0.6217, "num_tokens": 348924500.0, "reward": 0.1912499964237213, "reward_std": 0.15358057618141174, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.16821850836277008, "step": 1815 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1471.0, "completions/mean_length": 1440.2874755859375, "completions/mean_terminated_length": 495.0339050292969, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "epoch": 0.704422032583398, "grad_norm": 18.12678337097168, "kl": 26.21875, "learning_rate": 1.457395070118668e-06, "loss": 0.5986, "num_tokens": 349088267.0, "reward": 0.26500001549720764, "reward_std": 0.18216544389724731, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.23854783177375793, "step": 1816 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2111.0, "completions/mean_length": 1656.7625732421875, "completions/mean_terminated_length": 482.3148193359375, "completions/min_length": 144.0, "completions/min_terminated_length": 144.0, "epoch": 0.7048099301784329, "grad_norm": 14.043645858764648, "kl": 22.765625, "learning_rate": 1.455081806443255e-06, "loss": 0.437, "num_tokens": 349270184.0, "reward": 0.2912500202655792, "reward_std": 0.21374833583831787, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.20999999344348907, "rewards/unicoder_reward_fn/std": 0.3083644211292267, "step": 1817 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 766.0, "completions/mean_length": 1559.1375732421875, "completions/mean_terminated_length": 406.0181579589844, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.7051978277734678, "grad_norm": 22.53274917602539, "kl": 26.9375, "learning_rate": 1.452770587620422e-06, "loss": 0.5926, "num_tokens": 349440595.0, "reward": 0.19625000655651093, "reward_std": 0.1350720226764679, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.22690823674201965, "step": 1818 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 1434.0, "completions/mean_length": 1356.1624755859375, "completions/mean_terminated_length": 442.88336181640625, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.7055857253685027, "grad_norm": 33.98036193847656, "kl": 39.46875, "learning_rate": 1.4504614172998984e-06, "loss": 0.9745, "num_tokens": 349603624.0, "reward": 0.22875002026557922, "reward_std": 0.1585894078016281, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.26191842555999756, "step": 1819 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1667.0, "completions/mean_length": 1514.0125732421875, "completions/mean_terminated_length": 472.15789794921875, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.7059736229635376, "grad_norm": 129.30857849121094, "kl": 24.03125, "learning_rate": 1.4481542991281782e-06, "loss": 0.6006, "num_tokens": 349773449.0, "reward": 0.2562499940395355, "reward_std": 0.16958491504192352, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.17249999940395355, "rewards/unicoder_reward_fn/std": 0.22444023191928864, "step": 1820 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1429.0, "completions/mean_length": 2301.462646484375, "completions/mean_terminated_length": 414.8974304199219, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.7063615205585725, "grad_norm": 100.93355560302734, "kl": 51.8125, "learning_rate": 1.445849236748518e-06, "loss": 0.994, "num_tokens": 349998622.0, "reward": 0.21125002205371857, "reward_std": 0.2483011931180954, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2604280710220337, "step": 1821 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 704.0, "completions/mean_length": 2685.087646484375, "completions/mean_terminated_length": 454.9354553222656, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.7067494181536075, "grad_norm": 68.12870025634766, "kl": 76.75, "learning_rate": 1.4435462338009238e-06, "loss": 1.1636, "num_tokens": 350256597.0, "reward": 0.14625000953674316, "reward_std": 0.1477232724428177, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.16155888140201569, "step": 1822 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7375, "completions/max_length": 4096.0, "completions/max_terminated_length": 702.0, "completions/mean_length": 3149.862548828125, "completions/mean_terminated_length": 491.66668701171875, "completions/min_length": 286.0, "completions/min_terminated_length": 286.0, "epoch": 0.7071373157486424, "grad_norm": 55.08815383911133, "kl": 98.375, "learning_rate": 1.4412452939221535e-06, "loss": 1.2694, "num_tokens": 350557738.0, "reward": 0.11124999821186066, "reward_std": 0.17131407558918, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.1671617180109024, "step": 1823 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6375, "completions/max_length": 4096.0, "completions/max_terminated_length": 949.0, "completions/mean_length": 2781.300048828125, "completions/mean_terminated_length": 469.2413635253906, "completions/min_length": 271.0, "completions/min_terminated_length": 271.0, "epoch": 0.7075252133436772, "grad_norm": 30.137222290039062, "kl": 85.75, "learning_rate": 1.4389464207457042e-06, "loss": 1.255, "num_tokens": 350831146.0, "reward": 0.17499999701976776, "reward_std": 0.23064318299293518, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.2346556931734085, "step": 1824 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 945.0, "completions/mean_length": 3503.800048828125, "completions/mean_terminated_length": 451.69232177734375, "completions/min_length": 321.0, "completions/min_terminated_length": 321.0, "epoch": 0.7079131109387122, "grad_norm": 56.36732864379883, "kl": 101.75, "learning_rate": 1.43664961790181e-06, "loss": 1.2413, "num_tokens": 351156498.0, "reward": 0.07375001162290573, "reward_std": 0.12091248482465744, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.042500000447034836, "rewards/unicoder_reward_fn/std": 0.1338571012020111, "step": 1825 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 481.0, "completions/mean_length": 4004.699951171875, "completions/mean_terminated_length": 444.0, "completions/min_length": 407.0, "completions/min_terminated_length": 407.0, "epoch": 0.7083010085337471, "grad_norm": 71.68937683105469, "kl": 106.625, "learning_rate": 1.4343548890174374e-06, "loss": 1.1184, "num_tokens": 351520994.0, "reward": 0.017500001937150955, "reward_std": 0.044613540172576904, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531530380249, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944272249937057, "step": 1826 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.708688906128782, "grad_norm": 71.90422058105469, "kl": 97.25, "learning_rate": 1.432062237716275e-06, "loss": 0.9716, "num_tokens": 351898170.0, "reward": 0.003750000149011612, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1827 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 645.0, "completions/mean_length": 4007.237548828125, "completions/mean_terminated_length": 545.5, "completions/min_length": 446.0, "completions/min_terminated_length": 446.0, "epoch": 0.7090768037238169, "grad_norm": 93.09681701660156, "kl": 85.75, "learning_rate": 1.4297716676187337e-06, "loss": 0.9054, "num_tokens": 352268157.0, "reward": 0.007500000298023224, "reward_std": 0.018771234899759293, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1828 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 403.0, "completions/mean_length": 4001.762451171875, "completions/mean_terminated_length": 326.5, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "epoch": 0.7094647013188519, "grad_norm": 56.42093276977539, "kl": 79.75, "learning_rate": 1.427483182341936e-06, "loss": 0.8447, "num_tokens": 352639402.0, "reward": 0.003750000149011612, "reward_std": 0.008164633996784687, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1829 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 719.0, "completions/mean_length": 4011.462646484375, "completions/mean_terminated_length": 714.5, "completions/min_length": 710.0, "completions/min_terminated_length": 710.0, "epoch": 0.7098525989138867, "grad_norm": 43.73860549926758, "kl": 65.0625, "learning_rate": 1.4251967854997107e-06, "loss": 0.6978, "num_tokens": 353009559.0, "reward": 0.02500000037252903, "reward_std": 0.058346886187791824, "rewards/format_reward/mean": 0.125, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.08015017956495285, "step": 1830 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 529.0, "completions/mean_length": 3958.5625, "completions/mean_terminated_length": 431.0, "completions/min_length": 331.0, "completions/min_terminated_length": 331.0, "epoch": 0.7102404965089216, "grad_norm": 44.779273986816406, "kl": 54.5625, "learning_rate": 1.422912480702593e-06, "loss": 0.6334, "num_tokens": 353378524.0, "reward": 0.0162500012665987, "reward_std": 0.04053083062171936, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434914350509644, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.06708204001188278, "step": 1831 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 1.0, "completions/max_length": 4096.0, "completions/max_terminated_length": 0.0, "completions/mean_length": 4096.0, "completions/mean_terminated_length": 0.0, "completions/min_length": 4096.0, "completions/min_terminated_length": 0.0, "epoch": 0.7106283941039565, "grad_norm": 48.857154846191406, "kl": 43.5, "learning_rate": 1.420630271557811e-06, "loss": 0.4352, "num_tokens": 353762212.0, "reward": 0.005000000353902578, "reward_std": 0.01414213515818119, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1832 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.975, "completions/max_length": 4096.0, "completions/max_terminated_length": 397.0, "completions/mean_length": 4002.625, "completions/mean_terminated_length": 361.0, "completions/min_length": 325.0, "completions/min_terminated_length": 325.0, "epoch": 0.7110162916989915, "grad_norm": 39.35039520263672, "kl": 31.96875, "learning_rate": 1.4183501616692835e-06, "loss": 0.3671, "num_tokens": 354131206.0, "reward": 0.01875000074505806, "reward_std": 0.046745337545871735, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434914350509644, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.08944272249937057, "step": 1833 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9875, "completions/max_length": 4096.0, "completions/max_terminated_length": 408.0, "completions/mean_length": 4049.900146484375, "completions/mean_terminated_length": 408.0, "completions/min_length": 408.0, "completions/min_terminated_length": 408.0, "epoch": 0.7114041892940264, "grad_norm": 19.351051330566406, "kl": 24.46875, "learning_rate": 1.4160721546376168e-06, "loss": 0.2753, "num_tokens": 354501862.0, "reward": 0.002500000176951289, "reward_std": 0.007071067579090595, "rewards/format_reward/mean": 0.02500000037252903, "rewards/format_reward/std": 0.15710999071598053, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1834 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 518.0, "completions/mean_length": 3956.262451171875, "completions/mean_terminated_length": 369.66668701171875, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.7117920868890613, "grad_norm": 8.70661735534668, "kl": 15.0625, "learning_rate": 1.4137962540600943e-06, "loss": 0.2279, "num_tokens": 354861811.0, "reward": 0.007500000298023224, "reward_std": 0.018771236762404442, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360678762197495, "step": 1835 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.95, "completions/max_length": 4096.0, "completions/max_terminated_length": 434.0, "completions/mean_length": 3906.625, "completions/mean_terminated_length": 308.5, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "epoch": 0.7121799844840963, "grad_norm": 1.658432960510254, "kl": 9.703125, "learning_rate": 1.4115224635306738e-06, "loss": 0.1996, "num_tokens": 355232581.0, "reward": 0.021250000223517418, "reward_std": 0.048917364329099655, "rewards/format_reward/mean": 0.0625, "rewards/format_reward/std": 0.2435886710882187, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.09426598995923996, "step": 1836 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9625, "completions/max_length": 4096.0, "completions/max_terminated_length": 692.0, "completions/mean_length": 3962.987548828125, "completions/mean_terminated_length": 549.0, "completions/min_length": 314.0, "completions/min_terminated_length": 314.0, "epoch": 0.7125678820791311, "grad_norm": 4.057839870452881, "kl": 5.421875, "learning_rate": 1.4092507866399795e-06, "loss": 0.1359, "num_tokens": 355597140.0, "reward": 0.003750000149011612, "reward_std": 0.010606600902974606, "rewards/format_reward/mean": 0.03750000149011612, "rewards/format_reward/std": 0.1911821961402893, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1837 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 580.0, "completions/mean_length": 3617.675048828125, "completions/mean_terminated_length": 269.3999938964844, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.712955779674166, "grad_norm": 13.127884864807129, "kl": 2.515625, "learning_rate": 1.4069812269753016e-06, "loss": 0.1623, "num_tokens": 355936506.0, "reward": 0.021250000223517418, "reward_std": 0.05309201031923294, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434914350509644, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.09192044287919998, "step": 1838 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 690.0, "completions/mean_length": 3865.60009765625, "completions/mean_terminated_length": 409.6000061035156, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.7133436772692009, "grad_norm": 11.94927978515625, "kl": 1.88671875, "learning_rate": 1.4047137881205835e-06, "loss": 0.1537, "num_tokens": 356301442.0, "reward": 0.027500001713633537, "reward_std": 0.05711568146944046, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.12568798661231995, "step": 1839 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 496.0, "completions/mean_length": 3575.012451171875, "completions/mean_terminated_length": 307.0, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.7137315748642359, "grad_norm": 26.297901153564453, "kl": 1.83984375, "learning_rate": 1.4024484736564197e-06, "loss": 0.2143, "num_tokens": 356635067.0, "reward": 0.030000001192092896, "reward_std": 0.0726429745554924, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.061828672885894775, "step": 1840 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 659.0, "completions/mean_length": 3821.66259765625, "completions/mean_terminated_length": 438.16668701171875, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.7141194724592708, "grad_norm": 11.847723960876465, "kl": 1.75, "learning_rate": 1.4001852871600526e-06, "loss": 0.1216, "num_tokens": 356983864.0, "reward": 0.036250002682209015, "reward_std": 0.0986490398645401, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434914350509644, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.14493560791015625, "step": 1841 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 626.0, "completions/mean_length": 3716.85009765625, "completions/mean_terminated_length": 304.5, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.7145073700543056, "grad_norm": 12.360880851745605, "kl": 2.171875, "learning_rate": 1.3979242322053612e-06, "loss": 0.1066, "num_tokens": 357338300.0, "reward": 0.010000000707805157, "reward_std": 0.023400336503982544, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1842 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 437.0, "completions/mean_length": 3810.375, "completions/mean_terminated_length": 287.66668701171875, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.7148952676493405, "grad_norm": 12.815220832824707, "kl": 3.109375, "learning_rate": 1.3956653123628617e-06, "loss": 0.152, "num_tokens": 357689378.0, "reward": 0.02250000275671482, "reward_std": 0.05731619521975517, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 1843 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9125, "completions/max_length": 4096.0, "completions/max_terminated_length": 648.0, "completions/mean_length": 3761.137451171875, "completions/mean_terminated_length": 269.0, "completions/min_length": 98.0, "completions/min_terminated_length": 98.0, "epoch": 0.7152831652443755, "grad_norm": 9.190089225769043, "kl": 4.05078125, "learning_rate": 1.3934085311996954e-06, "loss": 0.1154, "num_tokens": 358032557.0, "reward": 0.02500000037252903, "reward_std": 0.06826870888471603, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.017500000074505806, "rewards/unicoder_reward_fn/std": 0.09109044820070267, "step": 1844 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 445.0, "completions/mean_length": 3853.387451171875, "completions/mean_terminated_length": 214.1999969482422, "completions/min_length": 64.0, "completions/min_terminated_length": 64.0, "epoch": 0.7156710628394104, "grad_norm": 6.924036502838135, "kl": 5.546875, "learning_rate": 1.3911538922796275e-06, "loss": 0.1062, "num_tokens": 358384252.0, "reward": 0.017500000074505806, "reward_std": 0.04705550894141197, "rewards/format_reward/mean": 0.07500000298023224, "rewards/format_reward/std": 0.2650531530380249, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.06284399330615997, "step": 1845 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 474.0, "completions/mean_length": 3712.712646484375, "completions/mean_terminated_length": 263.125, "completions/min_length": 144.0, "completions/min_terminated_length": 144.0, "epoch": 0.7160589604344453, "grad_norm": 12.27550983428955, "kl": 8.296875, "learning_rate": 1.3889013991630418e-06, "loss": 0.294, "num_tokens": 358735653.0, "reward": 0.015000000596046448, "reward_std": 0.03998443856835365, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.03142199665307999, "step": 1846 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9125, "completions/max_length": 4096.0, "completions/max_terminated_length": 485.0, "completions/mean_length": 3758.175048828125, "completions/mean_terminated_length": 235.1428680419922, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.7164468580294803, "grad_norm": 9.517016410827637, "kl": 10.953125, "learning_rate": 1.3866510554069312e-06, "loss": 0.2415, "num_tokens": 359088675.0, "reward": 0.021250000223517418, "reward_std": 0.0530330054461956, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.054191410541534424, "step": 1847 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9125, "completions/max_length": 4096.0, "completions/max_terminated_length": 715.0, "completions/mean_length": 3778.012451171875, "completions/mean_terminated_length": 461.857177734375, "completions/min_length": 267.0, "completions/min_terminated_length": 267.0, "epoch": 0.7168347556245152, "grad_norm": 1.6595836877822876, "kl": 13.984375, "learning_rate": 1.3844028645648971e-06, "loss": 0.2589, "num_tokens": 359440404.0, "reward": 0.01875000260770321, "reward_std": 0.0441502146422863, "rewards/format_reward/mean": 0.08749999850988388, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.07044164836406708, "step": 1848 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9375, "completions/max_length": 4096.0, "completions/max_terminated_length": 423.0, "completions/mean_length": 3857.925048828125, "completions/mean_terminated_length": 286.8000183105469, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.71722265321955, "grad_norm": 6.949640274047852, "kl": 18.8125, "learning_rate": 1.3821568301871386e-06, "loss": 0.2741, "num_tokens": 359796254.0, "reward": 0.017500001937150955, "reward_std": 0.04949747398495674, "rewards/format_reward/mean": 0.05000000074505806, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.08015017956495285, "step": 1849 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.925, "completions/max_length": 4096.0, "completions/max_terminated_length": 471.0, "completions/mean_length": 3812.775146484375, "completions/mean_terminated_length": 319.66668701171875, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.7176105508145849, "grad_norm": 4.359829425811768, "kl": 19.65625, "learning_rate": 1.3799129558204505e-06, "loss": 0.3438, "num_tokens": 360156276.0, "reward": 0.010000000707805157, "reward_std": 0.025842303410172462, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.0, "rewards/unicoder_reward_fn/std": 0.0, "step": 1850 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.9, "completions/max_length": 4096.0, "completions/max_terminated_length": 548.0, "completions/mean_length": 3720.875, "completions/mean_terminated_length": 344.75, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.7179984484096199, "grad_norm": 3.711099147796631, "kl": 20.59375, "learning_rate": 1.3776712450082174e-06, "loss": 0.3562, "num_tokens": 360506418.0, "reward": 0.030000001192092896, "reward_std": 0.05663713440299034, "rewards/format_reward/mean": 0.15000000596046448, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.06953734904527664, "step": 1851 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8625, "completions/max_length": 4096.0, "completions/max_terminated_length": 411.0, "completions/mean_length": 3575.987548828125, "completions/mean_terminated_length": 314.0909118652344, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "epoch": 0.7183863460046548, "grad_norm": 2.1114394664764404, "kl": 22.65625, "learning_rate": 1.3754317012904058e-06, "loss": 0.4439, "num_tokens": 360838553.0, "reward": 0.03875000402331352, "reward_std": 0.09983368217945099, "rewards/format_reward/mean": 0.11249999701976776, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.11797340214252472, "step": 1852 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8375, "completions/max_length": 4096.0, "completions/max_terminated_length": 521.0, "completions/mean_length": 3482.512451171875, "completions/mean_terminated_length": 320.69232177734375, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.7187742435996897, "grad_norm": 4.141829967498779, "kl": 23.21875, "learning_rate": 1.3731943282035624e-06, "loss": 0.4116, "num_tokens": 361163306.0, "reward": 0.019999999552965164, "reward_std": 0.035868797451257706, "rewards/format_reward/mean": 0.10000000149011612, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.04386400431394577, "step": 1853 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.875, "completions/max_length": 4096.0, "completions/max_terminated_length": 868.0, "completions/mean_length": 3640.91259765625, "completions/mean_terminated_length": 455.3000183105469, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.7191621411947245, "grad_norm": 6.3152289390563965, "kl": 24.375, "learning_rate": 1.3709591292808037e-06, "loss": 0.42, "num_tokens": 361508507.0, "reward": 0.04125000163912773, "reward_std": 0.10568752139806747, "rewards/format_reward/mean": 0.13750000298023224, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.12218993902206421, "step": 1854 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.8125, "completions/max_length": 4096.0, "completions/max_terminated_length": 689.0, "completions/mean_length": 3405.6875, "completions/mean_terminated_length": 414.3333435058594, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.7195500387897595, "grad_norm": 8.457054138183594, "kl": 24.6875, "learning_rate": 1.368726108051813e-06, "loss": 0.5545, "num_tokens": 361825434.0, "reward": 0.04374999925494194, "reward_std": 0.09601832181215286, "rewards/format_reward/mean": 0.1875, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.10250039398670197, "step": 1855 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.65, "completions/max_length": 4096.0, "completions/max_terminated_length": 1425.0, "completions/mean_length": 2799.28759765625, "completions/mean_terminated_length": 391.1071472167969, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.7199379363847944, "grad_norm": 14.682060241699219, "kl": 24.03125, "learning_rate": 1.3664952680428374e-06, "loss": 0.6267, "num_tokens": 362093505.0, "reward": 0.07750000804662704, "reward_std": 0.1252894401550293, "rewards/format_reward/mean": 0.2750000059604645, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.17283928394317627, "step": 1856 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 592.0, "completions/mean_length": 2920.9375, "completions/mean_terminated_length": 335.79998779296875, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.7203258339798293, "grad_norm": 11.192803382873535, "kl": 24.96875, "learning_rate": 1.3642666127766757e-06, "loss": 0.6289, "num_tokens": 362368988.0, "reward": 0.0637500062584877, "reward_std": 0.12288320064544678, "rewards/format_reward/mean": 0.2874999940395355, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.03500000014901161, "rewards/unicoder_reward_fn/std": 0.13416406512260437, "step": 1857 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.825, "completions/max_length": 4096.0, "completions/max_terminated_length": 457.0, "completions/mean_length": 3433.28759765625, "completions/mean_terminated_length": 309.0714416503906, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.7207137315748642, "grad_norm": 1.416722059249878, "kl": 31.25, "learning_rate": 1.3620401457726815e-06, "loss": 0.5804, "num_tokens": 362686739.0, "reward": 0.03125, "reward_std": 0.07672590017318726, "rewards/format_reward/mean": 0.16249999403953552, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.08282970637083054, "step": 1858 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 607.0, "completions/mean_length": 3016.53759765625, "completions/mean_terminated_length": 341.34783935546875, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.7211016291698992, "grad_norm": 6.957376480102539, "kl": 26.28125, "learning_rate": 1.3598158705467482e-06, "loss": 0.5482, "num_tokens": 362970462.0, "reward": 0.07750000804662704, "reward_std": 0.15691840648651123, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.17773184180259705, "step": 1859 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 1823.0, "completions/mean_length": 2970.25, "completions/mean_terminated_length": 343.5, "completions/min_length": 51.0, "completions/min_terminated_length": 51.0, "epoch": 0.721489526764934, "grad_norm": 17.358285903930664, "kl": 30.1875, "learning_rate": 1.3575937906113097e-06, "loss": 0.649, "num_tokens": 363265546.0, "reward": 0.05000000074505806, "reward_std": 0.08805958181619644, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.10250039398670197, "step": 1860 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 564.0, "completions/mean_length": 3009.512451171875, "completions/mean_terminated_length": 316.9130554199219, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.7218774243599689, "grad_norm": 1.5105916261672974, "kl": 34.90625, "learning_rate": 1.3553739094753349e-06, "loss": 0.6322, "num_tokens": 363546043.0, "reward": 0.07000000029802322, "reward_std": 0.11587842553853989, "rewards/format_reward/mean": 0.20000000298023224, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.14052343368530273, "step": 1861 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.75, "completions/max_length": 4096.0, "completions/max_terminated_length": 841.0, "completions/mean_length": 3166.25, "completions/mean_terminated_length": 377.0, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.7222653219550039, "grad_norm": 3.351858377456665, "kl": 35.625, "learning_rate": 1.3531562306443188e-06, "loss": 0.6329, "num_tokens": 363849191.0, "reward": 0.055000003427267075, "reward_std": 0.10162343829870224, "rewards/format_reward/mean": 0.22499999403953552, "rewards/format_reward/std": 0.4202169179916382, "rewards/unicoder_reward_fn/mean": 0.0325000025331974, "rewards/unicoder_reward_fn/std": 0.11225535720586777, "step": 1862 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 870.0, "completions/mean_length": 3043.35009765625, "completions/mean_terminated_length": 434.60870361328125, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.7226532195500388, "grad_norm": 3.142378807067871, "kl": 33.65625, "learning_rate": 1.3509407576202776e-06, "loss": 0.5725, "num_tokens": 364144371.0, "reward": 0.11250000447034836, "reward_std": 0.1897345632314682, "rewards/format_reward/mean": 0.30000001192092896, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.20547877252101898, "step": 1863 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.65, "completions/max_length": 4096.0, "completions/max_terminated_length": 3052.0, "completions/mean_length": 2809.47509765625, "completions/mean_terminated_length": 420.21429443359375, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.7230411171450737, "grad_norm": 7.796204090118408, "kl": 34.5, "learning_rate": 1.3487274939017475e-06, "loss": 0.7062, "num_tokens": 364419249.0, "reward": 0.05874999985098839, "reward_std": 0.08243938535451889, "rewards/format_reward/mean": 0.36250001192092896, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.02250000089406967, "rewards/unicoder_reward_fn/std": 0.0899718776345253, "step": 1864 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 1588.0, "completions/mean_length": 2534.85009765625, "completions/mean_terminated_length": 422.70587158203125, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.7234290147401086, "grad_norm": 4.96218729019165, "kl": 30.34375, "learning_rate": 1.3465164429837719e-06, "loss": 0.6641, "num_tokens": 364666461.0, "reward": 0.11375000327825546, "reward_std": 0.1474871039390564, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.1594293713569641, "step": 1865 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 1632.0, "completions/mean_length": 3113.72509765625, "completions/mean_terminated_length": 524.0909423828125, "completions/min_length": 188.0, "completions/min_terminated_length": 188.0, "epoch": 0.7238169123351436, "grad_norm": 2.792783737182617, "kl": 33.46875, "learning_rate": 1.3443076083579043e-06, "loss": 0.559, "num_tokens": 364970895.0, "reward": 0.07875000685453415, "reward_std": 0.10818679630756378, "rewards/format_reward/mean": 0.38749998807907104, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.12887045741081238, "step": 1866 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1475.0, "completions/mean_length": 2719.574951171875, "completions/mean_terminated_length": 425.5333557128906, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.7242048099301784, "grad_norm": 13.175843238830566, "kl": 38.125, "learning_rate": 1.3421009935121955e-06, "loss": 0.6999, "num_tokens": 365235173.0, "reward": 0.10625000298023224, "reward_std": 0.17821229994297028, "rewards/format_reward/mean": 0.36250001192092896, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.16868816316127777, "step": 1867 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 1374.0, "completions/mean_length": 2531.699951171875, "completions/mean_terminated_length": 415.29412841796875, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.7245927075252133, "grad_norm": 7.533036708831787, "kl": 37.5625, "learning_rate": 1.3398966019311896e-06, "loss": 0.7545, "num_tokens": 365490045.0, "reward": 0.13250000774860382, "reward_std": 0.18498365581035614, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.17957404255867004, "step": 1868 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1745.0, "completions/mean_length": 2390.512451171875, "completions/mean_terminated_length": 408.45947265625, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.7249806051202482, "grad_norm": 5.629456043243408, "kl": 38.3125, "learning_rate": 1.3376944370959243e-06, "loss": 0.7338, "num_tokens": 365728294.0, "reward": 0.13625000417232513, "reward_std": 0.16775372624397278, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.15861345827579498, "step": 1869 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 1406.0, "completions/mean_length": 2126.387451171875, "completions/mean_terminated_length": 344.3571472167969, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.7253685027152832, "grad_norm": 2.7701494693756104, "kl": 45.4375, "learning_rate": 1.3354945024839166e-06, "loss": 0.9076, "num_tokens": 365940277.0, "reward": 0.15250001847743988, "reward_std": 0.16811013221740723, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.20865754783153534, "step": 1870 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 2287.0, "completions/mean_length": 2439.60009765625, "completions/mean_terminated_length": 415.1111145019531, "completions/min_length": 118.0, "completions/min_terminated_length": 118.0, "epoch": 0.7257564003103181, "grad_norm": 13.55462646484375, "kl": 45.125, "learning_rate": 1.333296801569165e-06, "loss": 0.9282, "num_tokens": 366187389.0, "reward": 0.10500001162290573, "reward_std": 0.11003196239471436, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.15393571555614471, "step": 1871 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2278.0, "completions/mean_length": 1819.75, "completions/mean_terminated_length": 454.0, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.726144297905353, "grad_norm": 21.286684036254883, "kl": 42.78125, "learning_rate": 1.3311013378221388e-06, "loss": 0.841, "num_tokens": 366386313.0, "reward": 0.14750002324581146, "reward_std": 0.17359420657157898, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.17957404255867004, "step": 1872 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2739.0, "completions/mean_length": 1963.125, "completions/mean_terminated_length": 465.574462890625, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.726532195500388, "grad_norm": 8.648824691772461, "kl": 48.5625, "learning_rate": 1.3289081147097737e-06, "loss": 0.9402, "num_tokens": 366582227.0, "reward": 0.17249999940395355, "reward_std": 0.21835088729858398, "rewards/format_reward/mean": 0.4749999940395355, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.23248793184757233, "step": 1873 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1195.0, "completions/mean_length": 1832.362548828125, "completions/mean_terminated_length": 400.2652893066406, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.7269200930954228, "grad_norm": 15.087307929992676, "kl": 43.5625, "learning_rate": 1.3267171356954701e-06, "loss": 0.8767, "num_tokens": 366773560.0, "reward": 0.1887500137090683, "reward_std": 0.1738254725933075, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.19713132083415985, "step": 1874 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2262.0, "completions/mean_length": 1962.4375, "completions/mean_terminated_length": 385.45654296875, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.7273079906904577, "grad_norm": 30.828968048095703, "kl": 41.625, "learning_rate": 1.3245284042390815e-06, "loss": 0.7918, "num_tokens": 366978515.0, "reward": 0.13375000655651093, "reward_std": 0.15743954479694366, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.1695116013288498, "step": 1875 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1978.0, "completions/mean_length": 1756.2000732421875, "completions/mean_terminated_length": 425.7254943847656, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.7276958882854926, "grad_norm": 5.728309631347656, "kl": 47.1875, "learning_rate": 1.3223419237969154e-06, "loss": 0.9627, "num_tokens": 367167571.0, "reward": 0.1575000137090683, "reward_std": 0.17268089950084686, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.1935674399137497, "step": 1876 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 940.0, "completions/mean_length": 1340.625, "completions/mean_terminated_length": 359.8983154296875, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.7280837858805276, "grad_norm": 3.6356732845306396, "kl": 47.25, "learning_rate": 1.3201576978217211e-06, "loss": 0.984, "num_tokens": 367319277.0, "reward": 0.19625000655651093, "reward_std": 0.22449302673339844, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.24256019294261932, "step": 1877 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 1541.0, "completions/mean_length": 1689.6375732421875, "completions/mean_terminated_length": 393.90386962890625, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.7284716834755625, "grad_norm": 2.212872266769409, "kl": 56.3125, "learning_rate": 1.31797572976269e-06, "loss": 1.1056, "num_tokens": 367500336.0, "reward": 0.2212500125169754, "reward_std": 0.21526546776294708, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.27206429839134216, "step": 1878 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 963.0, "completions/mean_length": 1449.425048828125, "completions/mean_terminated_length": 381.5087890625, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.7288595810705973, "grad_norm": 4.4206342697143555, "kl": 44.875, "learning_rate": 1.3157960230654482e-06, "loss": 0.826, "num_tokens": 367660050.0, "reward": 0.2550000250339508, "reward_std": 0.26848724484443665, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.1850000023841858, "rewards/unicoder_reward_fn/std": 0.31466275453567505, "step": 1879 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2974.0, "completions/mean_length": 1478.800048828125, "completions/mean_terminated_length": 422.7368469238281, "completions/min_length": 90.0, "completions/min_terminated_length": 90.0, "epoch": 0.7292474786656322, "grad_norm": 3.3466756343841553, "kl": 40.21875, "learning_rate": 1.3136185811720497e-06, "loss": 0.7116, "num_tokens": 367829178.0, "reward": 0.20000000298023224, "reward_std": 0.1952095478773117, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.21132493019104004, "step": 1880 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 722.0, "completions/mean_length": 1348.9625244140625, "completions/mean_terminated_length": 371.2033996582031, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.7296353762606672, "grad_norm": 12.209512710571289, "kl": 46.4375, "learning_rate": 1.3114434075209717e-06, "loss": 0.9614, "num_tokens": 367984735.0, "reward": 0.18500001728534698, "reward_std": 0.17657534778118134, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.23350656032562256, "step": 1881 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 1336.0, "completions/mean_length": 1430.6375732421875, "completions/mean_terminated_length": 419.637939453125, "completions/min_length": 152.0, "completions/min_terminated_length": 152.0, "epoch": 0.7300232738557021, "grad_norm": 4.566563129425049, "kl": 43.84375, "learning_rate": 1.3092705055471117e-06, "loss": 0.7935, "num_tokens": 368152906.0, "reward": 0.2537500262260437, "reward_std": 0.20641222596168518, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.2627868950366974, "step": 1882 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1223.0, "completions/mean_length": 1216.3875732421875, "completions/mean_terminated_length": 380.3709411621094, "completions/min_length": 135.0, "completions/min_terminated_length": 135.0, "epoch": 0.730411171450737, "grad_norm": 5.448374271392822, "kl": 40.75, "learning_rate": 1.3070998786817776e-06, "loss": 0.6974, "num_tokens": 368297353.0, "reward": 0.24625001847743988, "reward_std": 0.19663293659687042, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.16750000417232513, "rewards/unicoder_reward_fn/std": 0.26897603273391724, "step": 1883 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3416.0, "completions/mean_length": 1498.175048828125, "completions/mean_terminated_length": 449.9298400878906, "completions/min_length": 201.0, "completions/min_terminated_length": 201.0, "epoch": 0.730799069045772, "grad_norm": 2.829681634902954, "kl": 57.25, "learning_rate": 1.3049315303526853e-06, "loss": 1.0256, "num_tokens": 368472599.0, "reward": 0.2900000512599945, "reward_std": 0.23634663224220276, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.2175000160932541, "rewards/unicoder_reward_fn/std": 0.2685050070285797, "step": 1884 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 818.0, "completions/mean_length": 1674.5625, "completions/mean_terminated_length": 441.0, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.7311869666408068, "grad_norm": 7.674860954284668, "kl": 57.1875, "learning_rate": 1.3027654639839537e-06, "loss": 1.0803, "num_tokens": 368651148.0, "reward": 0.23125000298023224, "reward_std": 0.1965857893228531, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.2425471395254135, "step": 1885 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1531.0, "completions/mean_length": 1311.7125244140625, "completions/mean_terminated_length": 444.4753723144531, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.7315748642358417, "grad_norm": 16.715482711791992, "kl": 40.0, "learning_rate": 1.3006016829960966e-06, "loss": 0.9454, "num_tokens": 368800453.0, "reward": 0.3062500059604645, "reward_std": 0.20586013793945312, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.23000001907348633, "rewards/unicoder_reward_fn/std": 0.3070480227470398, "step": 1886 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1109.0, "completions/mean_length": 1283.3375244140625, "completions/mean_terminated_length": 407.26226806640625, "completions/min_length": 159.0, "completions/min_terminated_length": 159.0, "epoch": 0.7319627618308766, "grad_norm": 63.13903045654297, "kl": 51.5625, "learning_rate": 1.2984401908060198e-06, "loss": 1.17, "num_tokens": 368957752.0, "reward": 0.23125000298023224, "reward_std": 0.1859249323606491, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.19537371397018433, "step": 1887 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1537.0, "completions/mean_length": 1868.9500732421875, "completions/mean_terminated_length": 460.0, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "epoch": 0.7323506594259116, "grad_norm": 7.844636917114258, "kl": 52.25, "learning_rate": 1.2962809908270136e-06, "loss": 0.9638, "num_tokens": 369150356.0, "reward": 0.2212500125169754, "reward_std": 0.20155644416809082, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.2699742317199707, "step": 1888 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1020.0, "completions/mean_length": 1946.9625244140625, "completions/mean_terminated_length": 438.0638122558594, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.7327385570209465, "grad_norm": 161.33016967773438, "kl": 50.4375, "learning_rate": 1.294124086468751e-06, "loss": 0.9084, "num_tokens": 369362201.0, "reward": 0.23500001430511475, "reward_std": 0.20250354707241058, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.264527291059494, "step": 1889 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1657.0, "completions/mean_length": 1983.925048828125, "completions/mean_terminated_length": 422.8260803222656, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.7331264546159814, "grad_norm": 37.87826156616211, "kl": 56.4375, "learning_rate": 1.2919694811372767e-06, "loss": 1.1362, "num_tokens": 369568243.0, "reward": 0.23749999701976776, "reward_std": 0.22534742951393127, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.2487436830997467, "step": 1890 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 996.0, "completions/mean_length": 1807.6375732421875, "completions/mean_terminated_length": 434.6199951171875, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "epoch": 0.7335143522110162, "grad_norm": 5.791954040527344, "kl": 54.75, "learning_rate": 1.2898171782350088e-06, "loss": 1.0348, "num_tokens": 369766318.0, "reward": 0.1875, "reward_std": 0.17149792611598969, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.17938008904457092, "step": 1891 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 906.0, "completions/mean_length": 2447.050048828125, "completions/mean_terminated_length": 431.6666564941406, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.7339022498060512, "grad_norm": 56.435203552246094, "kl": 53.3125, "learning_rate": 1.2876671811607256e-06, "loss": 1.0147, "num_tokens": 370015402.0, "reward": 0.1287500113248825, "reward_std": 0.13289710879325867, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.12887045741081238, "step": 1892 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 1444.0, "completions/mean_length": 1931.8499755859375, "completions/mean_terminated_length": 489.0833435058594, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.7342901474010861, "grad_norm": 24.460905075073242, "kl": 49.0, "learning_rate": 1.285519493309566e-06, "loss": 0.9157, "num_tokens": 370214622.0, "reward": 0.2225000113248825, "reward_std": 0.24009381234645844, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.24955657124519348, "step": 1893 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3878.0, "completions/mean_length": 1514.4000244140625, "completions/mean_terminated_length": 595.5254516601562, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.734678044996121, "grad_norm": 5.9136152267456055, "kl": 34.40625, "learning_rate": 1.2833741180730233e-06, "loss": 0.7321, "num_tokens": 370384110.0, "reward": 0.2262500375509262, "reward_std": 0.2007472962141037, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.25452861189842224, "step": 1894 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1706.0, "completions/mean_length": 1680.9375, "completions/mean_terminated_length": 450.6226501464844, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.735065942591156, "grad_norm": 10.529330253601074, "kl": 43.625, "learning_rate": 1.2812310588389367e-06, "loss": 0.9101, "num_tokens": 370560089.0, "reward": 0.18375001847743988, "reward_std": 0.15231037139892578, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.2084299474954605, "step": 1895 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1694.0, "completions/mean_length": 1385.2750244140625, "completions/mean_terminated_length": 420.440673828125, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.7354538401861909, "grad_norm": 5.389052867889404, "kl": 31.53125, "learning_rate": 1.2790903189914906e-06, "loss": 0.6505, "num_tokens": 370715471.0, "reward": 0.2512499988079071, "reward_std": 0.1816333383321762, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.23045086860656738, "step": 1896 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2211.0, "completions/mean_length": 1879.3250732421875, "completions/mean_terminated_length": 549.3200073242188, "completions/min_length": 202.0, "completions/min_terminated_length": 202.0, "epoch": 0.7358417377812257, "grad_norm": 56.0130500793457, "kl": 42.90625, "learning_rate": 1.2769519019112045e-06, "loss": 0.7796, "num_tokens": 370910865.0, "reward": 0.17625001072883606, "reward_std": 0.16985884308815002, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.2156972736120224, "step": 1897 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3346.0, "completions/mean_length": 1364.875, "completions/mean_terminated_length": 571.9677124023438, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.7362296353762606, "grad_norm": 6.878696918487549, "kl": 27.140625, "learning_rate": 1.2748158109749303e-06, "loss": 0.674, "num_tokens": 371076223.0, "reward": 0.23375001549720764, "reward_std": 0.20815400779247284, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.2760366201400757, "step": 1898 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2447.0, "completions/mean_length": 1421.112548828125, "completions/mean_terminated_length": 587.9508056640625, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.7366175329712956, "grad_norm": 2.5526981353759766, "kl": 34.9375, "learning_rate": 1.2726820495558483e-06, "loss": 0.6534, "num_tokens": 371244424.0, "reward": 0.24625001847743988, "reward_std": 0.20536935329437256, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.16750000417232513, "rewards/unicoder_reward_fn/std": 0.2632681429386139, "step": 1899 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3143.0, "completions/mean_length": 1620.0625, "completions/mean_terminated_length": 621.0, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.7370054305663305, "grad_norm": 2.8423221111297607, "kl": 44.75, "learning_rate": 1.2705506210234582e-06, "loss": 0.775, "num_tokens": 371414973.0, "reward": 0.1912500113248825, "reward_std": 0.1481325477361679, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.18195493519306183, "step": 1900 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3329.0, "completions/mean_length": 1627.550048828125, "completions/mean_terminated_length": 569.6428833007812, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.7373933281613654, "grad_norm": 4.159786224365234, "kl": 40.09375, "learning_rate": 1.2684215287435778e-06, "loss": 0.897, "num_tokens": 371587177.0, "reward": 0.1899999976158142, "reward_std": 0.1894853562116623, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.22240060567855835, "step": 1901 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2186.0, "completions/mean_length": 1412.5374755859375, "completions/mean_terminated_length": 633.4677124023438, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.7377812257564003, "grad_norm": 2.5896544456481934, "kl": 36.34375, "learning_rate": 1.2662947760783333e-06, "loss": 0.7479, "num_tokens": 371744380.0, "reward": 0.16250000894069672, "reward_std": 0.13141314685344696, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.18195493519306183, "step": 1902 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3745.0, "completions/mean_length": 1419.6875, "completions/mean_terminated_length": 586.0819091796875, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.7381691233514353, "grad_norm": 2.5518455505371094, "kl": 36.96875, "learning_rate": 1.2641703663861574e-06, "loss": 0.8239, "num_tokens": 371911715.0, "reward": 0.26124998927116394, "reward_std": 0.1912349909543991, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.20950515568256378, "step": 1903 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3069.0, "completions/mean_length": 1199.4749755859375, "completions/mean_terminated_length": 688.3235473632812, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "epoch": 0.7385570209464701, "grad_norm": 2.5841383934020996, "kl": 33.21875, "learning_rate": 1.262048303021784e-06, "loss": 0.6717, "num_tokens": 372054569.0, "reward": 0.26500001549720764, "reward_std": 0.20956817269325256, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.18500001728534698, "rewards/unicoder_reward_fn/std": 0.21054480969905853, "step": 1904 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1272.0, "completions/mean_length": 1688.5250244140625, "completions/mean_terminated_length": 529.370361328125, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.738944918541505, "grad_norm": 7.649024486541748, "kl": 52.40625, "learning_rate": 1.2599285893362389e-06, "loss": 0.8801, "num_tokens": 372229571.0, "reward": 0.20375001430511475, "reward_std": 0.2015257626771927, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.13500002026557922, "rewards/unicoder_reward_fn/std": 0.23977838456630707, "step": 1905 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3550.0, "completions/mean_length": 1669.3499755859375, "completions/mean_terminated_length": 566.3272705078125, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.73933281613654, "grad_norm": 2.637805461883545, "kl": 41.375, "learning_rate": 1.2578112286768417e-06, "loss": 0.8224, "num_tokens": 372410391.0, "reward": 0.1612500101327896, "reward_std": 0.13637521862983704, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.15186603367328644, "step": 1906 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2289.0, "completions/mean_length": 1219.2125244140625, "completions/mean_terminated_length": 608.98486328125, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.7397207137315749, "grad_norm": 2.0837721824645996, "kl": 44.375, "learning_rate": 1.255696224387193e-06, "loss": 0.7875, "num_tokens": 372560072.0, "reward": 0.16750000417232513, "reward_std": 0.12352629005908966, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.15839384496212006, "step": 1907 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2532.0, "completions/mean_length": 933.3624877929688, "completions/mean_terminated_length": 532.4647827148438, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.7401086113266098, "grad_norm": 2.8774936199188232, "kl": 34.390625, "learning_rate": 1.2535835798071716e-06, "loss": 0.6897, "num_tokens": 372684629.0, "reward": 0.32375001907348633, "reward_std": 0.2503390312194824, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.23749999701976776, "rewards/unicoder_reward_fn/std": 0.3354102075099945, "step": 1908 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3949.0, "completions/mean_length": 1531.0, "completions/mean_terminated_length": 618.0338745117188, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.7404965089216446, "grad_norm": 11.061651229858398, "kl": 51.46875, "learning_rate": 1.251473298272934e-06, "loss": 0.8177, "num_tokens": 372852421.0, "reward": 0.21625001728534698, "reward_std": 0.17689602077007294, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.14000001549720764, "rewards/unicoder_reward_fn/std": 0.18390974402427673, "step": 1909 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 2231.0, "completions/mean_length": 1523.0999755859375, "completions/mean_terminated_length": 547.1724243164062, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.7408844065166796, "grad_norm": 11.814313888549805, "kl": 55.4375, "learning_rate": 1.2493653831169015e-06, "loss": 0.8982, "num_tokens": 373020285.0, "reward": 0.16124999523162842, "reward_std": 0.14299774169921875, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.1617741882801056, "step": 1910 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3096.0, "completions/mean_length": 1388.4124755859375, "completions/mean_terminated_length": 602.3386840820312, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.7412723041117145, "grad_norm": 12.251595497131348, "kl": 59.0625, "learning_rate": 1.2472598376677608e-06, "loss": 1.0219, "num_tokens": 373187422.0, "reward": 0.27500003576278687, "reward_std": 0.2328552007675171, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.19750000536441803, "rewards/unicoder_reward_fn/std": 0.28013333678245544, "step": 1911 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 3932.0, "completions/mean_length": 1691.237548828125, "completions/mean_terminated_length": 779.086181640625, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "epoch": 0.7416602017067494, "grad_norm": 9.06608772277832, "kl": 52.3125, "learning_rate": 1.245156665250455e-06, "loss": 0.819, "num_tokens": 373377689.0, "reward": 0.2562499940395355, "reward_std": 0.23201142251491547, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.442768931388855, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.2685050070285797, "step": 1912 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1519.0, "completions/mean_length": 1376.0750732421875, "completions/mean_terminated_length": 528.8851928710938, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.7420480993017843, "grad_norm": 9.156319618225098, "kl": 56.375, "learning_rate": 1.2430558691861799e-06, "loss": 1.079, "num_tokens": 373535559.0, "reward": 0.23749999701976776, "reward_std": 0.14170058071613312, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.1600000113248825, "rewards/unicoder_reward_fn/std": 0.2612530589103699, "step": 1913 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1736.0, "completions/mean_length": 1186.2000732421875, "completions/mean_terminated_length": 514.7077026367188, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.7424359968968193, "grad_norm": 9.825206756591797, "kl": 48.3125, "learning_rate": 1.2409574527923801e-06, "loss": 0.9323, "num_tokens": 373675983.0, "reward": 0.2862500250339508, "reward_std": 0.1879032999277115, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.23480400443077087, "step": 1914 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 1913.0, "completions/mean_length": 1477.5625, "completions/mean_terminated_length": 604.7500610351562, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.7428238944918542, "grad_norm": 9.058696746826172, "kl": 56.5, "learning_rate": 1.2388614193827408e-06, "loss": 0.9963, "num_tokens": 373844748.0, "reward": 0.20375001430511475, "reward_std": 0.18279384076595306, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.2252424955368042, "step": 1915 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2635.0, "completions/mean_length": 1185.0374755859375, "completions/mean_terminated_length": 671.3382568359375, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.743211792086889, "grad_norm": 6.451031684875488, "kl": 32.796875, "learning_rate": 1.2367677722671857e-06, "loss": 0.7121, "num_tokens": 373986727.0, "reward": 0.2224999964237213, "reward_std": 0.15321846306324005, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.19250185787677765, "step": 1916 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1458.0, "completions/mean_length": 1350.0625, "completions/mean_terminated_length": 494.7704772949219, "completions/min_length": 220.0, "completions/min_terminated_length": 220.0, "epoch": 0.743599689681924, "grad_norm": 7.041995525360107, "kl": 50.6875, "learning_rate": 1.234676514751869e-06, "loss": 0.8535, "num_tokens": 374144372.0, "reward": 0.23500001430511475, "reward_std": 0.17556388676166534, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.1600000113248825, "rewards/unicoder_reward_fn/std": 0.21908901631832123, "step": 1917 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3221.0, "completions/mean_length": 1444.9500732421875, "completions/mean_terminated_length": 619.2130737304688, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.7439875872769589, "grad_norm": 2.271876811981201, "kl": 50.71875, "learning_rate": 1.2325876501391708e-06, "loss": 1.0263, "num_tokens": 374312824.0, "reward": 0.30375000834465027, "reward_std": 0.18194864690303802, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.22999998927116394, "rewards/unicoder_reward_fn/std": 0.27020856738090515, "step": 1918 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1495.0, "completions/mean_length": 1077.9375, "completions/mean_terminated_length": 545.3382568359375, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.7443754848719938, "grad_norm": 3.650634288787842, "kl": 28.78125, "learning_rate": 1.230501181727695e-06, "loss": 0.5759, "num_tokens": 374450211.0, "reward": 0.2525000274181366, "reward_std": 0.19470946490764618, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.2214880734682083, "step": 1919 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1400.0, "completions/mean_length": 1282.2750244140625, "completions/mean_terminated_length": 465.3870849609375, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.7447633824670287, "grad_norm": 4.699905872344971, "kl": 40.90625, "learning_rate": 1.2284171128122588e-06, "loss": 0.8773, "num_tokens": 374604385.0, "reward": 0.29749998450279236, "reward_std": 0.24234305322170258, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.2225000113248825, "rewards/unicoder_reward_fn/std": 0.3150145709514618, "step": 1920 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3050.0, "completions/mean_length": 911.0750122070312, "completions/mean_terminated_length": 557.1944580078125, "completions/min_length": 291.0, "completions/min_terminated_length": 291.0, "epoch": 0.7451512800620637, "grad_norm": 3.0231716632843018, "kl": 22.64208984375, "learning_rate": 1.2263354466838928e-06, "loss": 0.5285, "num_tokens": 374722719.0, "reward": 0.3212500512599945, "reward_std": 0.19659966230392456, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.23499998450279236, "rewards/unicoder_reward_fn/std": 0.27700936794281006, "step": 1921 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1776.0, "completions/mean_length": 1449.862548828125, "completions/mean_terminated_length": 508.0169372558594, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.7455391776570985, "grad_norm": 1.927215814590454, "kl": 56.25, "learning_rate": 1.224256186629831e-06, "loss": 1.1265, "num_tokens": 374886444.0, "reward": 0.24250002205371857, "reward_std": 0.24176548421382904, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.27760276198387146, "step": 1922 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1335.0, "completions/mean_length": 1187.2125244140625, "completions/mean_terminated_length": 460.015625, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.7459270752521334, "grad_norm": 2.438863515853882, "kl": 36.96875, "learning_rate": 1.2221793359335083e-06, "loss": 0.6661, "num_tokens": 375032629.0, "reward": 0.1575000137090683, "reward_std": 0.13001814484596252, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.08000000566244125, "rewards/unicoder_reward_fn/std": 0.17888544499874115, "step": 1923 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1906.0, "completions/mean_length": 1217.9000244140625, "completions/mean_terminated_length": 498.375, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.7463149728471683, "grad_norm": 14.807955741882324, "kl": 42.40625, "learning_rate": 1.220104897874557e-06, "loss": 0.9502, "num_tokens": 375175445.0, "reward": 0.28125, "reward_std": 0.17813639342784882, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.20249998569488525, "rewards/unicoder_reward_fn/std": 0.26718172430992126, "step": 1924 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2450.0, "completions/mean_length": 1352.1624755859375, "completions/mean_terminated_length": 555.5645141601562, "completions/min_length": 292.0, "completions/min_terminated_length": 292.0, "epoch": 0.7467028704422033, "grad_norm": 2.3013529777526855, "kl": 45.625, "learning_rate": 1.2180328757287966e-06, "loss": 0.8075, "num_tokens": 375331042.0, "reward": 0.23000001907348633, "reward_std": 0.1781087964773178, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.19420400261878967, "step": 1925 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3690.0, "completions/mean_length": 1324.4625244140625, "completions/mean_terminated_length": 576.5873413085938, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.7470907680372382, "grad_norm": 8.561144828796387, "kl": 46.5625, "learning_rate": 1.2159632727682333e-06, "loss": 0.9937, "num_tokens": 375490887.0, "reward": 0.24250002205371857, "reward_std": 0.15732209384441376, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.16499999165534973, "rewards/unicoder_reward_fn/std": 0.23977839946746826, "step": 1926 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3183.0, "completions/mean_length": 1043.0125732421875, "completions/mean_terminated_length": 504.25, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.747478665632273, "grad_norm": 8.595404624938965, "kl": 31.96875, "learning_rate": 1.2138960922610532e-06, "loss": 0.7421, "num_tokens": 375620896.0, "reward": 0.3787500560283661, "reward_std": 0.22473697364330292, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.29249998927116394, "rewards/unicoder_reward_fn/std": 0.2699156105518341, "step": 1927 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1484.0, "completions/mean_length": 1318.612548828125, "completions/mean_terminated_length": 453.5245666503906, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "epoch": 0.747866563227308, "grad_norm": 2.835367202758789, "kl": 55.4375, "learning_rate": 1.211831337471616e-06, "loss": 1.0882, "num_tokens": 375773769.0, "reward": 0.2525000274181366, "reward_std": 0.1819753348827362, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.2565471827983856, "step": 1928 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1268.0, "completions/mean_length": 919.6625366210938, "completions/mean_terminated_length": 465.8999938964844, "completions/min_length": 254.0, "completions/min_terminated_length": 254.0, "epoch": 0.7482544608223429, "grad_norm": 6.88527774810791, "kl": 34.71875, "learning_rate": 1.2097690116604504e-06, "loss": 0.6221, "num_tokens": 375909206.0, "reward": 0.3525000214576721, "reward_std": 0.22801409661769867, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.26500001549720764, "rewards/unicoder_reward_fn/std": 0.25011390447616577, "step": 1929 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1713.0, "completions/mean_length": 1195.6375732421875, "completions/mean_terminated_length": 470.546875, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.7486423584173778, "grad_norm": 12.8728609085083, "kl": 57.375, "learning_rate": 1.2077091180842524e-06, "loss": 1.1109, "num_tokens": 376056433.0, "reward": 0.21250000596046448, "reward_std": 0.1534619778394699, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.13499999046325684, "rewards/unicoder_reward_fn/std": 0.19297799468040466, "step": 1930 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 910.0, "completions/mean_length": 1468.5625, "completions/mean_terminated_length": 408.3684387207031, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.7490302560124127, "grad_norm": 17.890396118164062, "kl": 69.8125, "learning_rate": 1.205651659995874e-06, "loss": 1.1133, "num_tokens": 376227614.0, "reward": 0.32375001907348633, "reward_std": 0.2812372148036957, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.2550000250339508, "rewards/unicoder_reward_fn/std": 0.3101245164871216, "step": 1931 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2045.0, "completions/mean_length": 1346.1375732421875, "completions/mean_terminated_length": 489.6229248046875, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "epoch": 0.7494181536074477, "grad_norm": 9.691473960876465, "kl": 59.6875, "learning_rate": 1.2035966406443213e-06, "loss": 1.0866, "num_tokens": 376378865.0, "reward": 0.2212500125169754, "reward_std": 0.13947056233882904, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.20119893550872803, "step": 1932 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 1095.0, "completions/mean_length": 1440.125, "completions/mean_terminated_length": 432.72412109375, "completions/min_length": 231.0, "completions/min_terminated_length": 231.0, "epoch": 0.7498060512024826, "grad_norm": 15.596999168395996, "kl": 65.625, "learning_rate": 1.2015440632747526e-06, "loss": 1.176, "num_tokens": 376538147.0, "reward": 0.3162500560283661, "reward_std": 0.23290471732616425, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.24249999225139618, "rewards/unicoder_reward_fn/std": 0.3096649646759033, "step": 1933 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 1935.0, "completions/mean_length": 1384.8499755859375, "completions/mean_terminated_length": 481.13336181640625, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.7501939487975174, "grad_norm": 9.038253784179688, "kl": 61.0, "learning_rate": 1.1994939311284665e-06, "loss": 1.1436, "num_tokens": 376694647.0, "reward": 0.20875000953674316, "reward_std": 0.15626785159111023, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.19382885098457336, "step": 1934 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3471.0, "completions/mean_length": 1559.3499755859375, "completions/mean_terminated_length": 535.7894897460938, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.7505818463925523, "grad_norm": 16.362699508666992, "kl": 66.375, "learning_rate": 1.1974462474429016e-06, "loss": 1.1709, "num_tokens": 376864699.0, "reward": 0.2549999952316284, "reward_std": 0.2190280705690384, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.18500001728534698, "rewards/unicoder_reward_fn/std": 0.26003894209861755, "step": 1935 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 3901.0, "completions/mean_length": 1533.362548828125, "completions/mean_terminated_length": 561.3275756835938, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.7509697439875873, "grad_norm": 22.0006103515625, "kl": 59.1875, "learning_rate": 1.1954010154516293e-06, "loss": 1.1236, "num_tokens": 377044856.0, "reward": 0.21000002324581146, "reward_std": 0.19644257426261902, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.2355172038078308, "step": 1936 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1190.0, "completions/mean_length": 2011.3875732421875, "completions/mean_terminated_length": 470.58697509765625, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.7513576415826222, "grad_norm": 18.617652893066406, "kl": 75.0, "learning_rate": 1.1933582383843514e-06, "loss": 1.3345, "num_tokens": 377252919.0, "reward": 0.19625000655651093, "reward_std": 0.21978233754634857, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.206093892455101, "step": 1937 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3379.0, "completions/mean_length": 2410.112548828125, "completions/mean_terminated_length": 637.7692260742188, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.7517455391776571, "grad_norm": 41.83282470703125, "kl": 70.9375, "learning_rate": 1.19131791946689e-06, "loss": 1.1363, "num_tokens": 377491400.0, "reward": 0.12000000476837158, "reward_std": 0.11897759884595871, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.1550622582435608, "step": 1938 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.675, "completions/max_length": 4096.0, "completions/max_terminated_length": 3029.0, "completions/mean_length": 3022.5625, "completions/mean_terminated_length": 793.1154174804688, "completions/min_length": 76.0, "completions/min_terminated_length": 76.0, "epoch": 0.752133436772692, "grad_norm": 74.84978485107422, "kl": 61.875, "learning_rate": 1.1892800619211894e-06, "loss": 0.7447, "num_tokens": 377783573.0, "reward": 0.07500000298023224, "reward_std": 0.0806359276175499, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.02250000089406967, "rewards/unicoder_reward_fn/std": 0.0899718776345253, "step": 1939 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 2601.0, "completions/mean_length": 2347.449951171875, "completions/mean_terminated_length": 598.9000244140625, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.752521334367727, "grad_norm": 34.29100799560547, "kl": 50.625, "learning_rate": 1.1872446689653034e-06, "loss": 0.8064, "num_tokens": 378014417.0, "reward": 0.1262500137090683, "reward_std": 0.13580335676670074, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.15652476251125336, "step": 1940 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 2491.0, "completions/mean_length": 2345.962646484375, "completions/mean_terminated_length": 595.9249877929688, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.7529092319627618, "grad_norm": 8.536192893981934, "kl": 49.5625, "learning_rate": 1.185211743813395e-06, "loss": 0.9036, "num_tokens": 378249886.0, "reward": 0.17499999701976776, "reward_std": 0.16807661950588226, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.19051894545555115, "step": 1941 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 827.0, "completions/mean_length": 1718.4124755859375, "completions/mean_terminated_length": 438.173095703125, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.7532971295577967, "grad_norm": 37.37450408935547, "kl": 41.125, "learning_rate": 1.1831812896757313e-06, "loss": 1.0567, "num_tokens": 378432199.0, "reward": 0.288750022649765, "reward_std": 0.2609184682369232, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.2199999988079071, "rewards/unicoder_reward_fn/std": 0.2537815272808075, "step": 1942 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1422.0, "completions/mean_length": 1501.737548828125, "completions/mean_terminated_length": 454.9298400878906, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.7536850271528317, "grad_norm": 13.147748947143555, "kl": 32.578125, "learning_rate": 1.181153309758676e-06, "loss": 0.769, "num_tokens": 378600074.0, "reward": 0.2237500250339508, "reward_std": 0.20237798988819122, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.2787744402885437, "step": 1943 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1005.0, "completions/mean_length": 1850.4375, "completions/mean_terminated_length": 503.0999755859375, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.7540729247478666, "grad_norm": 19.406112670898438, "kl": 44.6875, "learning_rate": 1.179127807264687e-06, "loss": 1.1027, "num_tokens": 378815725.0, "reward": 0.22874999046325684, "reward_std": 0.20846009254455566, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.21452036499977112, "step": 1944 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 3924.0, "completions/mean_length": 1427.862548828125, "completions/mean_terminated_length": 538.4833374023438, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.7544608223429015, "grad_norm": 14.536700248718262, "kl": 32.4375, "learning_rate": 1.1771047853923086e-06, "loss": 0.736, "num_tokens": 378978058.0, "reward": 0.2474999874830246, "reward_std": 0.1999010592699051, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.17250001430511475, "rewards/unicoder_reward_fn/std": 0.26191845536231995, "step": 1945 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 1091.0, "completions/mean_length": 1457.9500732421875, "completions/mean_terminated_length": 457.3103332519531, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.7548487199379363, "grad_norm": 24.945775985717773, "kl": 35.90625, "learning_rate": 1.175084247336168e-06, "loss": 0.9634, "num_tokens": 379141750.0, "reward": 0.36250001192092896, "reward_std": 0.2734956741333008, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.291384756565094, "step": 1946 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1621.0, "completions/mean_length": 1428.550048828125, "completions/mean_terminated_length": 479.11865234375, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.7552366175329713, "grad_norm": 22.675901412963867, "kl": 32.5625, "learning_rate": 1.1730661962869718e-06, "loss": 0.7579, "num_tokens": 379310082.0, "reward": 0.33125001192092896, "reward_std": 0.22916793823242188, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.442768931388855, "rewards/unicoder_reward_fn/mean": 0.2574999928474426, "rewards/unicoder_reward_fn/std": 0.27038416266441345, "step": 1947 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2131.0, "completions/mean_length": 1697.0875244140625, "completions/mean_terminated_length": 475.0, "completions/min_length": 118.0, "completions/min_terminated_length": 118.0, "epoch": 0.7556245151280062, "grad_norm": 10.839215278625488, "kl": 41.0625, "learning_rate": 1.1710506354314967e-06, "loss": 0.8698, "num_tokens": 379490969.0, "reward": 0.17375001311302185, "reward_std": 0.1665140986442566, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.2181481420993805, "step": 1948 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1877.0, "completions/mean_length": 1624.0374755859375, "completions/mean_terminated_length": 433.8333435058594, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.7560124127230411, "grad_norm": 9.780569076538086, "kl": 51.125, "learning_rate": 1.1690375679525897e-06, "loss": 1.1259, "num_tokens": 379677460.0, "reward": 0.1875000149011612, "reward_std": 0.17017744481563568, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.18471670150756836, "step": 1949 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1205.0, "completions/mean_length": 1153.9625244140625, "completions/mean_terminated_length": 475.03076171875, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.756400310318076, "grad_norm": 13.62718677520752, "kl": 30.75, "learning_rate": 1.1670269970291585e-06, "loss": 0.6714, "num_tokens": 379821537.0, "reward": 0.2487500160932541, "reward_std": 0.1517448127269745, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.16499999165534973, "rewards/unicoder_reward_fn/std": 0.2355172038078308, "step": 1950 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2763.0, "completions/mean_length": 1953.9500732421875, "completions/mean_terminated_length": 449.9574279785156, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.756788207913111, "grad_norm": 7.711963653564453, "kl": 60.125, "learning_rate": 1.1650189258361685e-06, "loss": 1.1981, "num_tokens": 380018405.0, "reward": 0.21500001847743988, "reward_std": 0.19442477822303772, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.2580231726169586, "step": 1951 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2034.0, "completions/mean_length": 1812.8250732421875, "completions/mean_terminated_length": 514.5490112304688, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.7571761055081458, "grad_norm": 9.462559700012207, "kl": 59.25, "learning_rate": 1.1630133575446394e-06, "loss": 1.045, "num_tokens": 380209039.0, "reward": 0.19750002026557922, "reward_std": 0.22821874916553497, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.2510610520839691, "step": 1952 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 1960.0, "completions/mean_length": 1526.175048828125, "completions/mean_terminated_length": 551.413818359375, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.7575640031031807, "grad_norm": 6.738129615783691, "kl": 53.0625, "learning_rate": 1.1610102953216361e-06, "loss": 1.0751, "num_tokens": 380387781.0, "reward": 0.20624999701976776, "reward_std": 0.16823284327983856, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.22556538879871368, "step": 1953 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3468.0, "completions/mean_length": 2006.8499755859375, "completions/mean_terminated_length": 685.142822265625, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "epoch": 0.7579519006982157, "grad_norm": 15.296514511108398, "kl": 62.25, "learning_rate": 1.1590097423302683e-06, "loss": 1.1622, "num_tokens": 380596065.0, "reward": 0.08125000447034836, "reward_std": 0.07192482054233551, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.08282970637083054, "step": 1954 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1794.0, "completions/mean_length": 1939.0750732421875, "completions/mean_terminated_length": 574.4898071289062, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.7583397982932506, "grad_norm": 18.21587562561035, "kl": 64.9375, "learning_rate": 1.157011701729682e-06, "loss": 1.0852, "num_tokens": 380805663.0, "reward": 0.1587499976158142, "reward_std": 0.18259000778198242, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.19858677685260773, "step": 1955 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3728.0, "completions/mean_length": 1820.487548828125, "completions/mean_terminated_length": 661.26416015625, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.7587276958882855, "grad_norm": 13.02293872833252, "kl": 56.375, "learning_rate": 1.155016176675055e-06, "loss": 0.9152, "num_tokens": 380999182.0, "reward": 0.11250000447034836, "reward_std": 0.13219814002513885, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.042500000447034836, "rewards/unicoder_reward_fn/std": 0.13758771121501923, "step": 1956 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2027.0, "completions/mean_length": 1674.8499755859375, "completions/mean_terminated_length": 441.4339599609375, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.7591155934833204, "grad_norm": 13.061161041259766, "kl": 63.3125, "learning_rate": 1.153023170317595e-06, "loss": 1.0707, "num_tokens": 381167522.0, "reward": 0.20000003278255463, "reward_std": 0.2031344622373581, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.26422804594039917, "step": 1957 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2421.0, "completions/mean_length": 1327.4124755859375, "completions/mean_terminated_length": 523.6290283203125, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.7595034910783554, "grad_norm": 7.3155293464660645, "kl": 42.46875, "learning_rate": 1.15103268580453e-06, "loss": 0.781, "num_tokens": 381319547.0, "reward": 0.24375000596046448, "reward_std": 0.21978704631328583, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.22968056797981262, "step": 1958 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2473.0, "completions/mean_length": 1287.112548828125, "completions/mean_terminated_length": 584.890625, "completions/min_length": 123.0, "completions/min_terminated_length": 123.0, "epoch": 0.7598913886733902, "grad_norm": 10.637551307678223, "kl": 35.6875, "learning_rate": 1.1490447262791081e-06, "loss": 0.7271, "num_tokens": 381459900.0, "reward": 0.23250000178813934, "reward_std": 0.20092558860778809, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.2428600788116455, "step": 1959 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2410.0, "completions/mean_length": 1548.1375732421875, "completions/mean_terminated_length": 520.0526123046875, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.7602792862684251, "grad_norm": 5.826297760009766, "kl": 45.0625, "learning_rate": 1.1470592948805883e-06, "loss": 0.9095, "num_tokens": 381623415.0, "reward": 0.14124999940395355, "reward_std": 0.13276341557502747, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.0650000050663948, "rewards/unicoder_reward_fn/std": 0.1263909935951233, "step": 1960 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2922.0, "completions/mean_length": 1976.8375244140625, "completions/mean_terminated_length": 488.9148864746094, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.76066718386346, "grad_norm": 7.434117317199707, "kl": 49.4375, "learning_rate": 1.1450763947442362e-06, "loss": 0.9223, "num_tokens": 381827650.0, "reward": 0.14750002324581146, "reward_std": 0.1573801189661026, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.23514068126678467, "step": 1961 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2938.0, "completions/mean_length": 2125.987548828125, "completions/mean_terminated_length": 593.7555541992188, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.761055081458495, "grad_norm": 14.914464950561523, "kl": 59.1875, "learning_rate": 1.143096029001324e-06, "loss": 1.0214, "num_tokens": 382044433.0, "reward": 0.10375000536441803, "reward_std": 0.11805739253759384, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.1207571029663086, "step": 1962 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 2286.0, "completions/mean_length": 1793.5, "completions/mean_terminated_length": 553.6923217773438, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.7614429790535299, "grad_norm": 13.272429466247559, "kl": 47.5625, "learning_rate": 1.1411182007791178e-06, "loss": 0.8536, "num_tokens": 382242849.0, "reward": 0.1612500250339508, "reward_std": 0.17930567264556885, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.22897681593894958, "step": 1963 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2269.0, "completions/mean_length": 1821.675048828125, "completions/mean_terminated_length": 457.0799865722656, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.7618308766485647, "grad_norm": 4.436610221862793, "kl": 50.6875, "learning_rate": 1.13914291320088e-06, "loss": 0.9843, "num_tokens": 382440375.0, "reward": 0.17375001311302185, "reward_std": 0.1838536560535431, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.10249999910593033, "rewards/unicoder_reward_fn/std": 0.1855372041463852, "step": 1964 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3715.0, "completions/mean_length": 1716.2750244140625, "completions/mean_terminated_length": 634.581787109375, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.7622187742435997, "grad_norm": 8.858098030090332, "kl": 42.8125, "learning_rate": 1.1371701693858584e-06, "loss": 0.96, "num_tokens": 382629437.0, "reward": 0.15250001847743988, "reward_std": 0.15166909992694855, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.14405958354473114, "step": 1965 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2124.0, "completions/mean_length": 1345.4749755859375, "completions/mean_terminated_length": 488.7540588378906, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.7626066718386346, "grad_norm": 14.759074211120605, "kl": 36.03125, "learning_rate": 1.135199972449284e-06, "loss": 0.8318, "num_tokens": 382786211.0, "reward": 0.14124999940395355, "reward_std": 0.1420975923538208, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.1951143890619278, "step": 1966 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1399.0, "completions/mean_length": 1339.4124755859375, "completions/mean_terminated_length": 480.8032531738281, "completions/min_length": 123.0, "completions/min_terminated_length": 123.0, "epoch": 0.7629945694336695, "grad_norm": 5.4388651847839355, "kl": 36.28125, "learning_rate": 1.1332323255023677e-06, "loss": 0.7322, "num_tokens": 382935716.0, "reward": 0.19875000417232513, "reward_std": 0.1970692276954651, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.24151422083377838, "step": 1967 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3127.0, "completions/mean_length": 1383.4625244140625, "completions/mean_terminated_length": 538.57373046875, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.7633824670287044, "grad_norm": 11.67384147644043, "kl": 33.359375, "learning_rate": 1.1312672316522923e-06, "loss": 0.7742, "num_tokens": 383088457.0, "reward": 0.2537500262260437, "reward_std": 0.17228369414806366, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.17249999940395355, "rewards/unicoder_reward_fn/std": 0.26191845536231995, "step": 1968 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2335.0, "completions/mean_length": 1211.550048828125, "completions/mean_terminated_length": 490.4375, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.7637703646237394, "grad_norm": 17.33757781982422, "kl": 30.1875, "learning_rate": 1.1293046940022095e-06, "loss": 0.7032, "num_tokens": 383229981.0, "reward": 0.29124999046325684, "reward_std": 0.22772741317749023, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.32676976919174194, "step": 1969 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2172.0, "completions/mean_length": 1924.1500244140625, "completions/mean_terminated_length": 621.0399780273438, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.7641582622187743, "grad_norm": 8.14676284790039, "kl": 52.0, "learning_rate": 1.127344715651234e-06, "loss": 0.906, "num_tokens": 383426961.0, "reward": 0.17874999344348907, "reward_std": 0.14296236634254456, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.2314511090517044, "step": 1970 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2909.0, "completions/mean_length": 1296.6624755859375, "completions/mean_terminated_length": 541.2857666015625, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.7645461598138091, "grad_norm": 4.003472328186035, "kl": 35.9375, "learning_rate": 1.1253872996944385e-06, "loss": 0.6953, "num_tokens": 383574726.0, "reward": 0.26875001192092896, "reward_std": 0.1520296186208725, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.1900000125169754, "rewards/unicoder_reward_fn/std": 0.2442244440317154, "step": 1971 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2306.0, "completions/mean_length": 1017.75, "completions/mean_terminated_length": 474.5294189453125, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.764934057408844, "grad_norm": 5.763489723205566, "kl": 44.0, "learning_rate": 1.1234324492228512e-06, "loss": 0.9043, "num_tokens": 383693690.0, "reward": 0.2862500250339508, "reward_std": 0.2190176099538803, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.2652798891067505, "step": 1972 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1464.0, "completions/mean_length": 1073.4000244140625, "completions/mean_terminated_length": 486.92535400390625, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "epoch": 0.765321955003879, "grad_norm": 3.5196895599365234, "kl": 39.9375, "learning_rate": 1.1214801673234462e-06, "loss": 0.7947, "num_tokens": 383827634.0, "reward": 0.23499999940395355, "reward_std": 0.1719425469636917, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.23652280867099762, "step": 1973 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1286.0, "completions/mean_length": 1269.0125732421875, "completions/mean_terminated_length": 448.274169921875, "completions/min_length": 135.0, "completions/min_terminated_length": 135.0, "epoch": 0.7657098525989139, "grad_norm": 2.2341458797454834, "kl": 44.65625, "learning_rate": 1.1195304570791452e-06, "loss": 0.8967, "num_tokens": 383982067.0, "reward": 0.21125002205371857, "reward_std": 0.1576838344335556, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.22747932374477386, "step": 1974 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2550.0, "completions/mean_length": 1549.875, "completions/mean_terminated_length": 522.4912109375, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.7660977501939488, "grad_norm": 5.2665252685546875, "kl": 40.75, "learning_rate": 1.1175833215688051e-06, "loss": 0.755, "num_tokens": 384153905.0, "reward": 0.10000000149011612, "reward_std": 0.08611989766359329, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.02250000089406967, "rewards/unicoder_reward_fn/std": 0.0899718627333641, "step": 1975 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4061.0, "completions/mean_length": 1807.1500244140625, "completions/mean_terminated_length": 766.7636108398438, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.7664856477889838, "grad_norm": 8.373736381530762, "kl": 46.125, "learning_rate": 1.1156387638672195e-06, "loss": 0.8342, "num_tokens": 384341765.0, "reward": 0.1274999976158142, "reward_std": 0.12091951817274094, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.126265749335289, "step": 1976 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3856.0, "completions/mean_length": 2209.487548828125, "completions/mean_terminated_length": 742.2000122070312, "completions/min_length": 101.0, "completions/min_terminated_length": 101.0, "epoch": 0.7668735453840186, "grad_norm": 27.810651779174805, "kl": 57.875, "learning_rate": 1.1136967870451095e-06, "loss": 0.9317, "num_tokens": 384553276.0, "reward": 0.07125000655651093, "reward_std": 0.051135431975126266, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.004999999888241291, "rewards/unicoder_reward_fn/std": 0.03142199665307999, "step": 1977 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 3269.0, "completions/mean_length": 2408.900146484375, "completions/mean_terminated_length": 882.4761962890625, "completions/min_length": 169.0, "completions/min_terminated_length": 169.0, "epoch": 0.7672614429790535, "grad_norm": 14.77841567993164, "kl": 39.5625, "learning_rate": 1.1117573941691225e-06, "loss": 0.6461, "num_tokens": 384795828.0, "reward": 0.10750000923871994, "reward_std": 0.10235798358917236, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.030000001192092896, "rewards/unicoder_reward_fn/std": 0.13538506627082825, "step": 1978 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3475.0, "completions/mean_length": 2117.4375, "completions/mean_terminated_length": 930.2999877929688, "completions/min_length": 218.0, "completions/min_terminated_length": 218.0, "epoch": 0.7676493405740884, "grad_norm": 15.41100788116455, "kl": 50.875, "learning_rate": 1.1098205883018246e-06, "loss": 0.8416, "num_tokens": 385010895.0, "reward": 0.09375, "reward_std": 0.10605964809656143, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.12065224349498749, "step": 1979 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 2210.0, "completions/mean_length": 1820.9749755859375, "completions/mean_terminated_length": 595.9615478515625, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.7680372381691234, "grad_norm": 11.267465591430664, "kl": 53.625, "learning_rate": 1.1078863725016956e-06, "loss": 0.8936, "num_tokens": 385199037.0, "reward": 0.19875000417232513, "reward_std": 0.20987391471862793, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.1274999976158142, "rewards/unicoder_reward_fn/std": 0.25506266951560974, "step": 1980 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3723.0, "completions/mean_length": 1704.9124755859375, "completions/mean_terminated_length": 680.1607666015625, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.7684251357641583, "grad_norm": 13.307480812072754, "kl": 44.40625, "learning_rate": 1.1059547498231277e-06, "loss": 0.8392, "num_tokens": 385388342.0, "reward": 0.09250000864267349, "reward_std": 0.05049135908484459, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.05301063135266304, "step": 1981 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2469.0, "completions/mean_length": 1191.4500732421875, "completions/mean_terminated_length": 521.1692504882812, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.7688130333591932, "grad_norm": 5.558623790740967, "kl": 39.3125, "learning_rate": 1.104025723316416e-06, "loss": 0.7759, "num_tokens": 385526058.0, "reward": 0.3187500238418579, "reward_std": 0.28945860266685486, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.24000000953674316, "rewards/unicoder_reward_fn/std": 0.33284351229667664, "step": 1982 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2075.0, "completions/mean_length": 809.375, "completions/mean_terminated_length": 444.1944580078125, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.769200930954228, "grad_norm": 14.541690826416016, "kl": 20.83251953125, "learning_rate": 1.1020992960277558e-06, "loss": 0.5942, "num_tokens": 385631176.0, "reward": 0.26625004410743713, "reward_std": 0.17412471771240234, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.23893223702907562, "step": 1983 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1191.0, "completions/mean_length": 1075.8875732421875, "completions/mean_terminated_length": 489.8955078125, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "epoch": 0.769588828549263, "grad_norm": 10.135055541992188, "kl": 29.6015625, "learning_rate": 1.1001754709992403e-06, "loss": 0.6907, "num_tokens": 385765343.0, "reward": 0.15250001847743988, "reward_std": 0.11357302963733673, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.06750001013278961, "rewards/unicoder_reward_fn/std": 0.18264926970005035, "step": 1984 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2638.0, "completions/mean_length": 932.2250366210938, "completions/mean_terminated_length": 580.6944580078125, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.7699767261442979, "grad_norm": 4.0042877197265625, "kl": 23.671875, "learning_rate": 1.0982542512688515e-06, "loss": 0.5141, "num_tokens": 385885673.0, "reward": 0.2537499964237213, "reward_std": 0.19482402503490448, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.22897681593894958, "step": 1985 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1479.0, "completions/mean_length": 919.1875, "completions/mean_terminated_length": 516.492919921875, "completions/min_length": 210.0, "completions/min_terminated_length": 210.0, "epoch": 0.7703646237393328, "grad_norm": 4.889166355133057, "kl": 25.142578125, "learning_rate": 1.0963356398704564e-06, "loss": 0.5754, "num_tokens": 386004144.0, "reward": 0.24249999225139618, "reward_std": 0.17995135486125946, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.19610124826431274, "step": 1986 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2011.0, "completions/mean_length": 908.3500366210938, "completions/mean_terminated_length": 504.28167724609375, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "epoch": 0.7707525213343678, "grad_norm": 10.121176719665527, "kl": 17.66015625, "learning_rate": 1.094419639833806e-06, "loss": 0.3967, "num_tokens": 386126476.0, "reward": 0.29250001907348633, "reward_std": 0.17737238109111786, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.2505689859390259, "step": 1987 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2877.0, "completions/mean_length": 1033.2874755859375, "completions/mean_terminated_length": 645.0563354492188, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.7711404189294027, "grad_norm": 6.587246894836426, "kl": 24.546875, "learning_rate": 1.0925062541845246e-06, "loss": 0.5015, "num_tokens": 386260635.0, "reward": 0.21875, "reward_std": 0.18278062343597412, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.23834875226020813, "step": 1988 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1806.0, "completions/mean_length": 1468.0250244140625, "completions/mean_terminated_length": 532.64404296875, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.7715283165244375, "grad_norm": 5.825620651245117, "kl": 51.25, "learning_rate": 1.0905954859441089e-06, "loss": 0.9744, "num_tokens": 386433469.0, "reward": 0.19500000774860382, "reward_std": 0.17912128567695618, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.22411571443080902, "step": 1989 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3320.0, "completions/mean_length": 1737.7750244140625, "completions/mean_terminated_length": 665.8545532226562, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.7719162141194724, "grad_norm": 8.24437427520752, "kl": 53.0, "learning_rate": 1.0886873381299248e-06, "loss": 0.9449, "num_tokens": 386619379.0, "reward": 0.1612500101327896, "reward_std": 0.1847124844789505, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.22897681593894958, "step": 1990 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 3707.0, "completions/mean_length": 1614.675048828125, "completions/mean_terminated_length": 787.5667114257812, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.7723041117145074, "grad_norm": 4.900613784790039, "kl": 43.875, "learning_rate": 1.086781813755196e-06, "loss": 0.8415, "num_tokens": 386785185.0, "reward": 0.21250000596046448, "reward_std": 0.2312556356191635, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.2680331766605377, "step": 1991 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3382.0, "completions/mean_length": 2402.087646484375, "completions/mean_terminated_length": 1016.1591186523438, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.7726920093095423, "grad_norm": 15.019638061523438, "kl": 48.375, "learning_rate": 1.0848789158290075e-06, "loss": 0.7961, "num_tokens": 387033544.0, "reward": 0.11000000685453415, "reward_std": 0.10041846334934235, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.11810743808746338, "step": 1992 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 2353.0, "completions/mean_length": 2710.125, "completions/mean_terminated_length": 835.11767578125, "completions/min_length": 78.0, "completions/min_terminated_length": 78.0, "epoch": 0.7730799069045772, "grad_norm": 21.20973014831543, "kl": 50.0625, "learning_rate": 1.0829786473562934e-06, "loss": 0.7457, "num_tokens": 387304810.0, "reward": 0.10750000923871994, "reward_std": 0.11770918220281601, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.03500000014901161, "rewards/unicoder_reward_fn/std": 0.1303354948759079, "step": 1993 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2112.0, "completions/mean_length": 3069.85009765625, "completions/mean_terminated_length": 812.3200073242188, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.773467804499612, "grad_norm": 24.61343765258789, "kl": 49.8125, "learning_rate": 1.0810810113378373e-06, "loss": 0.7217, "num_tokens": 387595822.0, "reward": 0.07625000923871994, "reward_std": 0.059799302369356155, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.04386400431394577, "step": 1994 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 4039.0, "completions/mean_length": 3045.0625, "completions/mean_terminated_length": 982.1111450195312, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.773855702094647, "grad_norm": 22.30769920349121, "kl": 40.0625, "learning_rate": 1.0791860107702658e-06, "loss": 0.5737, "num_tokens": 387883595.0, "reward": 0.10500001162290573, "reward_std": 0.10033565014600754, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.0325000025331974, "rewards/unicoder_reward_fn/std": 0.13290807604789734, "step": 1995 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 3646.0, "completions/mean_length": 2088.425048828125, "completions/mean_terminated_length": 750.0416870117188, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.7742435996896819, "grad_norm": 5.693391799926758, "kl": 52.625, "learning_rate": 1.0772936486460431e-06, "loss": 0.9761, "num_tokens": 388099525.0, "reward": 0.14374999701976776, "reward_std": 0.16008834540843964, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.442768931388855, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.15294575691223145, "step": 1996 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 3507.0, "completions/mean_length": 2020.5250244140625, "completions/mean_terminated_length": 636.875, "completions/min_length": 131.0, "completions/min_terminated_length": 131.0, "epoch": 0.7746314972847168, "grad_norm": 5.13550329208374, "kl": 43.25, "learning_rate": 1.075403927953468e-06, "loss": 0.8077, "num_tokens": 388308303.0, "reward": 0.1875000149011612, "reward_std": 0.22064383327960968, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.2659471035003662, "step": 1997 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2999.0, "completions/mean_length": 1908.1500244140625, "completions/mean_terminated_length": 664.0784301757812, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.7750193948797518, "grad_norm": 8.917434692382812, "kl": 30.6875, "learning_rate": 1.073516851676667e-06, "loss": 0.6589, "num_tokens": 388508835.0, "reward": 0.18250000476837158, "reward_std": 0.15702790021896362, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.20124612748622894, "step": 1998 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2618.0, "completions/mean_length": 2299.16259765625, "completions/mean_terminated_length": 753.0465087890625, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "epoch": 0.7754072924747867, "grad_norm": 4.263120174407959, "kl": 36.4375, "learning_rate": 1.0716324227955904e-06, "loss": 0.7689, "num_tokens": 388734592.0, "reward": 0.13125000894069672, "reward_std": 0.12415174394845963, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.14917494356632233, "step": 1999 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 2762.0, "completions/mean_length": 2275.337646484375, "completions/mean_terminated_length": 628.0714111328125, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.7757951900698216, "grad_norm": 7.683692932128906, "kl": 31.3125, "learning_rate": 1.06975064428601e-06, "loss": 0.622, "num_tokens": 388965419.0, "reward": 0.14000001549720764, "reward_std": 0.12250363826751709, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.1671617180109024, "step": 2000 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 4092.0, "completions/mean_length": 2528.0, "completions/mean_terminated_length": 794.9473876953125, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "epoch": 0.7761830876648564, "grad_norm": 8.878157615661621, "kl": 29.15625, "learning_rate": 1.0678715191195095e-06, "loss": 0.5345, "num_tokens": 389215691.0, "reward": 0.12000000476837158, "reward_std": 0.1042691245675087, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.04500000178813934, "rewards/unicoder_reward_fn/std": 0.165276899933815, "step": 2001 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 3091.0, "completions/mean_length": 2386.9375, "completions/mean_terminated_length": 677.875, "completions/min_length": 279.0, "completions/min_terminated_length": 279.0, "epoch": 0.7765709852598914, "grad_norm": 6.358726978302002, "kl": 41.40625, "learning_rate": 1.0659950502634854e-06, "loss": 0.7997, "num_tokens": 389461102.0, "reward": 0.1275000125169754, "reward_std": 0.1207321435213089, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.15325580537319183, "step": 2002 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2676.0, "completions/mean_length": 1757.6375732421875, "completions/mean_terminated_length": 631.75927734375, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.7769588828549263, "grad_norm": 10.19265079498291, "kl": 35.65625, "learning_rate": 1.0641212406811374e-06, "loss": 0.7353, "num_tokens": 389652689.0, "reward": 0.1575000137090683, "reward_std": 0.1339363157749176, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.18718433380126953, "step": 2003 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 2950.0, "completions/mean_length": 1635.6500244140625, "completions/mean_terminated_length": 702.413818359375, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.7773467804499612, "grad_norm": 8.927242279052734, "kl": 31.90625, "learning_rate": 1.0622500933314661e-06, "loss": 0.7318, "num_tokens": 389834069.0, "reward": 0.17375001311302185, "reward_std": 0.16032753884792328, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.19342026114463806, "step": 2004 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2537.0, "completions/mean_length": 1239.800048828125, "completions/mean_terminated_length": 469.07940673828125, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.7777346780449961, "grad_norm": 6.205746650695801, "kl": 27.625, "learning_rate": 1.06038161116927e-06, "loss": 0.5488, "num_tokens": 389980469.0, "reward": 0.23125000298023224, "reward_std": 0.17721174657344818, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2249613255262375, "step": 2005 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3155.0, "completions/mean_length": 1054.0, "completions/mean_terminated_length": 463.76116943359375, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.7781225756400311, "grad_norm": 19.063207626342773, "kl": 33.25, "learning_rate": 1.0585157971451356e-06, "loss": 0.8608, "num_tokens": 390103333.0, "reward": 0.2175000160932541, "reward_std": 0.15997831523418427, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.13249999284744263, "rewards/unicoder_reward_fn/std": 0.18811187148094177, "step": 2006 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2960.0, "completions/mean_length": 1537.2750244140625, "completions/mean_terminated_length": 504.8070068359375, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.778510473235066, "grad_norm": 11.633496284484863, "kl": 37.96875, "learning_rate": 1.05665265420544e-06, "loss": 0.8465, "num_tokens": 390281123.0, "reward": 0.22625000774860382, "reward_std": 0.21565549075603485, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.25802314281463623, "step": 2007 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2569.0, "completions/mean_length": 1412.925048828125, "completions/mean_terminated_length": 633.9677124023438, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.7788983708301008, "grad_norm": 18.202674865722656, "kl": 24.40625, "learning_rate": 1.054792185292339e-06, "loss": 0.6369, "num_tokens": 390439053.0, "reward": 0.28125, "reward_std": 0.22256170213222504, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.1925000101327896, "rewards/unicoder_reward_fn/std": 0.25743377208709717, "step": 2008 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.23750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3056.0, "completions/mean_length": 1389.800048828125, "completions/mean_terminated_length": 546.8851928710938, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.7792862684251358, "grad_norm": 3.333266019821167, "kl": 31.96875, "learning_rate": 1.0529343933437663e-06, "loss": 0.5157, "num_tokens": 390595045.0, "reward": 0.2537500262260437, "reward_std": 0.19152316451072693, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.27020853757858276, "step": 2009 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2107.0, "completions/mean_length": 1788.3375244140625, "completions/mean_terminated_length": 612.73583984375, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.7796741660201707, "grad_norm": 3.1909940242767334, "kl": 49.125, "learning_rate": 1.0510792812934307e-06, "loss": 0.9734, "num_tokens": 390793016.0, "reward": 0.19500000774860382, "reward_std": 0.15449105203151703, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.11749999225139618, "rewards/unicoder_reward_fn/std": 0.23318108916282654, "step": 2010 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3095.0, "completions/mean_length": 1759.6375732421875, "completions/mean_terminated_length": 697.654541015625, "completions/min_length": 44.0, "completions/min_terminated_length": 44.0, "epoch": 0.7800620636152056, "grad_norm": 9.850004196166992, "kl": 33.296875, "learning_rate": 1.0492268520708056e-06, "loss": 0.8152, "num_tokens": 390984883.0, "reward": 0.1612500101327896, "reward_std": 0.1453971415758133, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.16379596292972565, "step": 2011 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2747.0, "completions/mean_length": 1899.487548828125, "completions/mean_terminated_length": 581.5800170898438, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.7804499612102405, "grad_norm": 9.297029495239258, "kl": 28.21875, "learning_rate": 1.0473771086011313e-06, "loss": 0.6456, "num_tokens": 391187658.0, "reward": 0.15000000596046448, "reward_std": 0.1131720095872879, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.0650000050663948, "rewards/unicoder_reward_fn/std": 0.15839384496212006, "step": 2012 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 4057.0, "completions/mean_length": 2414.212646484375, "completions/mean_terminated_length": 732.4249877929688, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "epoch": 0.7808378588052755, "grad_norm": 15.058629989624023, "kl": 43.09375, "learning_rate": 1.0455300538054045e-06, "loss": 0.6958, "num_tokens": 391435899.0, "reward": 0.10000000149011612, "reward_std": 0.08819307386875153, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.10732655227184296, "step": 2013 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3983.0, "completions/mean_length": 2848.03759765625, "completions/mean_terminated_length": 1070.6363525390625, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.7812257564003103, "grad_norm": 19.9174861907959, "kl": 59.9375, "learning_rate": 1.043685690600376e-06, "loss": 0.9406, "num_tokens": 391709358.0, "reward": 0.10250001400709152, "reward_std": 0.12255900353193283, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.14264677464962006, "step": 2014 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3667.0, "completions/mean_length": 3132.60009765625, "completions/mean_terminated_length": 1013.1199951171875, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.7816136539953452, "grad_norm": 33.34521484375, "kl": 55.25, "learning_rate": 1.0418440218985483e-06, "loss": 0.7794, "num_tokens": 392005422.0, "reward": 0.07125000655651093, "reward_std": 0.05146493390202522, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.0024999999441206455, "rewards/unicoder_reward_fn/std": 0.022360680624842644, "step": 2015 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2932.0, "completions/mean_length": 3006.675048828125, "completions/mean_terminated_length": 868.370361328125, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.7820015515903801, "grad_norm": 24.38065528869629, "kl": 62.375, "learning_rate": 1.040005050608166e-06, "loss": 0.9042, "num_tokens": 392305292.0, "reward": 0.08375000208616257, "reward_std": 0.10285482555627823, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.11194483935832977, "step": 2016 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3430.0, "completions/mean_length": 3055.22509765625, "completions/mean_terminated_length": 765.5199584960938, "completions/min_length": 150.0, "completions/min_terminated_length": 150.0, "epoch": 0.7823894491854151, "grad_norm": 21.216907501220703, "kl": 55.1875, "learning_rate": 1.038168779633218e-06, "loss": 0.8322, "num_tokens": 392601214.0, "reward": 0.09750000387430191, "reward_std": 0.12408428639173508, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298882484436035, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.14615321159362793, "step": 2017 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 3430.0, "completions/mean_length": 2913.425048828125, "completions/mean_terminated_length": 1139.5625, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.78277734678045, "grad_norm": 24.557302474975586, "kl": 44.3125, "learning_rate": 1.0363352118734257e-06, "loss": 0.6091, "num_tokens": 392888328.0, "reward": 0.07750000059604645, "reward_std": 0.06000840663909912, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.007500000298023224, "rewards/unicoder_reward_fn/std": 0.04974619299173355, "step": 2018 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2072.0, "completions/mean_length": 2053.78759765625, "completions/mean_terminated_length": 544.3261108398438, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.7831652443754848, "grad_norm": 9.174755096435547, "kl": 48.375, "learning_rate": 1.0345043502242421e-06, "loss": 0.8594, "num_tokens": 393098735.0, "reward": 0.17000000178813934, "reward_std": 0.18987850844860077, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.0949999988079071, "rewards/unicoder_reward_fn/std": 0.21102523803710938, "step": 2019 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3217.0, "completions/mean_length": 1838.2125244140625, "completions/mean_terminated_length": 688.0188598632812, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.7835531419705197, "grad_norm": 8.222766876220703, "kl": 38.4375, "learning_rate": 1.0326761975768494e-06, "loss": 0.9071, "num_tokens": 393305760.0, "reward": 0.12375000864267349, "reward_std": 0.10962130129337311, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.04500000178813934, "rewards/unicoder_reward_fn/std": 0.13861887156963348, "step": 2020 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3168.0, "completions/mean_length": 1898.8125, "completions/mean_terminated_length": 649.431396484375, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.7839410395655547, "grad_norm": 4.566086769104004, "kl": 42.625, "learning_rate": 1.030850756818148e-06, "loss": 0.8307, "num_tokens": 393494801.0, "reward": 0.1887500137090683, "reward_std": 0.16809654235839844, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.23379099369049072, "step": 2021 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 1418.0, "completions/mean_length": 1923.237548828125, "completions/mean_terminated_length": 474.72918701171875, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.7843289371605896, "grad_norm": 12.091663360595703, "kl": 36.0625, "learning_rate": 1.0290280308307602e-06, "loss": 0.8521, "num_tokens": 393702020.0, "reward": 0.1550000160932541, "reward_std": 0.16288648545742035, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.18106576800346375, "step": 2022 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2866.0, "completions/mean_length": 1683.4625244140625, "completions/mean_terminated_length": 521.870361328125, "completions/min_length": 101.0, "completions/min_terminated_length": 101.0, "epoch": 0.7847168347556245, "grad_norm": 3.5071871280670166, "kl": 43.328125, "learning_rate": 1.027208022493018e-06, "loss": 0.8372, "num_tokens": 393887161.0, "reward": 0.2225000113248825, "reward_std": 0.20250657200813293, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.21810461580753326, "step": 2023 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 3600.0, "completions/mean_length": 1881.8125, "completions/mean_terminated_length": 689.5577392578125, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.7851047323506595, "grad_norm": 5.300375938415527, "kl": 33.71875, "learning_rate": 1.0253907346789632e-06, "loss": 0.652, "num_tokens": 394083930.0, "reward": 0.13875000178813934, "reward_std": 0.11324012279510498, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.0625000074505806, "rewards/unicoder_reward_fn/std": 0.1334782987833023, "step": 2024 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3406.0, "completions/mean_length": 1881.5625, "completions/mean_terminated_length": 622.37255859375, "completions/min_length": 110.0, "completions/min_terminated_length": 110.0, "epoch": 0.7854926299456944, "grad_norm": 7.373656272888184, "kl": 41.375, "learning_rate": 1.0235761702583406e-06, "loss": 0.9369, "num_tokens": 394284623.0, "reward": 0.20250001549720764, "reward_std": 0.2231825888156891, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.24699458479881287, "step": 2025 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3725.0, "completions/mean_length": 2183.125, "completions/mean_terminated_length": 695.3333740234375, "completions/min_length": 139.0, "completions/min_terminated_length": 139.0, "epoch": 0.7858805275407292, "grad_norm": 8.433058738708496, "kl": 33.28125, "learning_rate": 1.0217643320965964e-06, "loss": 0.7116, "num_tokens": 394509665.0, "reward": 0.17125001549720764, "reward_std": 0.16886217892169952, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.23812292516231537, "step": 2026 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2061.0, "completions/mean_length": 1709.987548828125, "completions/mean_terminated_length": 494.4717102050781, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.7862684251357641, "grad_norm": 9.645795822143555, "kl": 27.59375, "learning_rate": 1.0199552230548698e-06, "loss": 0.6269, "num_tokens": 394700216.0, "reward": 0.1575000137090683, "reward_std": 0.15254633128643036, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.15385346114635468, "step": 2027 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7, "completions/max_length": 4096.0, "completions/max_terminated_length": 2696.0, "completions/mean_length": 3057.175048828125, "completions/mean_terminated_length": 633.25, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.7866563227307991, "grad_norm": 15.716290473937988, "kl": 43.4375, "learning_rate": 1.01814884598999e-06, "loss": 0.7291, "num_tokens": 394993030.0, "reward": 0.07500000298023224, "reward_std": 0.07640562206506729, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.06632295995950699, "step": 2028 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.725, "completions/max_length": 4096.0, "completions/max_terminated_length": 3848.0, "completions/mean_length": 3300.0625, "completions/mean_terminated_length": 1201.681884765625, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.787044220325834, "grad_norm": 20.991275787353516, "kl": 30.53125, "learning_rate": 1.0163452037544746e-06, "loss": 0.4025, "num_tokens": 395307979.0, "reward": 0.08500000089406967, "reward_std": 0.08319836854934692, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.0994923859834671, "step": 2029 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.7125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2708.0, "completions/mean_length": 3132.775146484375, "completions/mean_terminated_length": 745.6521606445312, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.7874321179208689, "grad_norm": 18.015336990356445, "kl": 36.65625, "learning_rate": 1.01454429919652e-06, "loss": 0.585, "num_tokens": 395608697.0, "reward": 0.07750000804662704, "reward_std": 0.06851384043693542, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.009999999776482582, "rewards/unicoder_reward_fn/std": 0.054191410541534424, "step": 2030 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2869.0, "completions/mean_length": 3056.837646484375, "completions/mean_terminated_length": 770.6799926757812, "completions/min_length": 159.0, "completions/min_terminated_length": 159.0, "epoch": 0.7878200155159037, "grad_norm": 15.151652336120605, "kl": 29.875, "learning_rate": 1.0127461351599997e-06, "loss": 0.5258, "num_tokens": 395907836.0, "reward": 0.08624999970197678, "reward_std": 0.077349953353405, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.07647288590669632, "step": 2031 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3515.0, "completions/mean_length": 2911.97509765625, "completions/mean_terminated_length": 938.6000366210938, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.7882079131109387, "grad_norm": 18.033653259277344, "kl": 33.53125, "learning_rate": 1.0109507144844616e-06, "loss": 0.5004, "num_tokens": 396198722.0, "reward": 0.07625000923871994, "reward_std": 0.0756082534790039, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.09192044287919998, "step": 2032 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 2937.0, "completions/mean_length": 2522.125, "completions/mean_terminated_length": 598.5, "completions/min_length": 37.0, "completions/min_terminated_length": 37.0, "epoch": 0.7885958107059736, "grad_norm": 2.952035427093506, "kl": 30.265625, "learning_rate": 1.009158040005119e-06, "loss": 0.5905, "num_tokens": 396444516.0, "reward": 0.125, "reward_std": 0.16586633026599884, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.057499997317790985, "rewards/unicoder_reward_fn/std": 0.17485980689525604, "step": 2033 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3295.0, "completions/mean_length": 2238.5, "completions/mean_terminated_length": 640.18603515625, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.7889837083010085, "grad_norm": 5.856095314025879, "kl": 30.0, "learning_rate": 1.0073681145528493e-06, "loss": 0.6768, "num_tokens": 396663404.0, "reward": 0.12375000864267349, "reward_std": 0.13001836836338043, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.17644520103931427, "step": 2034 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3325.0, "completions/mean_length": 1782.1375732421875, "completions/mean_terminated_length": 668.0555419921875, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.7893716058960435, "grad_norm": 11.465377807617188, "kl": 20.453125, "learning_rate": 1.0055809409541903e-06, "loss": 0.5949, "num_tokens": 396848599.0, "reward": 0.19875001907348633, "reward_std": 0.18661002814769745, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.2153007984161377, "step": 2035 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2111.0, "completions/mean_length": 1224.4625244140625, "completions/mean_terminated_length": 506.578125, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.7897595034910784, "grad_norm": 17.667652130126953, "kl": 15.9609375, "learning_rate": 1.0037965220313317e-06, "loss": 0.4877, "num_tokens": 396999132.0, "reward": 0.23875001072883606, "reward_std": 0.1681462824344635, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.24955657124519348, "step": 2036 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 2853.0, "completions/mean_length": 1451.987548828125, "completions/mean_terminated_length": 570.6500244140625, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "epoch": 0.7901474010861133, "grad_norm": 14.874980926513672, "kl": 31.90625, "learning_rate": 1.0020148606021157e-06, "loss": 0.8636, "num_tokens": 397162155.0, "reward": 0.22625000774860382, "reward_std": 0.22590792179107666, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.26529181003570557, "step": 2037 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1690.0, "completions/mean_length": 1535.5374755859375, "completions/mean_terminated_length": 438.1964416503906, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.7905352986811481, "grad_norm": 12.493184089660645, "kl": 25.359375, "learning_rate": 1.0002359594800287e-06, "loss": 0.5638, "num_tokens": 397340622.0, "reward": 0.22500000894069672, "reward_std": 0.18228457868099213, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.23545001447200775, "step": 2038 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2250.0, "completions/mean_length": 1294.6875, "completions/mean_terminated_length": 481.4031982421875, "completions/min_length": 169.0, "completions/min_terminated_length": 169.0, "epoch": 0.7909231962761831, "grad_norm": 16.375778198242188, "kl": 26.015625, "learning_rate": 9.984598214741975e-07, "loss": 0.7251, "num_tokens": 397496989.0, "reward": 0.18250001966953278, "reward_std": 0.1103348359465599, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932424664497375, "rewards/unicoder_reward_fn/mean": 0.09749999642372131, "rewards/unicoder_reward_fn/std": 0.22500351071357727, "step": 2039 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2679.0, "completions/mean_length": 1636.9000244140625, "completions/mean_terminated_length": 452.8888854980469, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "epoch": 0.791311093871218, "grad_norm": 10.862571716308594, "kl": 37.125, "learning_rate": 9.966864493893876e-07, "loss": 0.8481, "num_tokens": 397681197.0, "reward": 0.23000000417232513, "reward_std": 0.19429849088191986, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.21976685523986816, "step": 2040 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2695.0, "completions/mean_length": 2011.75, "completions/mean_terminated_length": 471.2174072265625, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.7916989914662529, "grad_norm": 2.9276435375213623, "kl": 51.3125, "learning_rate": 9.949158460259945e-07, "loss": 0.9841, "num_tokens": 397883705.0, "reward": 0.13750000298023224, "reward_std": 0.1492137908935547, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.07000000029802322, "rewards/unicoder_reward_fn/std": 0.1594293713569641, "step": 2041 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 3865.0, "completions/mean_length": 1727.5750732421875, "completions/mean_terminated_length": 452.2692565917969, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.7920868890612878, "grad_norm": 7.500136375427246, "kl": 40.6875, "learning_rate": 9.931480141800436e-07, "loss": 0.8943, "num_tokens": 398072159.0, "reward": 0.23875001072883606, "reward_std": 0.22574372589588165, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.1599999964237213, "rewards/unicoder_reward_fn/std": 0.2612530589103699, "step": 2042 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2256.0, "completions/mean_length": 1822.5125732421875, "completions/mean_terminated_length": 458.41998291015625, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.7924747866563228, "grad_norm": 2.973869562149048, "kl": 50.125, "learning_rate": 9.913829566431823e-07, "loss": 1.0206, "num_tokens": 398271696.0, "reward": 0.18375001847743988, "reward_std": 0.17830418050289154, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.20351345837116241, "step": 2043 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 3593.0, "completions/mean_length": 2218.97509765625, "completions/mean_terminated_length": 520.7142944335938, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.7928626842513576, "grad_norm": 4.295258045196533, "kl": 48.0, "learning_rate": 9.896206762026768e-07, "loss": 0.9816, "num_tokens": 398490982.0, "reward": 0.1612500101327896, "reward_std": 0.17395734786987305, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.08750000596046448, "rewards/unicoder_reward_fn/std": 0.2058480679988861, "step": 2044 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1992.0, "completions/mean_length": 1920.8250732421875, "completions/mean_terminated_length": 393.574462890625, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.7932505818463925, "grad_norm": 4.157526016235352, "kl": 46.875, "learning_rate": 9.878611756414093e-07, "loss": 0.9142, "num_tokens": 398692712.0, "reward": 0.2562499940395355, "reward_std": 0.2569544315338135, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.1824999898672104, "rewards/unicoder_reward_fn/std": 0.31210798025131226, "step": 2045 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 1438.0, "completions/mean_length": 2451.35009765625, "completions/mean_terminated_length": 441.22222900390625, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.7936384794414275, "grad_norm": 18.391189575195312, "kl": 58.375, "learning_rate": 9.861044577378709e-07, "loss": 0.8395, "num_tokens": 398941836.0, "reward": 0.14250001311302185, "reward_std": 0.1592307835817337, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.22160233557224274, "step": 2046 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 3261.0, "completions/mean_length": 2508.47509765625, "completions/mean_terminated_length": 568.1666870117188, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.7940263770364624, "grad_norm": 11.452363967895508, "kl": 60.625, "learning_rate": 9.843505252661592e-07, "loss": 1.057, "num_tokens": 399183610.0, "reward": 0.1575000137090683, "reward_std": 0.17536203563213348, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.22941863536834717, "step": 2047 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.575, "completions/max_length": 4096.0, "completions/max_terminated_length": 3764.0, "completions/mean_length": 2604.012451171875, "completions/mean_terminated_length": 585.441162109375, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "epoch": 0.7944142746314973, "grad_norm": 16.131338119506836, "kl": 73.625, "learning_rate": 9.825993809959733e-07, "loss": 1.21, "num_tokens": 399433307.0, "reward": 0.13250000774860382, "reward_std": 0.1651231199502945, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.16100946068763733, "step": 2048 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.65, "completions/max_length": 4096.0, "completions/max_terminated_length": 3168.0, "completions/mean_length": 2885.375, "completions/mean_terminated_length": 637.0714721679688, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.7948021722265322, "grad_norm": 18.512712478637695, "kl": 61.5625, "learning_rate": 9.808510276926075e-07, "loss": 0.9323, "num_tokens": 399707561.0, "reward": 0.12250000238418579, "reward_std": 0.15031877160072327, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.18582692742347717, "step": 2049 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5625, "completions/max_length": 4096.0, "completions/max_terminated_length": 993.0, "completions/mean_length": 2452.28759765625, "completions/mean_terminated_length": 338.94287109375, "completions/min_length": 80.0, "completions/min_terminated_length": 80.0, "epoch": 0.7951900698215671, "grad_norm": 5.715693473815918, "kl": 61.5625, "learning_rate": 9.791054681169524e-07, "loss": 1.1885, "num_tokens": 399950112.0, "reward": 0.11874999850988388, "reward_std": 0.1395873874425888, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.057499997317790985, "rewards/unicoder_reward_fn/std": 0.13195224106311798, "step": 2050 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1813.0, "completions/mean_length": 2119.175048828125, "completions/mean_terminated_length": 501.7727355957031, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "epoch": 0.795577967416602, "grad_norm": 4.877274036407471, "kl": 43.6875, "learning_rate": 9.773627050254835e-07, "loss": 0.7994, "num_tokens": 400167878.0, "reward": 0.24625001847743988, "reward_std": 0.2688957154750824, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.3185310363769531, "step": 2051 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2097.0, "completions/mean_length": 1765.0125732421875, "completions/mean_terminated_length": 439.5490417480469, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "epoch": 0.7959658650116369, "grad_norm": 11.28365707397461, "kl": 35.5625, "learning_rate": 9.756227411702613e-07, "loss": 0.9198, "num_tokens": 400362823.0, "reward": 0.1862500011920929, "reward_std": 0.15379571914672852, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.2156972736120224, "step": 2052 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.375, "completions/max_length": 4096.0, "completions/max_terminated_length": 1396.0, "completions/mean_length": 1779.987548828125, "completions/mean_terminated_length": 390.3800048828125, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "epoch": 0.7963537626066718, "grad_norm": 11.774561882019043, "kl": 26.09375, "learning_rate": 9.738855792989277e-07, "loss": 0.5626, "num_tokens": 400550918.0, "reward": 0.17249999940395355, "reward_std": 0.15026584267616272, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.1386873424053192, "step": 2053 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2393.0, "completions/mean_length": 1954.4124755859375, "completions/mean_terminated_length": 450.7446594238281, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.7967416602017068, "grad_norm": 9.310948371887207, "kl": 35.09375, "learning_rate": 9.721512221546967e-07, "loss": 0.8853, "num_tokens": 400747703.0, "reward": 0.20624999701976776, "reward_std": 0.19207440316677094, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.22330941259860992, "step": 2054 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 4038.0, "completions/mean_length": 2298.137451171875, "completions/mean_terminated_length": 500.2749938964844, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.7971295577967417, "grad_norm": 4.85117244720459, "kl": 32.75, "learning_rate": 9.704196724763566e-07, "loss": 0.6879, "num_tokens": 400980346.0, "reward": 0.12000000476837158, "reward_std": 0.09685108810663223, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.04500000178813934, "rewards/unicoder_reward_fn/std": 0.0898662880063057, "step": 2055 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 3639.0, "completions/mean_length": 2752.587646484375, "completions/mean_terminated_length": 737.46875, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.7975174553917765, "grad_norm": 2.5298142433166504, "kl": 36.1875, "learning_rate": 9.686909329982597e-07, "loss": 0.7447, "num_tokens": 401252185.0, "reward": 0.13750000298023224, "reward_std": 0.13160188496112823, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.15126053988933563, "step": 2056 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2683.0, "completions/mean_length": 2796.675048828125, "completions/mean_terminated_length": 631.1333618164062, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.7979053529868115, "grad_norm": 15.316948890686035, "kl": 42.5625, "learning_rate": 9.669650064503214e-07, "loss": 0.7363, "num_tokens": 401517095.0, "reward": 0.17750000953674316, "reward_std": 0.17141270637512207, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.28952163457870483, "step": 2057 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2428.0, "completions/mean_length": 2769.324951171875, "completions/mean_terminated_length": 558.2000122070312, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.7982932505818464, "grad_norm": 5.038659572601318, "kl": 39.8125, "learning_rate": 9.652418955580161e-07, "loss": 0.7791, "num_tokens": 401787305.0, "reward": 0.17250001430511475, "reward_std": 0.18778495490550995, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.19614964723587036, "step": 2058 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 3919.0, "completions/mean_length": 2412.16259765625, "completions/mean_terminated_length": 551.0789794921875, "completions/min_length": 219.0, "completions/min_terminated_length": 219.0, "epoch": 0.7986811481768813, "grad_norm": 5.803683280944824, "kl": 40.75, "learning_rate": 9.635216030423706e-07, "loss": 0.8494, "num_tokens": 402030550.0, "reward": 0.17500001192092896, "reward_std": 0.17450420558452606, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.23780377209186554, "step": 2059 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.48750000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3490.0, "completions/mean_length": 2289.737548828125, "completions/mean_terminated_length": 571.5853271484375, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "epoch": 0.7990690457719162, "grad_norm": 10.557476043701172, "kl": 37.53125, "learning_rate": 9.618041316199617e-07, "loss": 0.9169, "num_tokens": 402252345.0, "reward": 0.23375001549720764, "reward_std": 0.25282278656959534, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.265661358833313, "step": 2060 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3266.0, "completions/mean_length": 1905.9124755859375, "completions/mean_terminated_length": 520.346923828125, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.7994569433669512, "grad_norm": 14.980978965759277, "kl": 28.71875, "learning_rate": 9.600894840029116e-07, "loss": 0.8108, "num_tokens": 402453506.0, "reward": 0.1600000113248825, "reward_std": 0.13263483345508575, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.14821724593639374, "step": 2061 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 1122.0, "completions/mean_length": 2264.0, "completions/mean_terminated_length": 432.0, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "epoch": 0.799844840961986, "grad_norm": 4.194924831390381, "kl": 35.625, "learning_rate": 9.583776628988818e-07, "loss": 0.7484, "num_tokens": 402692778.0, "reward": 0.1925000101327896, "reward_std": 0.16760270297527313, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.21054480969905853, "step": 2062 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 2474.0, "completions/mean_length": 2330.47509765625, "completions/mean_terminated_length": 564.9500122070312, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.8002327385570209, "grad_norm": 6.488803386688232, "kl": 35.96875, "learning_rate": 9.566686710110729e-07, "loss": 0.8016, "num_tokens": 402927704.0, "reward": 0.17625001072883606, "reward_std": 0.18249905109405518, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.20374655723571777, "step": 2063 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 2123.0, "completions/mean_length": 2285.487548828125, "completions/mean_terminated_length": 474.9750061035156, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.8006206361520558, "grad_norm": 5.052083969116211, "kl": 39.40625, "learning_rate": 9.549625110382152e-07, "loss": 0.8338, "num_tokens": 403154879.0, "reward": 0.21375003457069397, "reward_std": 0.2468549758195877, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.442768931388855, "rewards/unicoder_reward_fn/mean": 0.14000000059604645, "rewards/unicoder_reward_fn/std": 0.26700401306152344, "step": 2064 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.525, "completions/max_length": 4096.0, "completions/max_terminated_length": 3213.0, "completions/mean_length": 2385.10009765625, "completions/mean_terminated_length": 494.1052551269531, "completions/min_length": 231.0, "completions/min_terminated_length": 231.0, "epoch": 0.8010085337470908, "grad_norm": 7.727807998657227, "kl": 39.875, "learning_rate": 9.53259185674569e-07, "loss": 0.8877, "num_tokens": 403388783.0, "reward": 0.257500022649765, "reward_std": 0.24566833674907684, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.1824999898672104, "rewards/unicoder_reward_fn/std": 0.29199227690696716, "step": 2065 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 2259.0, "completions/mean_length": 2296.625, "completions/mean_terminated_length": 497.25, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.8013964313421257, "grad_norm": 6.07381010055542, "kl": 46.25, "learning_rate": 9.515586976099175e-07, "loss": 0.8339, "num_tokens": 403617105.0, "reward": 0.1587500125169754, "reward_std": 0.16252413392066956, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.20337344706058502, "step": 2066 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.55, "completions/max_length": 4096.0, "completions/max_terminated_length": 3244.0, "completions/mean_length": 2487.050048828125, "completions/mean_terminated_length": 520.5555419921875, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.8017843289371606, "grad_norm": 4.421860218048096, "kl": 44.875, "learning_rate": 9.498610495295624e-07, "loss": 0.9058, "num_tokens": 403875309.0, "reward": 0.17875000834465027, "reward_std": 0.17105446755886078, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.10249999910593033, "rewards/unicoder_reward_fn/std": 0.19355109333992004, "step": 2067 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6, "completions/max_length": 4096.0, "completions/max_terminated_length": 3886.0, "completions/mean_length": 2681.762451171875, "completions/mean_terminated_length": 560.40625, "completions/min_length": 251.0, "completions/min_terminated_length": 251.0, "epoch": 0.8021722265321956, "grad_norm": 11.311745643615723, "kl": 48.3125, "learning_rate": 9.48166244114323e-07, "loss": 0.8513, "num_tokens": 404141786.0, "reward": 0.1887500137090683, "reward_std": 0.19427652657032013, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.2677142024040222, "step": 2068 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2197.0, "completions/mean_length": 2490.75, "completions/mean_terminated_length": 625.189208984375, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "epoch": 0.8025601241272304, "grad_norm": 8.430987358093262, "kl": 55.4375, "learning_rate": 9.464742840405275e-07, "loss": 1.048, "num_tokens": 404381718.0, "reward": 0.13375000655651093, "reward_std": 0.1441127508878708, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.06499999761581421, "rewards/unicoder_reward_fn/std": 0.14849455654621124, "step": 2069 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4076.0, "completions/mean_length": 2456.550048828125, "completions/mean_terminated_length": 733.025634765625, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.8029480217222653, "grad_norm": 7.470353126525879, "kl": 50.125, "learning_rate": 9.447851719800127e-07, "loss": 0.9218, "num_tokens": 404618162.0, "reward": 0.13375000655651093, "reward_std": 0.13485945761203766, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.442768931388855, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.15061058104038239, "step": 2070 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2108.0, "completions/mean_length": 2407.16259765625, "completions/mean_terminated_length": 444.45947265625, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.8033359193173002, "grad_norm": 2.233905792236328, "kl": 47.0625, "learning_rate": 9.430989106001168e-07, "loss": 0.8823, "num_tokens": 404854215.0, "reward": 0.21125002205371857, "reward_std": 0.23635955154895782, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2661968469619751, "step": 2071 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3451.0, "completions/mean_length": 2474.074951171875, "completions/mean_terminated_length": 589.1351318359375, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.8037238169123352, "grad_norm": 6.491055965423584, "kl": 36.59375, "learning_rate": 9.414155025636757e-07, "loss": 0.8795, "num_tokens": 405100797.0, "reward": 0.16250000894069672, "reward_std": 0.17033155262470245, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.1738433837890625, "step": 2072 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2113.0, "completions/mean_length": 2407.237548828125, "completions/mean_terminated_length": 444.62164306640625, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.8041117145073701, "grad_norm": 8.032769203186035, "kl": 52.6875, "learning_rate": 9.397349505290218e-07, "loss": 0.9826, "num_tokens": 405338512.0, "reward": 0.15625, "reward_std": 0.15033039450645447, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.20861205458641052, "step": 2073 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2070.0, "completions/mean_length": 1863.8875732421875, "completions/mean_terminated_length": 451.73468017578125, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "epoch": 0.804499612102405, "grad_norm": 3.125528573989868, "kl": 32.53125, "learning_rate": 9.380572571499759e-07, "loss": 0.5983, "num_tokens": 405527631.0, "reward": 0.23125000298023224, "reward_std": 0.2405518740415573, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.2876245081424713, "step": 2074 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2939.0, "completions/mean_length": 2129.9375, "completions/mean_terminated_length": 521.3409423828125, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.8048875096974398, "grad_norm": 7.352111339569092, "kl": 36.0, "learning_rate": 9.363824250758444e-07, "loss": 0.8136, "num_tokens": 405748762.0, "reward": 0.22625000774860382, "reward_std": 0.2388623058795929, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.24903611838817596, "step": 2075 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1183.0, "completions/mean_length": 2304.53759765625, "completions/mean_terminated_length": 421.20513916015625, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "epoch": 0.8052754072924748, "grad_norm": 2.203463077545166, "kl": 43.65625, "learning_rate": 9.347104569514154e-07, "loss": 0.9064, "num_tokens": 405972693.0, "reward": 0.1537500023841858, "reward_std": 0.1818312406539917, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.19833165407180786, "step": 2076 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3143.0, "completions/mean_length": 2120.800048828125, "completions/mean_terminated_length": 504.727294921875, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.8056633048875097, "grad_norm": 4.145862579345703, "kl": 31.09375, "learning_rate": 9.330413554169557e-07, "loss": 0.6519, "num_tokens": 406190165.0, "reward": 0.2237500250339508, "reward_std": 0.22728104889392853, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.26719358563423157, "step": 2077 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 731.0, "completions/mean_length": 1914.5750732421875, "completions/mean_terminated_length": 382.9361572265625, "completions/min_length": 233.0, "completions/min_terminated_length": 233.0, "epoch": 0.8060512024825446, "grad_norm": 10.427948951721191, "kl": 36.0625, "learning_rate": 9.313751231082044e-07, "loss": 0.858, "num_tokens": 406399123.0, "reward": 0.257500022649765, "reward_std": 0.2319079488515854, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.2512626647949219, "step": 2078 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.38749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3775.0, "completions/mean_length": 1866.2750244140625, "completions/mean_terminated_length": 455.63262939453125, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.8064391000775796, "grad_norm": 8.446389198303223, "kl": 34.0625, "learning_rate": 9.297117626563687e-07, "loss": 0.8304, "num_tokens": 406597953.0, "reward": 0.20500002801418304, "reward_std": 0.18967992067337036, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.23621487617492676, "step": 2079 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1019.0, "completions/mean_length": 1935.125, "completions/mean_terminated_length": 417.9148864746094, "completions/min_length": 196.0, "completions/min_terminated_length": 196.0, "epoch": 0.8068269976726145, "grad_norm": 2.9039723873138428, "kl": 43.1875, "learning_rate": 9.280512766881228e-07, "loss": 0.8493, "num_tokens": 406806579.0, "reward": 0.14625000953674316, "reward_std": 0.1470271497964859, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.1698659211397171, "step": 2080 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3154.0, "completions/mean_length": 1811.0625, "completions/mean_terminated_length": 511.7843322753906, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "epoch": 0.8072148952676493, "grad_norm": 2.0115673542022705, "kl": 37.03125, "learning_rate": 9.263936678256008e-07, "loss": 0.6651, "num_tokens": 407001744.0, "reward": 0.1262500137090683, "reward_std": 0.12383164465427399, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.12218993902206421, "step": 2081 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 3261.0, "completions/mean_length": 1956.925048828125, "completions/mean_terminated_length": 530.875, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "epoch": 0.8076027928626842, "grad_norm": 3.8537862300872803, "kl": 35.25, "learning_rate": 9.24738938686392e-07, "loss": 0.7656, "num_tokens": 407207770.0, "reward": 0.15000000596046448, "reward_std": 0.1017790213227272, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.14563262462615967, "step": 2082 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 1198.0, "completions/mean_length": 1966.425048828125, "completions/mean_terminated_length": 392.3913269042969, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.8079906904577192, "grad_norm": 6.502084255218506, "kl": 42.375, "learning_rate": 9.230870918835412e-07, "loss": 0.9341, "num_tokens": 407410012.0, "reward": 0.20250001549720764, "reward_std": 0.20722909271717072, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.23587971925735474, "step": 2083 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2900.0, "completions/mean_length": 2401.41259765625, "completions/mean_terminated_length": 619.923095703125, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.8083785880527541, "grad_norm": 14.198689460754395, "kl": 46.1875, "learning_rate": 9.214381300255387e-07, "loss": 0.7645, "num_tokens": 407639885.0, "reward": 0.125, "reward_std": 0.12334638088941574, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.05250000208616257, "rewards/unicoder_reward_fn/std": 0.15829390287399292, "step": 2084 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 3740.0, "completions/mean_length": 2331.087646484375, "completions/mean_terminated_length": 566.1749877929688, "completions/min_length": 193.0, "completions/min_terminated_length": 193.0, "epoch": 0.808766485647789, "grad_norm": 10.3096284866333, "kl": 47.0625, "learning_rate": 9.197920557163216e-07, "loss": 0.8548, "num_tokens": 407879236.0, "reward": 0.15250001847743988, "reward_std": 0.15348374843597412, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.18195493519306183, "step": 2085 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.6125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2362.0, "completions/mean_length": 2781.074951171875, "completions/mean_terminated_length": 702.6451416015625, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "epoch": 0.8091543832428238, "grad_norm": 9.624759674072266, "kl": 49.0625, "learning_rate": 9.181488715552652e-07, "loss": 0.8812, "num_tokens": 408145370.0, "reward": 0.16375000774860382, "reward_std": 0.19986578822135925, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.21345560252666473, "step": 2086 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2455.0, "completions/mean_length": 2409.85009765625, "completions/mean_terminated_length": 637.2307739257812, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "epoch": 0.8095422808378588, "grad_norm": 8.51860523223877, "kl": 51.0625, "learning_rate": 9.165085801371811e-07, "loss": 0.9579, "num_tokens": 408382118.0, "reward": 0.10000000149011612, "reward_std": 0.09924260526895523, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.0325000025331974, "rewards/unicoder_reward_fn/std": 0.10283944010734558, "step": 2087 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1016.0, "completions/mean_length": 2116.800048828125, "completions/mean_terminated_length": 413.7674255371094, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.8099301784328937, "grad_norm": 4.673130035400391, "kl": 51.5625, "learning_rate": 9.148711840523148e-07, "loss": 0.9758, "num_tokens": 408600686.0, "reward": 0.19875000417232513, "reward_std": 0.2235405296087265, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.2229832410812378, "step": 2088 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3161.0, "completions/mean_length": 2221.0625, "completions/mean_terminated_length": 607.7442016601562, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.8103180760279286, "grad_norm": 5.010804653167725, "kl": 41.5, "learning_rate": 9.132366858863362e-07, "loss": 0.9009, "num_tokens": 408833907.0, "reward": 0.1912500113248825, "reward_std": 0.1894081085920334, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.25774091482162476, "step": 2089 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 3142.0, "completions/mean_length": 1931.5750732421875, "completions/mean_terminated_length": 488.625, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.8107059736229635, "grad_norm": 2.4274797439575195, "kl": 40.25, "learning_rate": 9.116050882203428e-07, "loss": 0.8264, "num_tokens": 409036241.0, "reward": 0.21250000596046448, "reward_std": 0.21494410932064056, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.28064122796058655, "step": 2090 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4084.0, "completions/mean_length": 2038.112548828125, "completions/mean_terminated_length": 593.2127685546875, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.8110938712179985, "grad_norm": 7.763106346130371, "kl": 43.5, "learning_rate": 9.099763936308495e-07, "loss": 1.0468, "num_tokens": 409252210.0, "reward": 0.14750002324581146, "reward_std": 0.14295987784862518, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.15424376726150513, "step": 2091 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3502.0, "completions/mean_length": 1675.1624755859375, "completions/mean_terminated_length": 509.5740661621094, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.8114817688130334, "grad_norm": 7.268515110015869, "kl": 35.15625, "learning_rate": 9.083506046897864e-07, "loss": 0.8449, "num_tokens": 409436943.0, "reward": 0.21875, "reward_std": 0.22777102887630463, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.2685050070285797, "step": 2092 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3258.0, "completions/mean_length": 1604.0250244140625, "completions/mean_terminated_length": 404.1851806640625, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.8118696664080682, "grad_norm": 11.221675872802734, "kl": 28.34375, "learning_rate": 9.067277239644978e-07, "loss": 0.8173, "num_tokens": 409621121.0, "reward": 0.21250002086162567, "reward_std": 0.20920096337795258, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.2300110012292862, "step": 2093 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.35, "completions/max_length": 4096.0, "completions/max_terminated_length": 1749.0, "completions/mean_length": 1726.6500244140625, "completions/mean_terminated_length": 450.8461608886719, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.8122575640031032, "grad_norm": 3.9747018814086914, "kl": 40.34375, "learning_rate": 9.051077540177322e-07, "loss": 0.8645, "num_tokens": 409817989.0, "reward": 0.26250001788139343, "reward_std": 0.19480949640274048, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.27804699540138245, "step": 2094 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2543.0, "completions/mean_length": 1724.9000244140625, "completions/mean_terminated_length": 376.6274719238281, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "epoch": 0.8126454615981381, "grad_norm": 5.1966552734375, "kl": 40.8125, "learning_rate": 9.034906974076449e-07, "loss": 0.9534, "num_tokens": 409999773.0, "reward": 0.16624999046325684, "reward_std": 0.1291232407093048, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.09000000357627869, "rewards/unicoder_reward_fn/std": 0.1797325760126114, "step": 2095 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3148.0, "completions/mean_length": 1810.25, "completions/mean_terminated_length": 510.50982666015625, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.813033359193173, "grad_norm": 6.593033790588379, "kl": 42.375, "learning_rate": 9.018765566877882e-07, "loss": 0.9411, "num_tokens": 410197769.0, "reward": 0.22750000655651093, "reward_std": 0.20896072685718536, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.24437987804412842, "step": 2096 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2861.0, "completions/mean_length": 1725.237548828125, "completions/mean_terminated_length": 517.4906005859375, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.8134212567882079, "grad_norm": 2.492831230163574, "kl": 49.0625, "learning_rate": 9.0026533440711e-07, "loss": 1.0571, "num_tokens": 410388060.0, "reward": 0.18250000476837158, "reward_std": 0.14478500187397003, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.16831256449222565, "step": 2097 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3646.0, "completions/mean_length": 2418.337646484375, "completions/mean_terminated_length": 654.6410522460938, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.8138091543832429, "grad_norm": 11.406526565551758, "kl": 60.375, "learning_rate": 8.986570331099512e-07, "loss": 1.1101, "num_tokens": 410624975.0, "reward": 0.13875000178813934, "reward_std": 0.17450299859046936, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.18416763842105865, "step": 2098 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 3743.0, "completions/mean_length": 2379.875, "completions/mean_terminated_length": 663.75, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.8141970519782777, "grad_norm": 18.87868309020996, "kl": 61.75, "learning_rate": 8.970516553360384e-07, "loss": 1.0, "num_tokens": 410858605.0, "reward": 0.11124999821186066, "reward_std": 0.12216412276029587, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.042500000447034836, "rewards/unicoder_reward_fn/std": 0.1177586019039154, "step": 2099 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 3922.0, "completions/mean_length": 2304.800048828125, "completions/mean_terminated_length": 684.1904907226562, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.8145849495733126, "grad_norm": 14.148321151733398, "kl": 57.6875, "learning_rate": 8.954492036204829e-07, "loss": 0.9992, "num_tokens": 411089837.0, "reward": 0.15625, "reward_std": 0.1793701946735382, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375511169433594, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.23372331261634827, "step": 2100 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3638.0, "completions/mean_length": 2245.887451171875, "completions/mean_terminated_length": 878.4130859375, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.8149728471683475, "grad_norm": 17.921632766723633, "kl": 58.5625, "learning_rate": 8.938496804937744e-07, "loss": 0.9414, "num_tokens": 411313348.0, "reward": 0.1237499937415123, "reward_std": 0.1667783260345459, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.18608218431472778, "step": 2101 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4, "completions/max_length": 4096.0, "completions/max_terminated_length": 2244.0, "completions/mean_length": 1923.487548828125, "completions/mean_terminated_length": 475.1458435058594, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.8153607447633825, "grad_norm": 9.842001914978027, "kl": 38.90625, "learning_rate": 8.922530884817776e-07, "loss": 0.5843, "num_tokens": 411517219.0, "reward": 0.13125000894069672, "reward_std": 0.1064048781991005, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.12010016292333603, "step": 2102 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2074.0, "completions/mean_length": 2229.900146484375, "completions/mean_terminated_length": 624.18603515625, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.8157486423584174, "grad_norm": 13.875279426574707, "kl": 50.6875, "learning_rate": 8.906594301057305e-07, "loss": 0.7933, "num_tokens": 411743403.0, "reward": 0.14500001072883606, "reward_std": 0.164891317486763, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.17603221535682678, "step": 2103 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 2445.0, "completions/mean_length": 1761.5875244140625, "completions/mean_terminated_length": 572.3585205078125, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.8161365399534523, "grad_norm": 4.716017246246338, "kl": 35.9375, "learning_rate": 8.890687078822357e-07, "loss": 0.7076, "num_tokens": 411931802.0, "reward": 0.16375000774860382, "reward_std": 0.13229595124721527, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.16746433079242706, "step": 2104 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2417.0, "completions/mean_length": 2213.85009765625, "completions/mean_terminated_length": 594.3255615234375, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "epoch": 0.8165244375484872, "grad_norm": 6.854681968688965, "kl": 45.125, "learning_rate": 8.874809243232624e-07, "loss": 0.8785, "num_tokens": 412150046.0, "reward": 0.11625001579523087, "reward_std": 0.12689147889614105, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.13650241494178772, "step": 2105 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.5, "completions/max_length": 4096.0, "completions/max_terminated_length": 3389.0, "completions/mean_length": 2289.112548828125, "completions/mean_terminated_length": 482.2250061035156, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.8169123351435221, "grad_norm": 4.394760608673096, "kl": 42.0, "learning_rate": 8.858960819361365e-07, "loss": 0.8338, "num_tokens": 412384879.0, "reward": 0.16375000774860382, "reward_std": 0.1854313462972641, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.20124612748622894, "step": 2106 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 3865.0, "completions/mean_length": 1490.3250732421875, "completions/mean_terminated_length": 501.96551513671875, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.817300232738557, "grad_norm": 4.890352725982666, "kl": 36.71875, "learning_rate": 8.8431418322354e-07, "loss": 0.7564, "num_tokens": 412552113.0, "reward": 0.20999999344348907, "reward_std": 0.20525617897510529, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.2463788539171219, "step": 2107 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3607.0, "completions/mean_length": 1791.362548828125, "completions/mean_terminated_length": 617.3018798828125, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.8176881303335919, "grad_norm": 6.961374282836914, "kl": 31.40625, "learning_rate": 8.827352306835083e-07, "loss": 0.574, "num_tokens": 412744638.0, "reward": 0.15375001728534698, "reward_std": 0.1369778960943222, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.15893235802650452, "step": 2108 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.475, "completions/max_length": 4096.0, "completions/max_terminated_length": 3376.0, "completions/mean_length": 2323.550048828125, "completions/mean_terminated_length": 719.90478515625, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.8180760279286269, "grad_norm": 6.43569803237915, "kl": 34.34375, "learning_rate": 8.811592268094213e-07, "loss": 0.6502, "num_tokens": 412984834.0, "reward": 0.10875000804662704, "reward_std": 0.11660905182361603, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.13156795501708984, "step": 2109 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3678.0, "completions/mean_length": 2149.262451171875, "completions/mean_terminated_length": 635.1333618164062, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.8184639255236618, "grad_norm": 3.7912254333496094, "kl": 35.0, "learning_rate": 8.795861740900061e-07, "loss": 0.7765, "num_tokens": 413197535.0, "reward": 0.19625000655651093, "reward_std": 0.22866728901863098, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.23079390823841095, "step": 2110 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2665.0, "completions/mean_length": 2067.837646484375, "completions/mean_terminated_length": 643.8084716796875, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.8188518231186966, "grad_norm": 6.181737422943115, "kl": 26.3125, "learning_rate": 8.780160750093264e-07, "loss": 0.6058, "num_tokens": 413404450.0, "reward": 0.15250001847743988, "reward_std": 0.16920094192028046, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.17500454187393188, "step": 2111 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.32499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3856.0, "completions/mean_length": 1881.7874755859375, "completions/mean_terminated_length": 815.6851806640625, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.8192397207137315, "grad_norm": 5.636629581451416, "kl": 26.5, "learning_rate": 8.764489320467828e-07, "loss": 0.626, "num_tokens": 413602865.0, "reward": 0.1849999874830246, "reward_std": 0.18952830135822296, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.2181481420993805, "step": 2112 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.44999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3832.0, "completions/mean_length": 2298.887451171875, "completions/mean_terminated_length": 828.5227661132812, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.8196276183087665, "grad_norm": 4.56549072265625, "kl": 28.15625, "learning_rate": 8.748847476771092e-07, "loss": 0.6094, "num_tokens": 413837864.0, "reward": 0.1862500160932541, "reward_std": 0.20841650664806366, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.10749999433755875, "rewards/unicoder_reward_fn/std": 0.2524687945842743, "step": 2113 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.42500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 4088.0, "completions/mean_length": 2241.699951171875, "completions/mean_terminated_length": 871.1304321289062, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.8200155159038014, "grad_norm": 3.883087635040283, "kl": 30.78125, "learning_rate": 8.73323524370365e-07, "loss": 0.6654, "num_tokens": 414060408.0, "reward": 0.125, "reward_std": 0.13834530115127563, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.05499999597668648, "rewards/unicoder_reward_fn/std": 0.18548600375652313, "step": 2114 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.36250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3755.0, "completions/mean_length": 1942.7249755859375, "completions/mean_terminated_length": 718.313720703125, "completions/min_length": 92.0, "completions/min_terminated_length": 92.0, "epoch": 0.8204034134988363, "grad_norm": 6.64755392074585, "kl": 27.09375, "learning_rate": 8.717652645919353e-07, "loss": 0.5787, "num_tokens": 414273674.0, "reward": 0.10125000774860382, "reward_std": 0.07923507690429688, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.08641964197158813, "step": 2115 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.4125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3278.0, "completions/mean_length": 2055.987548828125, "completions/mean_terminated_length": 623.6383056640625, "completions/min_length": 110.0, "completions/min_terminated_length": 110.0, "epoch": 0.8207913110938713, "grad_norm": 9.064362525939941, "kl": 33.15625, "learning_rate": 8.702099708025252e-07, "loss": 0.6828, "num_tokens": 414485473.0, "reward": 0.10249999910593033, "reward_std": 0.09471316635608673, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.0325000025331974, "rewards/unicoder_reward_fn/std": 0.14032061398029327, "step": 2116 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2438.0, "completions/mean_length": 1622.5875244140625, "completions/mean_terminated_length": 624.5438842773438, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.8211792086889061, "grad_norm": 9.300564765930176, "kl": 25.90625, "learning_rate": 8.686576454581553e-07, "loss": 0.636, "num_tokens": 414663440.0, "reward": 0.1262500137090683, "reward_std": 0.11635226011276245, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.15008436143398285, "step": 2117 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 3820.0, "completions/mean_length": 1461.8375244140625, "completions/mean_terminated_length": 583.7833862304688, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.821567106283941, "grad_norm": 8.37649154663086, "kl": 23.875, "learning_rate": 8.671082910101603e-07, "loss": 0.6084, "num_tokens": 414830107.0, "reward": 0.21124999225139618, "reward_std": 0.17879627645015717, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.1250000149011612, "rewards/unicoder_reward_fn/std": 0.27627724409103394, "step": 2118 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3589.0, "completions/mean_length": 1557.0374755859375, "completions/mean_terminated_length": 532.5438842773438, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.8219550038789759, "grad_norm": 3.699815511703491, "kl": 32.0625, "learning_rate": 8.655619099051814e-07, "loss": 0.6205, "num_tokens": 415010686.0, "reward": 0.12000000476837158, "reward_std": 0.11308523267507553, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.042500000447034836, "rewards/unicoder_reward_fn/std": 0.18195493519306183, "step": 2119 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3634.0, "completions/mean_length": 1203.9124755859375, "completions/mean_terminated_length": 642.7611694335938, "completions/min_length": 89.0, "completions/min_terminated_length": 89.0, "epoch": 0.8223429014740109, "grad_norm": 7.910285472869873, "kl": 27.59375, "learning_rate": 8.640185045851654e-07, "loss": 0.6612, "num_tokens": 415151359.0, "reward": 0.1537499874830246, "reward_std": 0.12033797800540924, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.14563262462615967, "step": 2120 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.25, "completions/max_length": 4096.0, "completions/max_terminated_length": 3532.0, "completions/mean_length": 1399.4749755859375, "completions/mean_terminated_length": 500.63336181640625, "completions/min_length": 93.0, "completions/min_terminated_length": 93.0, "epoch": 0.8227307990690458, "grad_norm": 4.262950420379639, "kl": 39.15625, "learning_rate": 8.624780774873606e-07, "loss": 0.7868, "num_tokens": 415316885.0, "reward": 0.1925000101327896, "reward_std": 0.21493735909461975, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.24396513402462006, "step": 2121 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3935.0, "completions/mean_length": 1079.8375244140625, "completions/mean_terminated_length": 599.0, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.8231186966640807, "grad_norm": 28.85478973388672, "kl": 21.203125, "learning_rate": 8.609406310443114e-07, "loss": 0.7122, "num_tokens": 415438904.0, "reward": 0.3162500262260437, "reward_std": 0.2758750319480896, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.23000001907348633, "rewards/unicoder_reward_fn/std": 0.30539456009864807, "step": 2122 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3400.0, "completions/mean_length": 1007.6875, "completions/mean_terminated_length": 462.6911926269531, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.8235065942591155, "grad_norm": 8.502440452575684, "kl": 23.09375, "learning_rate": 8.594061676838556e-07, "loss": 0.5283, "num_tokens": 415564039.0, "reward": 0.19875000417232513, "reward_std": 0.15756013989448547, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.1587929129600525, "step": 2123 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2065.0, "completions/mean_length": 854.6875, "completions/mean_terminated_length": 337.95654296875, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.8238944918541505, "grad_norm": 9.147631645202637, "kl": 27.21875, "learning_rate": 8.578746898291198e-07, "loss": 0.6649, "num_tokens": 415669870.0, "reward": 0.26875001192092896, "reward_std": 0.18780554831027985, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.18000000715255737, "rewards/unicoder_reward_fn/std": 0.24567139148712158, "step": 2124 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3803.0, "completions/mean_length": 1041.3875732421875, "completions/mean_terminated_length": 448.70147705078125, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.8242823894491854, "grad_norm": 6.520763397216797, "kl": 31.8125, "learning_rate": 8.563461998985173e-07, "loss": 0.7061, "num_tokens": 415796205.0, "reward": 0.15125000476837158, "reward_std": 0.09895149618387222, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.12487969547510147, "step": 2125 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2662.0, "completions/mean_length": 1137.112548828125, "completions/mean_terminated_length": 509.4697265625, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.8246702870442203, "grad_norm": 3.3481740951538086, "kl": 36.28125, "learning_rate": 8.548207003057421e-07, "loss": 0.7132, "num_tokens": 415935686.0, "reward": 0.1550000160932541, "reward_std": 0.13136498630046844, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.18582692742347717, "step": 2126 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3375, "completions/max_length": 4096.0, "completions/max_terminated_length": 3622.0, "completions/mean_length": 1748.112548828125, "completions/mean_terminated_length": 552.0188598632812, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.8250581846392553, "grad_norm": 14.884313583374023, "kl": 61.25, "learning_rate": 8.532981934597658e-07, "loss": 1.0224, "num_tokens": 416129487.0, "reward": 0.17249999940395355, "reward_std": 0.17074254155158997, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.2110702097415924, "step": 2127 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2658.0, "completions/mean_length": 1460.5875244140625, "completions/mean_terminated_length": 522.559326171875, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.8254460822342902, "grad_norm": 14.763761520385742, "kl": 55.375, "learning_rate": 8.517786817648352e-07, "loss": 0.8979, "num_tokens": 416295638.0, "reward": 0.20875000953674316, "reward_std": 0.2289596050977707, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.2664344906806946, "step": 2128 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 1377.0, "completions/mean_length": 1456.300048828125, "completions/mean_terminated_length": 391.15789794921875, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.825833979829325, "grad_norm": 12.387807846069336, "kl": 57.375, "learning_rate": 8.502621676204667e-07, "loss": 0.9646, "num_tokens": 416453374.0, "reward": 0.13375000655651093, "reward_std": 0.10840243101119995, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.11462653428316116, "step": 2129 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 3467.0, "completions/mean_length": 1562.1875, "completions/mean_terminated_length": 601.086181640625, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.8262218774243599, "grad_norm": 8.017066955566406, "kl": 54.75, "learning_rate": 8.487486534214423e-07, "loss": 0.9455, "num_tokens": 416625493.0, "reward": 0.17750000953674316, "reward_std": 0.15326452255249023, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.17134982347488403, "step": 2130 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2290.0, "completions/mean_length": 1037.2000732421875, "completions/mean_terminated_length": 443.70147705078125, "completions/min_length": 90.0, "completions/min_terminated_length": 90.0, "epoch": 0.8266097750193949, "grad_norm": 3.544353723526001, "kl": 35.546875, "learning_rate": 8.472381415578085e-07, "loss": 0.6426, "num_tokens": 416763549.0, "reward": 0.17875000834465027, "reward_std": 0.16332700848579407, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.1987142115831375, "step": 2131 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2465.0, "completions/mean_length": 860.9000244140625, "completions/mean_terminated_length": 450.81689453125, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.8269976726144298, "grad_norm": 5.654169082641602, "kl": 32.640625, "learning_rate": 8.457306344148685e-07, "loss": 0.6261, "num_tokens": 416880509.0, "reward": 0.23499999940395355, "reward_std": 0.19087792932987213, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.23971238732337952, "step": 2132 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.21250000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3728.0, "completions/mean_length": 1333.8250732421875, "completions/mean_terminated_length": 588.4761962890625, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.8273855702094647, "grad_norm": 3.3009133338928223, "kl": 44.625, "learning_rate": 8.442261343731834e-07, "loss": 0.8902, "num_tokens": 417037383.0, "reward": 0.17625001072883606, "reward_std": 0.18289732933044434, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.0924999937415123, "rewards/unicoder_reward_fn/std": 0.21095022559165955, "step": 2133 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3216.0, "completions/mean_length": 1113.625, "completions/mean_terminated_length": 587.3235473632812, "completions/min_length": 118.0, "completions/min_terminated_length": 118.0, "epoch": 0.8277734678044996, "grad_norm": 6.647755146026611, "kl": 32.78125, "learning_rate": 8.427246438085633e-07, "loss": 0.6157, "num_tokens": 417164801.0, "reward": 0.26250001788139343, "reward_std": 0.2628338038921356, "rewards/format_reward/mean": 0.925000011920929, "rewards/format_reward/std": 0.2650531232357025, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.295279324054718, "step": 2134 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3125.0, "completions/mean_length": 1121.0250244140625, "completions/mean_terminated_length": 543.791015625, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.8281613653995346, "grad_norm": 3.7421252727508545, "kl": 33.203125, "learning_rate": 8.412261650920662e-07, "loss": 0.6269, "num_tokens": 417303035.0, "reward": 0.1850000023841858, "reward_std": 0.17458127439022064, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.1855372041463852, "step": 2135 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3879.0, "completions/mean_length": 1625.425048828125, "completions/mean_terminated_length": 628.5263061523438, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.8285492629945694, "grad_norm": 3.8894870281219482, "kl": 44.625, "learning_rate": 8.397307005899957e-07, "loss": 0.8984, "num_tokens": 417486893.0, "reward": 0.1912500113248825, "reward_std": 0.16486607491970062, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.20609387755393982, "step": 2136 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3288.0, "completions/mean_length": 903.7999877929688, "completions/mean_terminated_length": 499.1549072265625, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.8289371605896043, "grad_norm": 3.472851037979126, "kl": 26.90625, "learning_rate": 8.382382526638932e-07, "loss": 0.5133, "num_tokens": 417599181.0, "reward": 0.1587499976158142, "reward_std": 0.11872177571058273, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.14563262462615967, "step": 2137 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.3125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2726.0, "completions/mean_length": 1696.7874755859375, "completions/mean_terminated_length": 606.236328125, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.8293250581846393, "grad_norm": 8.290671348571777, "kl": 43.4375, "learning_rate": 8.36748823670539e-07, "loss": 0.8056, "num_tokens": 417777268.0, "reward": 0.11375000327825546, "reward_std": 0.11474180221557617, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.11810743808746338, "step": 2138 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.2875, "completions/max_length": 4096.0, "completions/max_terminated_length": 3888.0, "completions/mean_length": 1700.5999755859375, "completions/mean_terminated_length": 734.0350952148438, "completions/min_length": 110.0, "completions/min_terminated_length": 110.0, "epoch": 0.8297129557796742, "grad_norm": 8.336878776550293, "kl": 43.4375, "learning_rate": 8.352624159619438e-07, "loss": 0.8255, "num_tokens": 417963100.0, "reward": 0.17125001549720764, "reward_std": 0.181444451212883, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.22041387856006622, "step": 2139 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2017.0, "completions/mean_length": 1028.4625244140625, "completions/mean_terminated_length": 487.1323547363281, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.8301008533747091, "grad_norm": 3.860307216644287, "kl": 31.0625, "learning_rate": 8.337790318853481e-07, "loss": 0.6644, "num_tokens": 418082329.0, "reward": 0.1275000125169754, "reward_std": 0.11460492759943008, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.04749999940395355, "rewards/unicoder_reward_fn/std": 0.1598852425813675, "step": 2140 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3937.0, "completions/mean_length": 1234.3375244140625, "completions/mean_terminated_length": 778.1304321289062, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.830488750969744, "grad_norm": 4.837555885314941, "kl": 32.8125, "learning_rate": 8.32298673783219e-07, "loss": 0.6976, "num_tokens": 418224956.0, "reward": 0.13125000894069672, "reward_std": 0.12643150985240936, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.15751230716705322, "step": 2141 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3320.0, "completions/mean_length": 1021.5499877929688, "completions/mean_terminated_length": 631.8309936523438, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.8308766485647789, "grad_norm": 4.876056671142578, "kl": 23.65625, "learning_rate": 8.308213439932422e-07, "loss": 0.436, "num_tokens": 418348176.0, "reward": 0.17499999701976776, "reward_std": 0.16709789633750916, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.09250000864267349, "rewards/unicoder_reward_fn/std": 0.2036222666501999, "step": 2142 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3749.0, "completions/mean_length": 658.5499877929688, "completions/mean_terminated_length": 524.6233520507812, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.8312645461598138, "grad_norm": 12.607420921325684, "kl": 17.765625, "learning_rate": 8.293470448483251e-07, "loss": 0.4492, "num_tokens": 418448332.0, "reward": 0.20000000298023224, "reward_std": 0.18009793758392334, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927276134491, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.2108452022075653, "step": 2143 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2345.0, "completions/mean_length": 989.2999877929688, "completions/mean_terminated_length": 545.4857177734375, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.8316524437548487, "grad_norm": 10.36239242553711, "kl": 28.8125, "learning_rate": 8.278757786765862e-07, "loss": 0.73, "num_tokens": 418593412.0, "reward": 0.19875000417232513, "reward_std": 0.1916392594575882, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.2355172038078308, "step": 2144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2062.0, "completions/mean_length": 902.3125, "completions/mean_terminated_length": 446.0714416503906, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.8320403413498836, "grad_norm": 2.738147735595703, "kl": 29.8125, "learning_rate": 8.264075478013553e-07, "loss": 0.5788, "num_tokens": 418724829.0, "reward": 0.16500000655651093, "reward_std": 0.11770067363977432, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.1734059453010559, "step": 2145 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3247.0, "completions/mean_length": 741.6749877929688, "completions/mean_terminated_length": 469.7027282714844, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.8324282389449186, "grad_norm": 7.522624492645264, "kl": 24.265625, "learning_rate": 8.249423545411702e-07, "loss": 0.3779, "num_tokens": 418825787.0, "reward": 0.2212500125169754, "reward_std": 0.21904289722442627, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.2699156105518341, "step": 2146 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1495.0, "completions/mean_length": 720.3125, "completions/mean_terminated_length": 396.6164245605469, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.8328161365399535, "grad_norm": 3.1847431659698486, "kl": 24.609375, "learning_rate": 8.234802012097704e-07, "loss": 0.4105, "num_tokens": 418927212.0, "reward": 0.18125002086162567, "reward_std": 0.1690477579832077, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.0949999988079071, "rewards/unicoder_reward_fn/std": 0.24229909479618073, "step": 2147 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3839.0, "completions/mean_length": 919.3875122070312, "completions/mean_terminated_length": 566.4305419921875, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.8332040341349883, "grad_norm": 4.3289713859558105, "kl": 27.40625, "learning_rate": 8.220210901160963e-07, "loss": 0.5521, "num_tokens": 419041715.0, "reward": 0.20875000953674316, "reward_std": 0.17327769100666046, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.23099946975708008, "step": 2148 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3717.0, "completions/mean_length": 1319.7874755859375, "completions/mean_terminated_length": 625.734375, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.8335919317300233, "grad_norm": 10.383828163146973, "kl": 47.125, "learning_rate": 8.205650235642829e-07, "loss": 0.8886, "num_tokens": 419197386.0, "reward": 0.09875000268220901, "reward_std": 0.08475096523761749, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.09209241718053818, "step": 2149 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 4056.0, "completions/mean_length": 1225.800048828125, "completions/mean_terminated_length": 563.4461669921875, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.8339798293250582, "grad_norm": 6.9711222648620605, "kl": 46.25, "learning_rate": 8.191120038536576e-07, "loss": 0.9402, "num_tokens": 419341394.0, "reward": 0.1300000101327896, "reward_std": 0.12848694622516632, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.16146089136600494, "step": 2150 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3825.0, "completions/mean_length": 1115.112548828125, "completions/mean_terminated_length": 589.0735473632812, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.8343677269200931, "grad_norm": 6.520629405975342, "kl": 40.875, "learning_rate": 8.17662033278738e-07, "loss": 0.7654, "num_tokens": 419473907.0, "reward": 0.11749999970197678, "reward_std": 0.09828957170248032, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.12718968093395233, "step": 2151 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3585.0, "completions/mean_length": 1197.612548828125, "completions/mean_terminated_length": 686.1323852539062, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.834755624515128, "grad_norm": 4.7049407958984375, "kl": 44.875, "learning_rate": 8.162151141292243e-07, "loss": 0.8259, "num_tokens": 419617780.0, "reward": 0.18375001847743988, "reward_std": 0.21216335892677307, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.10500000417232513, "rewards/unicoder_reward_fn/std": 0.252531498670578, "step": 2152 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2581.0, "completions/mean_length": 1139.8250732421875, "completions/mean_terminated_length": 512.757568359375, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "epoch": 0.835143522110163, "grad_norm": 4.9883294105529785, "kl": 40.34375, "learning_rate": 8.147712486900009e-07, "loss": 0.8274, "num_tokens": 419768870.0, "reward": 0.14124999940395355, "reward_std": 0.1057014986872673, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.20174868404865265, "step": 2153 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1454.0, "completions/mean_length": 975.9625244140625, "completions/mean_terminated_length": 425.3676452636719, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.8355314197051978, "grad_norm": 4.2283854484558105, "kl": 40.40625, "learning_rate": 8.133304392411272e-07, "loss": 0.7404, "num_tokens": 419895651.0, "reward": 0.1275000125169754, "reward_std": 0.12366483360528946, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.04500000178813934, "rewards/unicoder_reward_fn/std": 0.1386188566684723, "step": 2154 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 4062.0, "completions/mean_length": 999.2999877929688, "completions/mean_terminated_length": 505.6231994628906, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.8359193173002327, "grad_norm": 5.533079624176025, "kl": 34.0625, "learning_rate": 8.118926880578379e-07, "loss": 0.6091, "num_tokens": 420021395.0, "reward": 0.15000000596046448, "reward_std": 0.1272527277469635, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.13214397430419922, "step": 2155 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1907.0, "completions/mean_length": 741.4625244140625, "completions/mean_terminated_length": 368.7361145019531, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.8363072148952676, "grad_norm": 2.633612632751465, "kl": 30.4375, "learning_rate": 8.104579974105391e-07, "loss": 0.5659, "num_tokens": 420120168.0, "reward": 0.16500000655651093, "reward_std": 0.15783238410949707, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.20524761080741882, "step": 2156 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3535.0, "completions/mean_length": 578.4874877929688, "completions/mean_terminated_length": 393.3552551269531, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.8366951124903026, "grad_norm": 4.631697177886963, "kl": 21.1875, "learning_rate": 8.090263695648025e-07, "loss": 0.2947, "num_tokens": 420216703.0, "reward": 0.29375001788139343, "reward_std": 0.24081765115261078, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.20749998092651367, "rewards/unicoder_reward_fn/std": 0.2782290279865265, "step": 2157 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 2901.0, "completions/mean_length": 617.7000122070312, "completions/mean_terminated_length": 385.8133544921875, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.8370830100853375, "grad_norm": 8.86310863494873, "kl": 23.234375, "learning_rate": 8.075978067813647e-07, "loss": 0.456, "num_tokens": 420324383.0, "reward": 0.20374999940395355, "reward_std": 0.1526297777891159, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.19815605878829956, "step": 2158 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 4032.0, "completions/mean_length": 954.9874877929688, "completions/mean_terminated_length": 506.27142333984375, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.8374709076803724, "grad_norm": 4.348371505737305, "kl": 38.53125, "learning_rate": 8.061723113161207e-07, "loss": 0.6843, "num_tokens": 420448014.0, "reward": 0.26250001788139343, "reward_std": 0.24346132576465607, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.31072595715522766, "step": 2159 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2359.0, "completions/mean_length": 820.8875122070312, "completions/mean_terminated_length": 405.7323913574219, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.8378588052754073, "grad_norm": 3.5103650093078613, "kl": 32.234375, "learning_rate": 8.047498854201223e-07, "loss": 0.587, "num_tokens": 420557669.0, "reward": 0.26750001311302185, "reward_std": 0.19016163051128387, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.2867429554462433, "step": 2160 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3070.0, "completions/mean_length": 1210.5374755859375, "completions/mean_terminated_length": 489.171875, "completions/min_length": 89.0, "completions/min_terminated_length": 89.0, "epoch": 0.8382467028704422, "grad_norm": 4.234157085418701, "kl": 47.1875, "learning_rate": 8.033305313395748e-07, "loss": 0.9954, "num_tokens": 420695928.0, "reward": 0.12125001102685928, "reward_std": 0.09365855157375336, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.04749999940395355, "rewards/unicoder_reward_fn/std": 0.12825508415699005, "step": 2161 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 2437.0, "completions/mean_length": 954.9125366210938, "completions/mean_terminated_length": 506.18572998046875, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.8386346004654771, "grad_norm": 5.4618120193481445, "kl": 33.5625, "learning_rate": 8.01914251315831e-07, "loss": 0.6764, "num_tokens": 420813673.0, "reward": 0.16124999523162842, "reward_std": 0.08591743558645248, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.14088328182697296, "step": 2162 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3231.0, "completions/mean_length": 1213.3375244140625, "completions/mean_terminated_length": 704.6323852539062, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.839022498060512, "grad_norm": 16.344804763793945, "kl": 54.8125, "learning_rate": 8.005010475853916e-07, "loss": 0.8553, "num_tokens": 420957436.0, "reward": 0.09000001102685928, "reward_std": 0.09233538806438446, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.027500003576278687, "rewards/unicoder_reward_fn/std": 0.10430602729320526, "step": 2163 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.275, "completions/max_length": 4096.0, "completions/max_terminated_length": 3342.0, "completions/mean_length": 1646.0750732421875, "completions/mean_terminated_length": 716.7930908203125, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.839410395655547, "grad_norm": 20.244901657104492, "kl": 69.375, "learning_rate": 7.990909223798976e-07, "loss": 1.1174, "num_tokens": 421129690.0, "reward": 0.1300000101327896, "reward_std": 0.18533717095851898, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.18718433380126953, "step": 2164 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2996.0, "completions/mean_length": 1147.0625, "completions/mean_terminated_length": 466.5384521484375, "completions/min_length": 93.0, "completions/min_terminated_length": 93.0, "epoch": 0.8397982932505819, "grad_norm": 3.060513734817505, "kl": 48.90625, "learning_rate": 7.976838779261285e-07, "loss": 0.9523, "num_tokens": 421270279.0, "reward": 0.20000000298023224, "reward_std": 0.1814897358417511, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.2572985291481018, "step": 2165 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.26249999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3335.0, "completions/mean_length": 1552.1500244140625, "completions/mean_terminated_length": 646.7118530273438, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.8401861908456167, "grad_norm": 18.204200744628906, "kl": 65.5, "learning_rate": 7.96279916446001e-07, "loss": 1.0372, "num_tokens": 421441795.0, "reward": 0.11124999821186066, "reward_std": 0.16595065593719482, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.1799437254667282, "step": 2166 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3613.0, "completions/mean_length": 1659.1624755859375, "completions/mean_terminated_length": 614.8035888671875, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.8405740884406516, "grad_norm": 23.087949752807617, "kl": 60.4375, "learning_rate": 7.948790401565609e-07, "loss": 0.9912, "num_tokens": 421622264.0, "reward": 0.07625000923871994, "reward_std": 0.0707007423043251, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.019999999552965164, "rewards/unicoder_reward_fn/std": 0.0817529484629631, "step": 2167 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3950.0, "completions/mean_length": 1357.175048828125, "completions/mean_terminated_length": 562.0322265625, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.8409619860356866, "grad_norm": 9.2931547164917, "kl": 52.625, "learning_rate": 7.93481251269983e-07, "loss": 0.9268, "num_tokens": 421775758.0, "reward": 0.1562500149011612, "reward_std": 0.17601311206817627, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.21054480969905853, "step": 2168 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3003.0, "completions/mean_length": 1713.7249755859375, "completions/mean_terminated_length": 692.7500610351562, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.8413498836307215, "grad_norm": 5.586559772491455, "kl": 57.375, "learning_rate": 7.920865519935673e-07, "loss": 1.2001, "num_tokens": 421972216.0, "reward": 0.07500000298023224, "reward_std": 0.06810330599546432, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.06953734904527664, "step": 2169 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.30000000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3330.0, "completions/mean_length": 1685.0125732421875, "completions/mean_terminated_length": 651.732177734375, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "epoch": 0.8417377812257564, "grad_norm": 10.405555725097656, "kl": 46.4375, "learning_rate": 7.906949445297343e-07, "loss": 0.8117, "num_tokens": 422155665.0, "reward": 0.08250000327825546, "reward_std": 0.08932297676801682, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.030000001192092896, "rewards/unicoder_reward_fn/std": 0.09599578380584717, "step": 2170 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3448.0, "completions/mean_length": 1141.75, "completions/mean_terminated_length": 670.7825927734375, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.8421256788207913, "grad_norm": 2.668062686920166, "kl": 32.5625, "learning_rate": 7.893064310760214e-07, "loss": 0.6348, "num_tokens": 422296165.0, "reward": 0.1237499937415123, "reward_std": 0.09835522621870041, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.11462653428316116, "step": 2171 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3310.0, "completions/mean_length": 1102.5999755859375, "completions/mean_terminated_length": 625.3912963867188, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.8425135764158262, "grad_norm": 6.948172569274902, "kl": 34.25, "learning_rate": 7.879210138250818e-07, "loss": 0.6969, "num_tokens": 422437877.0, "reward": 0.14875000715255737, "reward_std": 0.1633489578962326, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.2046918123960495, "step": 2172 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3718.0, "completions/mean_length": 884.5499877929688, "completions/mean_terminated_length": 670.453369140625, "completions/min_length": 89.0, "completions/min_terminated_length": 89.0, "epoch": 0.8429014740108611, "grad_norm": 7.9406657218933105, "kl": 26.65625, "learning_rate": 7.865386949646781e-07, "loss": 0.5815, "num_tokens": 422562561.0, "reward": 0.17374999821186066, "reward_std": 0.1645432710647583, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.19614964723587036, "step": 2173 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2613.0, "completions/mean_length": 1020.2125244140625, "completions/mean_terminated_length": 423.41790771484375, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.843289371605896, "grad_norm": 9.741548538208008, "kl": 32.125, "learning_rate": 7.851594766776803e-07, "loss": 0.6664, "num_tokens": 422693882.0, "reward": 0.20125000178813934, "reward_std": 0.22404156625270844, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.22809059917926788, "step": 2174 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2857.0, "completions/mean_length": 964.8125, "completions/mean_terminated_length": 465.6376953125, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.843677269200931, "grad_norm": 12.863410949707031, "kl": 30.65625, "learning_rate": 7.837833611420618e-07, "loss": 0.7044, "num_tokens": 422813627.0, "reward": 0.1900000125169754, "reward_std": 0.1866304725408554, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.22432740032672882, "step": 2175 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3769.0, "completions/mean_length": 839.6749877929688, "completions/mean_terminated_length": 527.4246826171875, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.8440651667959659, "grad_norm": 5.0568084716796875, "kl": 22.921875, "learning_rate": 7.824103505308983e-07, "loss": 0.4546, "num_tokens": 422924345.0, "reward": 0.1300000101327896, "reward_std": 0.14035457372665405, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.20500075817108154, "step": 2176 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3852.0, "completions/mean_length": 972.0499877929688, "completions/mean_terminated_length": 624.9444580078125, "completions/min_length": 92.0, "completions/min_terminated_length": 92.0, "epoch": 0.8444530643910008, "grad_norm": 8.490194320678711, "kl": 27.78125, "learning_rate": 7.810404470123599e-07, "loss": 0.5906, "num_tokens": 423043701.0, "reward": 0.19875000417232513, "reward_std": 0.20639614760875702, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.2517784833908081, "step": 2177 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3171.0, "completions/mean_length": 706.2374877929688, "completions/mean_terminated_length": 527.8289794921875, "completions/min_length": 70.0, "completions/min_terminated_length": 70.0, "epoch": 0.8448409619860356, "grad_norm": 6.10894250869751, "kl": 16.5390625, "learning_rate": 7.796736527497109e-07, "loss": 0.3084, "num_tokens": 423147848.0, "reward": 0.1900000125169754, "reward_std": 0.13910938799381256, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932424664497375, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.1799437254667282, "step": 2178 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3794.0, "completions/mean_length": 706.5625, "completions/mean_terminated_length": 381.5479431152344, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.8452288595810706, "grad_norm": 4.272561073303223, "kl": 25.21875, "learning_rate": 7.783099699013075e-07, "loss": 0.4661, "num_tokens": 423254397.0, "reward": 0.24500003457069397, "reward_std": 0.1514124870300293, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.1599999964237213, "rewards/unicoder_reward_fn/std": 0.2631840109825134, "step": 2179 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1749.0, "completions/mean_length": 700.6124877929688, "completions/mean_terminated_length": 375.02740478515625, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.8456167571761055, "grad_norm": 6.143833637237549, "kl": 24.3125, "learning_rate": 7.769494006205894e-07, "loss": 0.4447, "num_tokens": 423363414.0, "reward": 0.24750001728534698, "reward_std": 0.18519429862499237, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.2580844461917877, "step": 2180 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2644.0, "completions/mean_length": 816.75, "completions/mean_terminated_length": 401.0704040527344, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.8460046547711404, "grad_norm": 3.976184368133545, "kl": 37.25, "learning_rate": 7.755919470560824e-07, "loss": 0.6654, "num_tokens": 423472658.0, "reward": 0.20125003159046173, "reward_std": 0.1775720864534378, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.17018233239650726, "step": 2181 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.1875, "completions/max_length": 4096.0, "completions/max_terminated_length": 2473.0, "completions/mean_length": 1108.7125244140625, "completions/mean_terminated_length": 419.3384704589844, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.8463925523661753, "grad_norm": 3.896690607070923, "kl": 44.9375, "learning_rate": 7.742376113513908e-07, "loss": 0.7313, "num_tokens": 423610131.0, "reward": 0.2525000274181366, "reward_std": 0.24689900875091553, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.28139570355415344, "step": 2182 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3507.0, "completions/mean_length": 980.2125244140625, "completions/mean_terminated_length": 535.0999755859375, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.8467804499612103, "grad_norm": 4.2443623542785645, "kl": 42.0, "learning_rate": 7.728863956451945e-07, "loss": 0.756, "num_tokens": 423731876.0, "reward": 0.19499999284744263, "reward_std": 0.19193756580352783, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.11749999225139618, "rewards/unicoder_reward_fn/std": 0.19794833660125732, "step": 2183 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3391.0, "completions/mean_length": 1065.050048828125, "completions/mean_terminated_length": 680.8450317382812, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.8471683475562451, "grad_norm": 3.1922197341918945, "kl": 48.9375, "learning_rate": 7.715383020712489e-07, "loss": 0.9451, "num_tokens": 423868000.0, "reward": 0.14125001430511475, "reward_std": 0.15564163029193878, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.1745881289243698, "step": 2184 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3469.0, "completions/mean_length": 1046.3875732421875, "completions/mean_terminated_length": 560.2174072265625, "completions/min_length": 99.0, "completions/min_terminated_length": 99.0, "epoch": 0.84755624515128, "grad_norm": 7.109743118286133, "kl": 44.25, "learning_rate": 7.701933327583766e-07, "loss": 0.8188, "num_tokens": 424000639.0, "reward": 0.13125000894069672, "reward_std": 0.12200471013784409, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.14574122428894043, "step": 2185 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3448.0, "completions/mean_length": 1165.4124755859375, "completions/mean_terminated_length": 596.791015625, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.847944142746315, "grad_norm": 8.764766693115234, "kl": 48.5625, "learning_rate": 7.688514898304689e-07, "loss": 0.9001, "num_tokens": 424136480.0, "reward": 0.0949999988079071, "reward_std": 0.07446636259555817, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.022499999031424522, "rewards/unicoder_reward_fn/std": 0.06359484046697617, "step": 2186 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3763.0, "completions/mean_length": 1082.9000244140625, "completions/mean_terminated_length": 551.1764526367188, "completions/min_length": 77.0, "completions/min_terminated_length": 77.0, "epoch": 0.8483320403413499, "grad_norm": 8.123834609985352, "kl": 48.875, "learning_rate": 7.675127754064782e-07, "loss": 0.8136, "num_tokens": 424275072.0, "reward": 0.15000000596046448, "reward_std": 0.14992491900920868, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.17572635412216187, "step": 2187 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1824.0, "completions/mean_length": 1192.5374755859375, "completions/mean_terminated_length": 466.671875, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.8487199379363848, "grad_norm": 9.406347274780273, "kl": 47.875, "learning_rate": 7.661771916004171e-07, "loss": 0.7483, "num_tokens": 424419563.0, "reward": 0.15250001847743988, "reward_std": 0.165935680270195, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.17915059626102448, "step": 2188 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3522.0, "completions/mean_length": 1335.5, "completions/mean_terminated_length": 645.375, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.8491078355314197, "grad_norm": 3.4014925956726074, "kl": 49.1875, "learning_rate": 7.648447405213555e-07, "loss": 0.9552, "num_tokens": 424575547.0, "reward": 0.1287500113248825, "reward_std": 0.11975838989019394, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.442768931388855, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.134916752576828, "step": 2189 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2922.0, "completions/mean_length": 849.1000366210938, "completions/mean_terminated_length": 585.8378295898438, "completions/min_length": 28.0, "completions/min_terminated_length": 28.0, "epoch": 0.8494957331264547, "grad_norm": 7.356469631195068, "kl": 27.4375, "learning_rate": 7.635154242734146e-07, "loss": 0.5668, "num_tokens": 424691835.0, "reward": 0.17499999701976776, "reward_std": 0.14785689115524292, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.177107572555542, "step": 2190 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 4059.0, "completions/mean_length": 900.5499877929688, "completions/mean_terminated_length": 594.136962890625, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "epoch": 0.8498836307214895, "grad_norm": 5.513930320739746, "kl": 34.96875, "learning_rate": 7.621892449557671e-07, "loss": 0.7034, "num_tokens": 424817279.0, "reward": 0.17625001072883606, "reward_std": 0.16765575110912323, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.22501757740974426, "step": 2191 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 1862.0, "completions/mean_length": 837.4500122070312, "completions/mean_terminated_length": 475.3888854980469, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.8502715283165244, "grad_norm": 15.989855766296387, "kl": 24.828125, "learning_rate": 7.608662046626308e-07, "loss": 0.5281, "num_tokens": 424931667.0, "reward": 0.27125000953674316, "reward_std": 0.2313869297504425, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.1899999976158142, "rewards/unicoder_reward_fn/std": 0.31000202894210815, "step": 2192 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2455.0, "completions/mean_length": 936.125, "completions/mean_terminated_length": 535.5774536132812, "completions/min_length": 78.0, "completions/min_terminated_length": 78.0, "epoch": 0.8506594259115593, "grad_norm": 5.216641902923584, "kl": 35.34375, "learning_rate": 7.595463054832665e-07, "loss": 0.7153, "num_tokens": 425056533.0, "reward": 0.23000001907348633, "reward_std": 0.24663515388965607, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.28859108686447144, "step": 2193 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3667.0, "completions/mean_length": 1065.862548828125, "completions/mean_terminated_length": 582.797119140625, "completions/min_length": 123.0, "completions/min_terminated_length": 123.0, "epoch": 0.8510473235065943, "grad_norm": 3.7454211711883545, "kl": 41.875, "learning_rate": 7.58229549501976e-07, "loss": 0.7721, "num_tokens": 425186890.0, "reward": 0.20999999344348907, "reward_std": 0.27631786465644836, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.14000000059604645, "rewards/unicoder_reward_fn/std": 0.26510089635849, "step": 2194 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 1843.0, "completions/mean_length": 999.1875, "completions/mean_terminated_length": 452.6911926269531, "completions/min_length": 163.0, "completions/min_terminated_length": 163.0, "epoch": 0.8514352211016292, "grad_norm": 4.854992866516113, "kl": 38.875, "learning_rate": 7.569159387980958e-07, "loss": 0.7335, "num_tokens": 425313761.0, "reward": 0.17500001192092896, "reward_std": 0.19691717624664307, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.10249999910593033, "rewards/unicoder_reward_fn/std": 0.20621666312217712, "step": 2195 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.16249999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3074.0, "completions/mean_length": 1177.1875, "completions/mean_terminated_length": 610.8507080078125, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.851823118696664, "grad_norm": 2.577587842941284, "kl": 44.375, "learning_rate": 7.556054754459975e-07, "loss": 0.9201, "num_tokens": 425455160.0, "reward": 0.13500000536441803, "reward_std": 0.15551166236400604, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.16821852326393127, "step": 2196 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3214.0, "completions/mean_length": 1145.425048828125, "completions/mean_terminated_length": 675.0435180664062, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.852211016291699, "grad_norm": 9.164984703063965, "kl": 41.6875, "learning_rate": 7.542981615150812e-07, "loss": 0.77, "num_tokens": 425589090.0, "reward": 0.10625000298023224, "reward_std": 0.10480234771966934, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.11648871749639511, "step": 2197 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3749.0, "completions/mean_length": 1080.8875732421875, "completions/mean_terminated_length": 650.1571655273438, "completions/min_length": 43.0, "completions/min_terminated_length": 43.0, "epoch": 0.8525989138867339, "grad_norm": 10.730462074279785, "kl": 44.5625, "learning_rate": 7.529939990697732e-07, "loss": 0.8873, "num_tokens": 425726305.0, "reward": 0.10500001162290573, "reward_std": 0.11184868961572647, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.16866940259933472, "step": 2198 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3301.0, "completions/mean_length": 965.7875366210938, "completions/mean_terminated_length": 665.630126953125, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "epoch": 0.8529868114817688, "grad_norm": 5.057347774505615, "kl": 39.375, "learning_rate": 7.51692990169525e-07, "loss": 0.7798, "num_tokens": 425849536.0, "reward": 0.14625000953674316, "reward_std": 0.16681618988513947, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.18384088575839996, "step": 2199 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 4055.0, "completions/mean_length": 1045.125, "completions/mean_terminated_length": 706.138916015625, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.8533747090768037, "grad_norm": 3.9820971488952637, "kl": 39.28125, "learning_rate": 7.503951368688069e-07, "loss": 0.823, "num_tokens": 425974122.0, "reward": 0.1587500125169754, "reward_std": 0.19891731441020966, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.2226281464099884, "step": 2200 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3016.0, "completions/mean_length": 781.7250366210938, "completions/mean_terminated_length": 513.0, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.8537626066718387, "grad_norm": 7.681419849395752, "kl": 19.890625, "learning_rate": 7.491004412171067e-07, "loss": 0.3391, "num_tokens": 426067892.0, "reward": 0.17499999701976776, "reward_std": 0.1439046710729599, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.18540070950984955, "step": 2201 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3357.0, "completions/mean_length": 1097.9124755859375, "completions/mean_terminated_length": 568.8382568359375, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.8541505042668736, "grad_norm": 4.337605953216553, "kl": 45.09375, "learning_rate": 7.478089052589256e-07, "loss": 0.9291, "num_tokens": 426201205.0, "reward": 0.17499999701976776, "reward_std": 0.15341876447200775, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.2225286215543747, "step": 2202 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 2228.0, "completions/mean_length": 623.4125366210938, "completions/mean_terminated_length": 440.6447448730469, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.8545384018619084, "grad_norm": 5.939497947692871, "kl": 25.40625, "learning_rate": 7.465205310337741e-07, "loss": 0.5038, "num_tokens": 426294246.0, "reward": 0.23000001907348633, "reward_std": 0.19127725064754486, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.14499999582767487, "rewards/unicoder_reward_fn/std": 0.26428794860839844, "step": 2203 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3964.0, "completions/mean_length": 1119.875, "completions/mean_terminated_length": 694.7142944335938, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.8549262994569433, "grad_norm": 11.163219451904297, "kl": 37.875, "learning_rate": 7.452353205761725e-07, "loss": 0.6484, "num_tokens": 426432196.0, "reward": 0.14625000953674316, "reward_std": 0.14632193744182587, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.17208696901798248, "step": 2204 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 1277.0, "completions/mean_length": 689.1000366210938, "completions/mean_terminated_length": 412.8648681640625, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.8553141970519783, "grad_norm": 2.1618523597717285, "kl": 26.46875, "learning_rate": 7.439532759156423e-07, "loss": 0.4452, "num_tokens": 426533972.0, "reward": 0.23625002801418304, "reward_std": 0.18383701145648956, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.26809218525886536, "step": 2205 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3656.0, "completions/mean_length": 848.0375366210938, "completions/mean_terminated_length": 536.5890502929688, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.8557020946470132, "grad_norm": 6.932923793792725, "kl": 33.46875, "learning_rate": 7.426743990767081e-07, "loss": 0.6271, "num_tokens": 426648703.0, "reward": 0.19875000417232513, "reward_std": 0.17240548133850098, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.1224999874830246, "rewards/unicoder_reward_fn/std": 0.20744070410728455, "step": 2206 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 2240.0, "completions/mean_length": 814.4750366210938, "completions/mean_terminated_length": 398.5070495605469, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "epoch": 0.8560899922420481, "grad_norm": 9.411273002624512, "kl": 31.40625, "learning_rate": 7.413986920788901e-07, "loss": 0.6716, "num_tokens": 426758125.0, "reward": 0.2787500023841858, "reward_std": 0.18659405410289764, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.2828427255153656, "step": 2207 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3732.0, "completions/mean_length": 836.2750244140625, "completions/mean_terminated_length": 523.6986083984375, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "epoch": 0.8564778898370831, "grad_norm": 3.8371121883392334, "kl": 39.0, "learning_rate": 7.401261569367039e-07, "loss": 0.6688, "num_tokens": 426868955.0, "reward": 0.13625000417232513, "reward_std": 0.11682917922735214, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.13941560685634613, "step": 2208 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3348.0, "completions/mean_length": 718.8500366210938, "completions/mean_terminated_length": 493.7066650390625, "completions/min_length": 131.0, "completions/min_terminated_length": 131.0, "epoch": 0.8568657874321179, "grad_norm": 4.994949817657471, "kl": 34.03125, "learning_rate": 7.388567956596567e-07, "loss": 0.4424, "num_tokens": 426963863.0, "reward": 0.257500022649765, "reward_std": 0.27195942401885986, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.31917616724967957, "step": 2209 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3813.0, "completions/mean_length": 1117.0250244140625, "completions/mean_terminated_length": 591.3235473632812, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.8572536850271528, "grad_norm": 4.435609817504883, "kl": 48.25, "learning_rate": 7.375906102522424e-07, "loss": 0.8146, "num_tokens": 427095209.0, "reward": 0.17500001192092896, "reward_std": 0.16869807243347168, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.10500000417232513, "rewards/unicoder_reward_fn/std": 0.19086743891239166, "step": 2210 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3963.0, "completions/mean_length": 1061.4124755859375, "completions/mean_terminated_length": 525.8970336914062, "completions/min_length": 90.0, "completions/min_terminated_length": 90.0, "epoch": 0.8576415826221877, "grad_norm": 5.489527702331543, "kl": 46.25, "learning_rate": 7.363276027139416e-07, "loss": 0.7994, "num_tokens": 427226258.0, "reward": 0.23749999701976776, "reward_std": 0.2534277141094208, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.29008948802948, "step": 2211 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.19999999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3671.0, "completions/mean_length": 1391.7750244140625, "completions/mean_terminated_length": 715.71875, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.8580294802172227, "grad_norm": 7.912156581878662, "kl": 56.6875, "learning_rate": 7.350677750392151e-07, "loss": 0.9743, "num_tokens": 427392704.0, "reward": 0.19875000417232513, "reward_std": 0.21454660594463348, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.14000000059604645, "rewards/unicoder_reward_fn/std": 0.23035472631454468, "step": 2212 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3863.0, "completions/mean_length": 1118.8875732421875, "completions/mean_terminated_length": 741.5070190429688, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.8584173778122576, "grad_norm": 9.601451873779297, "kl": 50.5625, "learning_rate": 7.338111292175024e-07, "loss": 0.8333, "num_tokens": 427532383.0, "reward": 0.20000000298023224, "reward_std": 0.21132878959178925, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.23512722551822662, "step": 2213 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 2624.0, "completions/mean_length": 710.7750244140625, "completions/mean_terminated_length": 532.6052856445312, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.8588052754072925, "grad_norm": 8.292153358459473, "kl": 30.0625, "learning_rate": 7.325576672332193e-07, "loss": 0.6542, "num_tokens": 427639725.0, "reward": 0.1850000023841858, "reward_std": 0.1629026234149933, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.10500000417232513, "rewards/unicoder_reward_fn/std": 0.1935020387172699, "step": 2214 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 2866.0, "completions/mean_length": 766.7374877929688, "completions/mean_terminated_length": 447.4931335449219, "completions/min_length": 58.0, "completions/min_terminated_length": 58.0, "epoch": 0.8591931730023273, "grad_norm": 5.554002285003662, "kl": 38.125, "learning_rate": 7.313073910657535e-07, "loss": 0.761, "num_tokens": 427753216.0, "reward": 0.1875, "reward_std": 0.20288227498531342, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.22675475478172302, "step": 2215 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3987.0, "completions/mean_length": 798.3125, "completions/mean_terminated_length": 530.9324340820312, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.8595810705973623, "grad_norm": 7.48199987411499, "kl": 35.78125, "learning_rate": 7.300603026894607e-07, "loss": 0.7269, "num_tokens": 427865305.0, "reward": 0.2550000250339508, "reward_std": 0.2797822058200836, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.34005770087242126, "step": 2216 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3049.0, "completions/mean_length": 900.6625366210938, "completions/mean_terminated_length": 594.26025390625, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.8599689681923972, "grad_norm": 3.947054147720337, "kl": 34.59375, "learning_rate": 7.288164040736652e-07, "loss": 0.6332, "num_tokens": 427996742.0, "reward": 0.23499999940395355, "reward_std": 0.19314806163311005, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.1600000113248825, "rewards/unicoder_reward_fn/std": 0.236857071518898, "step": 2217 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 4066.0, "completions/mean_length": 964.8125, "completions/mean_terminated_length": 842.8181762695312, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "epoch": 0.8603568657874321, "grad_norm": 3.744572639465332, "kl": 45.75, "learning_rate": 7.275756971826516e-07, "loss": 0.8254, "num_tokens": 428132335.0, "reward": 0.13750000298023224, "reward_std": 0.14816944301128387, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.1395970731973648, "step": 2218 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3896.0, "completions/mean_length": 846.4375, "completions/mean_terminated_length": 534.8356323242188, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.8607447633824671, "grad_norm": 2.024371862411499, "kl": 39.15625, "learning_rate": 7.263381839756651e-07, "loss": 0.7175, "num_tokens": 428248474.0, "reward": 0.1925000250339508, "reward_std": 0.1978348344564438, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.12250000238418579, "rewards/unicoder_reward_fn/std": 0.20986735820770264, "step": 2219 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3432.0, "completions/mean_length": 876.2999877929688, "completions/mean_terminated_length": 661.6533203125, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "epoch": 0.861132660977502, "grad_norm": 3.843626022338867, "kl": 38.5625, "learning_rate": 7.251038664069093e-07, "loss": 0.6924, "num_tokens": 428370594.0, "reward": 0.18125000596046448, "reward_std": 0.18547005951404572, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.2201121747493744, "step": 2220 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3946.0, "completions/mean_length": 1082.7249755859375, "completions/mean_terminated_length": 793.7808227539062, "completions/min_length": 41.0, "completions/min_terminated_length": 41.0, "epoch": 0.8615205585725368, "grad_norm": 12.556710243225098, "kl": 54.5, "learning_rate": 7.238727464255394e-07, "loss": 0.973, "num_tokens": 428496972.0, "reward": 0.11625000089406967, "reward_std": 0.16064520180225372, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.1671617180109024, "step": 2221 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3294.0, "completions/mean_length": 911.1375122070312, "completions/mean_terminated_length": 652.9053955078125, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.8619084561675717, "grad_norm": 14.789372444152832, "kl": 46.25, "learning_rate": 7.226448259756618e-07, "loss": 0.7309, "num_tokens": 428605983.0, "reward": 0.10625000298023224, "reward_std": 0.13229425251483917, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.17574435472488403, "step": 2222 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 4015.0, "completions/mean_length": 1114.8499755859375, "completions/mean_terminated_length": 957.9473876953125, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.8622963537626067, "grad_norm": 26.29814338684082, "kl": 53.875, "learning_rate": 7.214201069963303e-07, "loss": 0.7369, "num_tokens": 428733491.0, "reward": 0.08500000089406967, "reward_std": 0.13227711617946625, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.15342091023921967, "step": 2223 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.15000000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 3708.0, "completions/mean_length": 1396.612548828125, "completions/mean_terminated_length": 920.25, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.8626842513576416, "grad_norm": 24.14809226989746, "kl": 56.375, "learning_rate": 7.201985914215439e-07, "loss": 0.851, "num_tokens": 428900204.0, "reward": 0.0625, "reward_std": 0.10571872442960739, "rewards/format_reward/mean": 0.30000001192092896, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.032499998807907104, "rewards/unicoder_reward_fn/std": 0.12505695223808289, "step": 2224 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3707.0, "completions/mean_length": 1024.1624755859375, "completions/mean_terminated_length": 945.3974609375, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.8630721489526765, "grad_norm": 13.96992015838623, "kl": 48.8125, "learning_rate": 7.18980281180242e-07, "loss": 0.7912, "num_tokens": 429031201.0, "reward": 0.14875002205371857, "reward_std": 0.23793397843837738, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.27369946241378784, "step": 2225 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3984.0, "completions/mean_length": 940.375, "completions/mean_terminated_length": 684.5135498046875, "completions/min_length": 56.0, "completions/min_terminated_length": 56.0, "epoch": 0.8634600465477114, "grad_norm": 22.556028366088867, "kl": 47.6875, "learning_rate": 7.177651781963033e-07, "loss": 0.7307, "num_tokens": 429148727.0, "reward": 0.07500001043081284, "reward_std": 0.12463817745447159, "rewards/format_reward/mean": 0.4000000059604645, "rewards/format_reward/std": 0.49298882484436035, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.1303354948759079, "step": 2226 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.07499999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 3643.0, "completions/mean_length": 1064.237548828125, "completions/mean_terminated_length": 818.4189453125, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.8638479441427463, "grad_norm": 6.355398654937744, "kl": 38.125, "learning_rate": 7.165532843885408e-07, "loss": 0.7044, "num_tokens": 429285898.0, "reward": 0.10125000774860382, "reward_std": 0.10474123060703278, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.12010017037391663, "step": 2227 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3611.0, "completions/mean_length": 933.2625122070312, "completions/mean_terminated_length": 722.413330078125, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.8642358417377812, "grad_norm": 8.020662307739258, "kl": 36.0625, "learning_rate": 7.153446016706999e-07, "loss": 0.6317, "num_tokens": 429406815.0, "reward": 0.09875000268220901, "reward_std": 0.12755881249904633, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.1606946885585785, "step": 2228 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3922.0, "completions/mean_length": 750.6375122070312, "completions/mean_terminated_length": 708.2911376953125, "completions/min_length": 32.0, "completions/min_terminated_length": 32.0, "epoch": 0.8646237393328161, "grad_norm": 6.981775760650635, "kl": 32.71875, "learning_rate": 7.141391319514565e-07, "loss": 0.4858, "num_tokens": 429512226.0, "reward": 0.10500001162290573, "reward_std": 0.17534731328487396, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.16098982095718384, "step": 2229 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1915.0, "completions/max_terminated_length": 1915.0, "completions/mean_length": 465.375, "completions/mean_terminated_length": 465.375, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.8650116369278511, "grad_norm": 9.889999389648438, "kl": 15.765625, "learning_rate": 7.129368771344108e-07, "loss": 0.2746, "num_tokens": 429593200.0, "reward": 0.2525000274181366, "reward_std": 0.19265785813331604, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.2707233726978302, "step": 2230 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1567.0, "completions/max_terminated_length": 1567.0, "completions/mean_length": 356.20001220703125, "completions/mean_terminated_length": 356.20001220703125, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.865399534522886, "grad_norm": 9.922170639038086, "kl": 8.2265625, "learning_rate": 7.117378391180885e-07, "loss": 0.1714, "num_tokens": 429661704.0, "reward": 0.351250022649765, "reward_std": 0.19753162562847137, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.26249998807907104, "rewards/unicoder_reward_fn/std": 0.3223234713077545, "step": 2231 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2161.0, "completions/max_terminated_length": 2161.0, "completions/mean_length": 414.1625061035156, "completions/mean_terminated_length": 414.1625061035156, "completions/min_length": 37.0, "completions/min_terminated_length": 37.0, "epoch": 0.8657874321179209, "grad_norm": 8.274140357971191, "kl": 10.3125, "learning_rate": 7.105420197959335e-07, "loss": 0.1435, "num_tokens": 429747373.0, "reward": 0.21250000596046448, "reward_std": 0.12050973623991013, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.19342024624347687, "step": 2232 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2575.0, "completions/max_terminated_length": 2575.0, "completions/mean_length": 443.7124938964844, "completions/mean_terminated_length": 443.7124938964844, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.8661753297129557, "grad_norm": 27.376205444335938, "kl": 11.140625, "learning_rate": 7.093494210563075e-07, "loss": 0.4008, "num_tokens": 429832118.0, "reward": 0.2787500023841858, "reward_std": 0.171135812997818, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.19749999046325684, "rewards/unicoder_reward_fn/std": 0.2370573878288269, "step": 2233 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3710.0, "completions/max_terminated_length": 3710.0, "completions/mean_length": 432.76251220703125, "completions/mean_terminated_length": 432.76251220703125, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "epoch": 0.8665632273079907, "grad_norm": 15.538558006286621, "kl": 10.01171875, "learning_rate": 7.081600447824871e-07, "loss": 0.2836, "num_tokens": 429921563.0, "reward": 0.27250000834465027, "reward_std": 0.18007388710975647, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932424664497375, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.2870076894760132, "step": 2234 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3251.0, "completions/mean_length": 515.6375122070312, "completions/mean_terminated_length": 470.31646728515625, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.8669511249030256, "grad_norm": 16.851566314697266, "kl": 11.765625, "learning_rate": 7.069738928526598e-07, "loss": 0.3039, "num_tokens": 430007022.0, "reward": 0.3162499964237213, "reward_std": 0.18906955420970917, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.23250000178813934, "rewards/unicoder_reward_fn/std": 0.24534913897514343, "step": 2235 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1937.0, "completions/max_terminated_length": 1937.0, "completions/mean_length": 351.2250061035156, "completions/mean_terminated_length": 351.2250061035156, "completions/min_length": 99.0, "completions/min_terminated_length": 99.0, "epoch": 0.8673390224980605, "grad_norm": 10.878514289855957, "kl": 6.078125, "learning_rate": 7.057909671399216e-07, "loss": 0.1577, "num_tokens": 430083080.0, "reward": 0.26750001311302185, "reward_std": 0.1425681710243225, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.28638955950737, "step": 2236 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1706.0, "completions/max_terminated_length": 1706.0, "completions/mean_length": 371.8374938964844, "completions/mean_terminated_length": 371.8374938964844, "completions/min_length": 64.0, "completions/min_terminated_length": 64.0, "epoch": 0.8677269200930954, "grad_norm": 16.25736427307129, "kl": 8.34375, "learning_rate": 7.046112695122731e-07, "loss": 0.2147, "num_tokens": 430155067.0, "reward": 0.3487499952316284, "reward_std": 0.1474233865737915, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.26249998807907104, "rewards/unicoder_reward_fn/std": 0.3127562403678894, "step": 2237 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2146.0, "completions/mean_length": 406.1499938964844, "completions/mean_terminated_length": 359.44305419921875, "completions/min_length": 93.0, "completions/min_terminated_length": 93.0, "epoch": 0.8681148176881304, "grad_norm": 8.820608139038086, "kl": 12.181640625, "learning_rate": 7.034348018326182e-07, "loss": 0.289, "num_tokens": 430236559.0, "reward": 0.3075000047683716, "reward_std": 0.1345175951719284, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.2199999988079071, "rewards/unicoder_reward_fn/std": 0.29784032702445984, "step": 2238 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1485.0, "completions/max_terminated_length": 1485.0, "completions/mean_length": 400.0375061035156, "completions/mean_terminated_length": 400.0375061035156, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.8685027152831652, "grad_norm": 15.090910911560059, "kl": 8.421875, "learning_rate": 7.022615659587599e-07, "loss": 0.1658, "num_tokens": 430315386.0, "reward": 0.288750022649765, "reward_std": 0.20744581520557404, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.19749999046325684, "rewards/unicoder_reward_fn/std": 0.29251202940940857, "step": 2239 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2718.0, "completions/max_terminated_length": 2718.0, "completions/mean_length": 455.3125, "completions/mean_terminated_length": 455.3125, "completions/min_length": 75.0, "completions/min_terminated_length": 75.0, "epoch": 0.8688906128782001, "grad_norm": 9.721830368041992, "kl": 17.59375, "learning_rate": 7.010915637433981e-07, "loss": 0.3758, "num_tokens": 430407811.0, "reward": 0.2874999940395355, "reward_std": 0.16522736847400665, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.20499999821186066, "rewards/unicoder_reward_fn/std": 0.248489111661911, "step": 2240 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2139.0, "completions/mean_length": 421.20001220703125, "completions/mean_terminated_length": 374.6835632324219, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.869278510473235, "grad_norm": 15.296956062316895, "kl": 11.75, "learning_rate": 6.999247970341253e-07, "loss": 0.2581, "num_tokens": 430488475.0, "reward": 0.29749998450279236, "reward_std": 0.12342055141925812, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.2514640688896179, "step": 2241 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3318.0, "completions/mean_length": 657.875, "completions/mean_terminated_length": 523.9220581054688, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.86966640806827, "grad_norm": 11.86921215057373, "kl": 21.9375, "learning_rate": 6.987612676734259e-07, "loss": 0.4339, "num_tokens": 430585497.0, "reward": 0.24375000596046448, "reward_std": 0.1862855851650238, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.16499999165534973, "rewards/unicoder_reward_fn/std": 0.27517545223236084, "step": 2242 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3213.0, "completions/mean_length": 710.8500366210938, "completions/mean_terminated_length": 668.0, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.8700543056633049, "grad_norm": 3.629300355911255, "kl": 37.6875, "learning_rate": 6.976009774986714e-07, "loss": 0.6286, "num_tokens": 430686325.0, "reward": 0.16250000894069672, "reward_std": 0.1721961945295334, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.20867271721363068, "step": 2243 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3634.0, "completions/mean_length": 551.5875244140625, "completions/mean_terminated_length": 460.70513916015625, "completions/min_length": 125.0, "completions/min_terminated_length": 125.0, "epoch": 0.8704422032583398, "grad_norm": 10.655518531799316, "kl": 22.25, "learning_rate": 6.964439283421188e-07, "loss": 0.4814, "num_tokens": 430770924.0, "reward": 0.27625003457069397, "reward_std": 0.18885378539562225, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.1900000125169754, "rewards/unicoder_reward_fn/std": 0.2641441822052002, "step": 2244 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2586.0, "completions/mean_length": 570.5250244140625, "completions/mean_terminated_length": 480.1282043457031, "completions/min_length": 52.0, "completions/min_terminated_length": 52.0, "epoch": 0.8708301008533748, "grad_norm": 6.644959926605225, "kl": 38.4375, "learning_rate": 6.952901220309063e-07, "loss": 0.7025, "num_tokens": 430860446.0, "reward": 0.23749999701976776, "reward_std": 0.2177319973707199, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.22070084512233734, "step": 2245 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2935.0, "completions/mean_length": 632.0, "completions/mean_terminated_length": 543.1795043945312, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.8712179984484096, "grad_norm": 14.827471733093262, "kl": 50.0625, "learning_rate": 6.941395603870512e-07, "loss": 0.7468, "num_tokens": 430957918.0, "reward": 0.1912500113248825, "reward_std": 0.20933154225349426, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.13500002026557922, "rewards/unicoder_reward_fn/std": 0.23335744440555573, "step": 2246 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 4090.0, "completions/mean_length": 729.4625244140625, "completions/mean_terminated_length": 598.2987060546875, "completions/min_length": 77.0, "completions/min_terminated_length": 77.0, "epoch": 0.8716058960434445, "grad_norm": 2.849254846572876, "kl": 46.09375, "learning_rate": 6.929922452274479e-07, "loss": 0.8601, "num_tokens": 431065067.0, "reward": 0.21875, "reward_std": 0.1948016732931137, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.2134111076593399, "step": 2247 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2966.0, "completions/mean_length": 705.0875244140625, "completions/mean_terminated_length": 572.9739990234375, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.8719937936384794, "grad_norm": 2.8005237579345703, "kl": 43.0, "learning_rate": 6.918481783638633e-07, "loss": 0.7974, "num_tokens": 431174594.0, "reward": 0.1612500101327896, "reward_std": 0.13558074831962585, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.0925000011920929, "rewards/unicoder_reward_fn/std": 0.2011203020811081, "step": 2248 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.13749999999999996, "completions/max_length": 4096.0, "completions/max_terminated_length": 2846.0, "completions/mean_length": 1211.75, "completions/mean_terminated_length": 751.9420166015625, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.8723816912335144, "grad_norm": 23.461572647094727, "kl": 65.0625, "learning_rate": 6.907073616029357e-07, "loss": 0.9214, "num_tokens": 431322462.0, "reward": 0.10000000149011612, "reward_std": 0.16081757843494415, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.057499997317790985, "rewards/unicoder_reward_fn/std": 0.15652476251125336, "step": 2249 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 4067.0, "completions/mean_length": 1218.7249755859375, "completions/mean_terminated_length": 1026.90673828125, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.8727695888285493, "grad_norm": 35.86103820800781, "kl": 73.625, "learning_rate": 6.895697967461699e-07, "loss": 0.9297, "num_tokens": 431468744.0, "reward": 0.07750000804662704, "reward_std": 0.13799776136875153, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.1706094741821289, "step": 2250 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.17500000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3145.0, "completions/mean_length": 1471.0625, "completions/mean_terminated_length": 914.2576293945312, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.8731574864235842, "grad_norm": 27.45381736755371, "kl": 71.375, "learning_rate": 6.884354855899361e-07, "loss": 1.0079, "num_tokens": 431648245.0, "reward": 0.0637500062584877, "reward_std": 0.08081668615341187, "rewards/format_reward/mean": 0.38749998807907104, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.06656109541654587, "step": 2251 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.08750000000000002, "completions/max_length": 4096.0, "completions/max_terminated_length": 4002.0, "completions/mean_length": 1046.8499755859375, "completions/mean_terminated_length": 754.4657592773438, "completions/min_length": 57.0, "completions/min_terminated_length": 57.0, "epoch": 0.873545384018619, "grad_norm": 40.203983306884766, "kl": 63.0, "learning_rate": 6.873044299254668e-07, "loss": 0.7146, "num_tokens": 431784793.0, "reward": 0.05000000074505806, "reward_std": 0.0799492746591568, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.08015017956495285, "step": 2252 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 4089.0, "completions/mean_length": 1397.125, "completions/mean_terminated_length": 1097.25, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.873933281613654, "grad_norm": 30.65158462524414, "kl": 71.625, "learning_rate": 6.86176631538853e-07, "loss": 0.9979, "num_tokens": 431954283.0, "reward": 0.07750000804662704, "reward_std": 0.1614074558019638, "rewards/format_reward/mean": 0.32499998807907104, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.17422522604465485, "step": 2253 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.22499999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3827.0, "completions/mean_length": 1799.4625244140625, "completions/mean_terminated_length": 1132.725830078125, "completions/min_length": 51.0, "completions/min_terminated_length": 51.0, "epoch": 0.8743211792086889, "grad_norm": 36.933921813964844, "kl": 76.625, "learning_rate": 6.850520922110425e-07, "loss": 1.0396, "num_tokens": 432150456.0, "reward": 0.03875000402331352, "reward_std": 0.06547292321920395, "rewards/format_reward/mean": 0.23749999701976776, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.015000000596046448, "rewards/unicoder_reward_fn/std": 0.05301063135266304, "step": 2254 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.125, "completions/max_length": 4096.0, "completions/max_terminated_length": 3364.0, "completions/mean_length": 1294.800048828125, "completions/mean_terminated_length": 894.6286010742188, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.8747090768037238, "grad_norm": 33.26833724975586, "kl": 69.0625, "learning_rate": 6.839308137178363e-07, "loss": 0.9323, "num_tokens": 432303200.0, "reward": 0.03750000149011612, "reward_std": 0.07238657772541046, "rewards/format_reward/mean": 0.25, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.012500000186264515, "rewards/unicoder_reward_fn/std": 0.11180339753627777, "step": 2255 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 3830.0, "completions/mean_length": 1179.25, "completions/mean_terminated_length": 809.5211181640625, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.8750969743987588, "grad_norm": 30.898822784423828, "kl": 63.3125, "learning_rate": 6.828127978298859e-07, "loss": 0.9066, "num_tokens": 432438556.0, "reward": 0.08875000476837158, "reward_std": 0.13669292628765106, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.04500000178813934, "rewards/unicoder_reward_fn/std": 0.1771075576543808, "step": 2256 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.09999999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3783.0, "completions/mean_length": 1431.9375, "completions/mean_terminated_length": 1135.9305419921875, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.8754848719937937, "grad_norm": 29.630002975463867, "kl": 61.1875, "learning_rate": 6.816980463126917e-07, "loss": 0.9003, "num_tokens": 432604511.0, "reward": 0.0625, "reward_std": 0.08300554007291794, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.02500000037252903, "rewards/unicoder_reward_fn/std": 0.10250040143728256, "step": 2257 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3102.0, "completions/mean_length": 820.2875366210938, "completions/mean_terminated_length": 647.881591796875, "completions/min_length": 70.0, "completions/min_terminated_length": 70.0, "epoch": 0.8758727695888285, "grad_norm": 23.582319259643555, "kl": 51.6875, "learning_rate": 6.805865609265979e-07, "loss": 0.7048, "num_tokens": 432714934.0, "reward": 0.125, "reward_std": 0.15281625092029572, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.18169385194778442, "step": 2258 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3867.0, "completions/mean_length": 1020.0875244140625, "completions/mean_terminated_length": 815.0266723632812, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.8762606671838634, "grad_norm": 20.49862289428711, "kl": 53.0625, "learning_rate": 6.794783434267918e-07, "loss": 0.748, "num_tokens": 432841477.0, "reward": 0.08624999970197678, "reward_std": 0.15150046348571777, "rewards/format_reward/mean": 0.36250001192092896, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.16381525993347168, "step": 2259 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3862.0, "completions/mean_length": 865.5625, "completions/mean_terminated_length": 650.2000122070312, "completions/min_length": 57.0, "completions/min_terminated_length": 57.0, "epoch": 0.8766485647788984, "grad_norm": 13.47038459777832, "kl": 39.0625, "learning_rate": 6.783733955633004e-07, "loss": 0.5775, "num_tokens": 432960442.0, "reward": 0.11750000715255737, "reward_std": 0.14183960855007172, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.06499999761581421, "rewards/unicoder_reward_fn/std": 0.17364302277565002, "step": 2260 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2912.0, "completions/mean_length": 679.75, "completions/mean_terminated_length": 546.6493530273438, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.8770364623739333, "grad_norm": 4.024224758148193, "kl": 30.375, "learning_rate": 6.772717190809862e-07, "loss": 0.4636, "num_tokens": 433069686.0, "reward": 0.11500000953674316, "reward_std": 0.1334487944841385, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.16456693410873413, "step": 2261 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 2912.0, "completions/mean_length": 758.3500366210938, "completions/mean_terminated_length": 582.6842041015625, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.8774243599689682, "grad_norm": 4.192765235900879, "kl": 31.53125, "learning_rate": 6.761733157195481e-07, "loss": 0.5629, "num_tokens": 433183050.0, "reward": 0.18375001847743988, "reward_std": 0.21574907004833221, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.1274999976158142, "rewards/unicoder_reward_fn/std": 0.2704777717590332, "step": 2262 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2615.0, "completions/mean_length": 714.3250122070312, "completions/mean_terminated_length": 582.5714111328125, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.877812257564003, "grad_norm": 8.990415573120117, "kl": 28.5625, "learning_rate": 6.750781872135142e-07, "loss": 0.6016, "num_tokens": 433281084.0, "reward": 0.16625002026557922, "reward_std": 0.16376416385173798, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.21810461580753326, "step": 2263 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4015.0, "completions/max_terminated_length": 4015.0, "completions/mean_length": 587.4125366210938, "completions/mean_terminated_length": 587.4125366210938, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.878200155159038, "grad_norm": 11.294803619384766, "kl": 25.75, "learning_rate": 6.739863352922415e-07, "loss": 0.572, "num_tokens": 433371581.0, "reward": 0.1275000125169754, "reward_std": 0.14782188832759857, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.0624999925494194, "rewards/unicoder_reward_fn/std": 0.1871674358844757, "step": 2264 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2621.0, "completions/max_terminated_length": 2621.0, "completions/mean_length": 526.5125122070312, "completions/mean_terminated_length": 526.5125122070312, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.8785880527540729, "grad_norm": 17.729949951171875, "kl": 20.09375, "learning_rate": 6.728977616799144e-07, "loss": 0.4995, "num_tokens": 433457990.0, "reward": 0.16250000894069672, "reward_std": 0.1257992535829544, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.09499999135732651, "rewards/unicoder_reward_fn/std": 0.14222465455532074, "step": 2265 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3754.0, "completions/mean_length": 448.8625183105469, "completions/mean_terminated_length": 402.6961975097656, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.8789759503491078, "grad_norm": 18.09683609008789, "kl": 11.4453125, "learning_rate": 6.718124680955382e-07, "loss": 0.3161, "num_tokens": 433545691.0, "reward": 0.2537499964237213, "reward_std": 0.16280092298984528, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.2580844759941101, "step": 2266 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1977.0, "completions/max_terminated_length": 1977.0, "completions/mean_length": 364.9624938964844, "completions/mean_terminated_length": 364.9624938964844, "completions/min_length": 37.0, "completions/min_terminated_length": 37.0, "epoch": 0.8793638479441428, "grad_norm": 15.945184707641602, "kl": 8.1171875, "learning_rate": 6.707304562529396e-07, "loss": 0.1465, "num_tokens": 433619472.0, "reward": 0.3525000214576721, "reward_std": 0.20075754821300507, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.26750001311302185, "rewards/unicoder_reward_fn/std": 0.277317613363266, "step": 2267 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2873.0, "completions/mean_length": 450.0375061035156, "completions/mean_terminated_length": 403.8860778808594, "completions/min_length": 118.0, "completions/min_terminated_length": 118.0, "epoch": 0.8797517455391777, "grad_norm": 15.093119621276855, "kl": 11.53125, "learning_rate": 6.696517278607639e-07, "loss": 0.2708, "num_tokens": 433699275.0, "reward": 0.4000000059604645, "reward_std": 0.25069981813430786, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.3149999976158142, "rewards/unicoder_reward_fn/std": 0.2895216643810272, "step": 2268 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2568.0, "completions/mean_length": 611.2625122070312, "completions/mean_terminated_length": 475.4934997558594, "completions/min_length": 152.0, "completions/min_terminated_length": 152.0, "epoch": 0.8801396431342126, "grad_norm": 25.013248443603516, "kl": 20.578125, "learning_rate": 6.685762846224697e-07, "loss": 0.6135, "num_tokens": 433804056.0, "reward": 0.32375001907348633, "reward_std": 0.223231241106987, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.24500000476837158, "rewards/unicoder_reward_fn/std": 0.30684182047843933, "step": 2269 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2593.0, "completions/max_terminated_length": 2593.0, "completions/mean_length": 491.63751220703125, "completions/mean_terminated_length": 491.63751220703125, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.8805275407292474, "grad_norm": 14.71107292175293, "kl": 16.9375, "learning_rate": 6.675041282363286e-07, "loss": 0.3917, "num_tokens": 433889963.0, "reward": 0.19500002264976501, "reward_std": 0.133056640625, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.20584805309772491, "step": 2270 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2122.0, "completions/max_terminated_length": 2122.0, "completions/mean_length": 376.0500183105469, "completions/mean_terminated_length": 376.0500183105469, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.8809154383242824, "grad_norm": 1196.8973388671875, "kl": 92.9453125, "learning_rate": 6.664352603954226e-07, "loss": 1.05, "num_tokens": 433971311.0, "reward": 0.35625001788139343, "reward_std": 0.185844287276268, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.27000001072883606, "rewards/unicoder_reward_fn/std": 0.32154691219329834, "step": 2271 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1656.0, "completions/max_terminated_length": 1656.0, "completions/mean_length": 381.3999938964844, "completions/mean_terminated_length": 381.3999938964844, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "epoch": 0.8813033359193173, "grad_norm": 11.58834171295166, "kl": 12.3125, "learning_rate": 6.653696827876393e-07, "loss": 0.3028, "num_tokens": 434058143.0, "reward": 0.2912500202655792, "reward_std": 0.14164100587368011, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.20500001311302185, "rewards/unicoder_reward_fn/std": 0.24229910969734192, "step": 2272 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2209.0, "completions/mean_length": 497.9250183105469, "completions/mean_terminated_length": 405.66668701171875, "completions/min_length": 41.0, "completions/min_terminated_length": 41.0, "epoch": 0.8816912335143522, "grad_norm": 10.510783195495605, "kl": 18.9375, "learning_rate": 6.643073970956708e-07, "loss": 0.3458, "num_tokens": 434148801.0, "reward": 0.21250000596046448, "reward_std": 0.1409127563238144, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.13499999046325684, "rewards/unicoder_reward_fn/std": 0.17653486132621765, "step": 2273 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2369.0, "completions/max_terminated_length": 2369.0, "completions/mean_length": 372.5874938964844, "completions/mean_terminated_length": 372.5874938964844, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.8820791311093871, "grad_norm": 17.383134841918945, "kl": 14.09375, "learning_rate": 6.632484049970123e-07, "loss": 0.3096, "num_tokens": 434222096.0, "reward": 0.5250000357627869, "reward_std": 0.230287104845047, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.4350000023841858, "rewards/unicoder_reward_fn/std": 0.3349286615848541, "step": 2274 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3068.0, "completions/max_terminated_length": 3068.0, "completions/mean_length": 431.7375183105469, "completions/mean_terminated_length": 431.7375183105469, "completions/min_length": 99.0, "completions/min_terminated_length": 99.0, "epoch": 0.8824670287044221, "grad_norm": 15.404497146606445, "kl": 19.796875, "learning_rate": 6.621927081639564e-07, "loss": 0.4243, "num_tokens": 434295691.0, "reward": 0.3787500560283661, "reward_std": 0.2129461020231247, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.29500001668930054, "rewards/unicoder_reward_fn/std": 0.2984771430492401, "step": 2275 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2345.0, "completions/max_terminated_length": 2345.0, "completions/mean_length": 612.4125366210938, "completions/mean_terminated_length": 612.4125366210938, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.8828549262994569, "grad_norm": 8.43836784362793, "kl": 29.25, "learning_rate": 6.61140308263592e-07, "loss": 0.5033, "num_tokens": 434392004.0, "reward": 0.26625004410743713, "reward_std": 0.23735125362873077, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.19500000774860382, "rewards/unicoder_reward_fn/std": 0.26809218525886536, "step": 2276 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3480.0, "completions/mean_length": 689.625, "completions/mean_terminated_length": 510.34210205078125, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.8832428238944918, "grad_norm": 4.391368865966797, "kl": 32.15625, "learning_rate": 6.60091206957803e-07, "loss": 0.5903, "num_tokens": 434493990.0, "reward": 0.2762499749660492, "reward_std": 0.2171163409948349, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.30604666471481323, "step": 2277 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2663.0, "completions/mean_length": 680.8875122070312, "completions/mean_terminated_length": 637.6582641601562, "completions/min_length": 43.0, "completions/min_terminated_length": 43.0, "epoch": 0.8836307214895268, "grad_norm": 2.8412578105926514, "kl": 41.9375, "learning_rate": 6.590454059032626e-07, "loss": 0.7881, "num_tokens": 434606013.0, "reward": 0.1912499964237213, "reward_std": 0.15538696944713593, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.11749999225139618, "rewards/unicoder_reward_fn/std": 0.170461043715477, "step": 2278 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2979.0, "completions/mean_length": 668.3624877929688, "completions/mean_terminated_length": 580.474365234375, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.8840186190845617, "grad_norm": 4.75491189956665, "kl": 49.5625, "learning_rate": 6.580029067514345e-07, "loss": 0.8987, "num_tokens": 434704426.0, "reward": 0.15000000596046448, "reward_std": 0.15755723416805267, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.15186603367328644, "step": 2279 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3946.0, "completions/mean_length": 689.7125244140625, "completions/mean_terminated_length": 602.371826171875, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.8844065166795966, "grad_norm": 5.409670352935791, "kl": 47.0, "learning_rate": 6.569637111485666e-07, "loss": 0.7729, "num_tokens": 434805531.0, "reward": 0.20124998688697815, "reward_std": 0.18759910762310028, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.14000001549720764, "rewards/unicoder_reward_fn/std": 0.23254236578941345, "step": 2280 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 4066.0, "completions/mean_length": 976.3375244140625, "completions/mean_terminated_length": 854.792236328125, "completions/min_length": 49.0, "completions/min_terminated_length": 49.0, "epoch": 0.8847944142746315, "grad_norm": 22.566877365112305, "kl": 65.25, "learning_rate": 6.559278207356905e-07, "loss": 0.9255, "num_tokens": 434929918.0, "reward": 0.10625000298023224, "reward_std": 0.16308389604091644, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.15719051659107208, "step": 2281 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2852.0, "completions/mean_length": 718.6124877929688, "completions/mean_terminated_length": 587.0260009765625, "completions/min_length": 89.0, "completions/min_terminated_length": 89.0, "epoch": 0.8851823118696664, "grad_norm": 5.248780727386475, "kl": 45.6875, "learning_rate": 6.548952371486189e-07, "loss": 0.7781, "num_tokens": 435035455.0, "reward": 0.20625002682209015, "reward_std": 0.16954758763313293, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.21763981878757477, "step": 2282 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.11250000000000004, "completions/max_length": 4096.0, "completions/max_terminated_length": 4001.0, "completions/mean_length": 1337.8499755859375, "completions/mean_terminated_length": 988.225341796875, "completions/min_length": 70.0, "completions/min_terminated_length": 70.0, "epoch": 0.8855702094647013, "grad_norm": 23.290239334106445, "kl": 69.25, "learning_rate": 6.538659620179417e-07, "loss": 1.0427, "num_tokens": 435185643.0, "reward": 0.1587500125169754, "reward_std": 0.20653271675109863, "rewards/format_reward/mean": 0.38749998807907104, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.25576889514923096, "step": 2283 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3608.0, "completions/mean_length": 740.375, "completions/mean_terminated_length": 654.3333129882812, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.8859581070597362, "grad_norm": 6.261320114135742, "kl": 50.75, "learning_rate": 6.528399969690255e-07, "loss": 0.8547, "num_tokens": 435284273.0, "reward": 0.17624999582767487, "reward_std": 0.1424066126346588, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.1738433837890625, "step": 2284 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3607.0, "completions/mean_length": 650.6000366210938, "completions/mean_terminated_length": 516.3636474609375, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.8863460046547711, "grad_norm": 5.729568958282471, "kl": 46.625, "learning_rate": 6.518173436220089e-07, "loss": 0.7328, "num_tokens": 435382081.0, "reward": 0.24625001847743988, "reward_std": 0.23290549218654633, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.2608530521392822, "step": 2285 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3323.0, "completions/mean_length": 582.4000244140625, "completions/mean_terminated_length": 445.5065002441406, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.8867339022498061, "grad_norm": 2.8928775787353516, "kl": 31.84375, "learning_rate": 6.507980035918006e-07, "loss": 0.5818, "num_tokens": 435483985.0, "reward": 0.19750000536441803, "reward_std": 0.15170350670814514, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.1613040268421173, "step": 2286 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3823.0, "completions/mean_length": 728.1625366210938, "completions/mean_terminated_length": 550.9078979492188, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.887121799844841, "grad_norm": 2.1430776119232178, "kl": 47.3125, "learning_rate": 6.497819784880785e-07, "loss": 0.89, "num_tokens": 435586286.0, "reward": 0.30500003695487976, "reward_std": 0.23633353412151337, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.24000000953674316, "rewards/unicoder_reward_fn/std": 0.3124423921108246, "step": 2287 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3426.0, "completions/mean_length": 729.9750366210938, "completions/mean_terminated_length": 643.6666870117188, "completions/min_length": 41.0, "completions/min_terminated_length": 41.0, "epoch": 0.8875096974398758, "grad_norm": 5.095607280731201, "kl": 43.75, "learning_rate": 6.487692699152841e-07, "loss": 0.7352, "num_tokens": 435688372.0, "reward": 0.27625003457069397, "reward_std": 0.24809232354164124, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.20750001072883606, "rewards/unicoder_reward_fn/std": 0.2764032185077667, "step": 2288 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3765.0, "completions/mean_length": 550.5625, "completions/mean_terminated_length": 505.6835632324219, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.8878975950349108, "grad_norm": 10.540570259094238, "kl": 28.03125, "learning_rate": 6.477598794726231e-07, "loss": 0.5525, "num_tokens": 435783857.0, "reward": 0.27125000953674316, "reward_std": 0.2064138501882553, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.19500000774860382, "rewards/unicoder_reward_fn/std": 0.24020034074783325, "step": 2289 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3740.0, "completions/mean_length": 995.3375244140625, "completions/mean_terminated_length": 874.532470703125, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.8882854926299457, "grad_norm": 13.543930053710938, "kl": 54.25, "learning_rate": 6.467538087540606e-07, "loss": 0.8461, "num_tokens": 435910180.0, "reward": 0.19500002264976501, "reward_std": 0.2203173190355301, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.14250001311302185, "rewards/unicoder_reward_fn/std": 0.27038416266441345, "step": 2290 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3876.0, "completions/mean_length": 683.6000366210938, "completions/mean_terminated_length": 596.1026000976562, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.8886733902249806, "grad_norm": 3.8693275451660156, "kl": 34.15625, "learning_rate": 6.457510593483192e-07, "loss": 0.6225, "num_tokens": 436002316.0, "reward": 0.24000000953674316, "reward_std": 0.1895100176334381, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.2289768010377884, "step": 2291 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3720.0, "completions/mean_length": 586.1000366210938, "completions/mean_terminated_length": 541.6708984375, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.8890612878200155, "grad_norm": 7.6723222732543945, "kl": 29.9375, "learning_rate": 6.447516328388777e-07, "loss": 0.5664, "num_tokens": 436102332.0, "reward": 0.257500022649765, "reward_std": 0.20729206502437592, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.25890469551086426, "step": 2292 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3210.0, "completions/mean_length": 717.4750366210938, "completions/mean_terminated_length": 539.6578979492188, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.8894491854150505, "grad_norm": 4.108172416687012, "kl": 36.1875, "learning_rate": 6.437555308039666e-07, "loss": 0.5719, "num_tokens": 436206658.0, "reward": 0.20000000298023224, "reward_std": 0.16808536648750305, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.12999999523162842, "rewards/unicoder_reward_fn/std": 0.23188823461532593, "step": 2293 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3856.0, "completions/mean_length": 711.8375244140625, "completions/mean_terminated_length": 625.0640869140625, "completions/min_length": 47.0, "completions/min_terminated_length": 47.0, "epoch": 0.8898370830100854, "grad_norm": 7.680708885192871, "kl": 38.34375, "learning_rate": 6.427627548165674e-07, "loss": 0.7346, "num_tokens": 436310405.0, "reward": 0.24500003457069397, "reward_std": 0.2120208591222763, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.17749999463558197, "rewards/unicoder_reward_fn/std": 0.25851327180862427, "step": 2294 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 2555.0, "completions/mean_length": 707.4375, "completions/mean_terminated_length": 529.0921020507812, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.8902249806051202, "grad_norm": 8.511635780334473, "kl": 27.296875, "learning_rate": 6.417733064444083e-07, "loss": 0.4731, "num_tokens": 436424904.0, "reward": 0.26875004172325134, "reward_std": 0.2316754311323166, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.28541553020477295, "step": 2295 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3219.0, "completions/mean_length": 614.25, "completions/mean_terminated_length": 478.597412109375, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.8906128782001551, "grad_norm": 13.3451566696167, "kl": 29.84375, "learning_rate": 6.407871872499635e-07, "loss": 0.6497, "num_tokens": 436525484.0, "reward": 0.29500001668930054, "reward_std": 0.2153489589691162, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.2150000035762787, "rewards/unicoder_reward_fn/std": 0.24396513402462006, "step": 2296 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3159.0, "completions/mean_length": 707.5, "completions/mean_terminated_length": 575.4805297851562, "completions/min_length": 62.0, "completions/min_terminated_length": 62.0, "epoch": 0.8910007757951901, "grad_norm": 5.162117004394531, "kl": 31.34375, "learning_rate": 6.398043987904504e-07, "loss": 0.5845, "num_tokens": 436636996.0, "reward": 0.2550000250339508, "reward_std": 0.22323980927467346, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.2533322274684906, "step": 2297 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3820.0, "completions/mean_length": 785.2000122070312, "completions/mean_terminated_length": 656.207763671875, "completions/min_length": 40.0, "completions/min_terminated_length": 40.0, "epoch": 0.891388673390225, "grad_norm": 1.9792393445968628, "kl": 42.375, "learning_rate": 6.388249426178253e-07, "loss": 0.8167, "num_tokens": 436749828.0, "reward": 0.20374999940395355, "reward_std": 0.14021489024162292, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.19602054357528687, "step": 2298 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3428.0, "completions/mean_length": 788.25, "completions/mean_terminated_length": 659.3766479492188, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.8917765709852599, "grad_norm": 10.903046607971191, "kl": 51.125, "learning_rate": 6.378488202787834e-07, "loss": 0.787, "num_tokens": 436861392.0, "reward": 0.10750000923871994, "reward_std": 0.16511695086956024, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.1689693182706833, "step": 2299 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3950.0, "completions/mean_length": 685.7374877929688, "completions/mean_terminated_length": 598.2948608398438, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.8921644685802949, "grad_norm": 5.05714750289917, "kl": 41.6875, "learning_rate": 6.368760333147555e-07, "loss": 0.6414, "num_tokens": 436951339.0, "reward": 0.23125000298023224, "reward_std": 0.23564846813678741, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.2466869205236435, "step": 2300 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3113.0, "completions/mean_length": 835.8375244140625, "completions/mean_terminated_length": 708.8181762695312, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.8925523661753297, "grad_norm": 3.5692012310028076, "kl": 47.5625, "learning_rate": 6.359065832619045e-07, "loss": 0.8117, "num_tokens": 437063078.0, "reward": 0.21375000476837158, "reward_std": 0.1993628293275833, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.26908189058303833, "step": 2301 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3004.0, "completions/mean_length": 890.2125244140625, "completions/mean_terminated_length": 721.4868774414062, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.8929402637703646, "grad_norm": 7.042531490325928, "kl": 47.375, "learning_rate": 6.349404716511247e-07, "loss": 0.8141, "num_tokens": 437179183.0, "reward": 0.20750001072883606, "reward_std": 0.18801270425319672, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.14250001311302185, "rewards/unicoder_reward_fn/std": 0.23854784667491913, "step": 2302 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3550.0, "completions/mean_length": 910.1500244140625, "completions/mean_terminated_length": 828.4615478515625, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.8933281613653995, "grad_norm": 6.644141674041748, "kl": 53.0625, "learning_rate": 6.339777000080384e-07, "loss": 0.9704, "num_tokens": 437293635.0, "reward": 0.2212500125169754, "reward_std": 0.23740659654140472, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.29126524925231934, "step": 2303 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 4031.0, "completions/mean_length": 938.8875122070312, "completions/mean_terminated_length": 815.8831176757812, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.8937160589604345, "grad_norm": 10.47295093536377, "kl": 50.4375, "learning_rate": 6.330182698529928e-07, "loss": 0.7693, "num_tokens": 437416914.0, "reward": 0.21500001847743988, "reward_std": 0.23450933396816254, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.5031546354293823, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.2964344918727875, "step": 2304 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2747.0, "completions/mean_length": 721.4249877929688, "completions/mean_terminated_length": 678.7088623046875, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "epoch": 0.8941039565554694, "grad_norm": 4.768476963043213, "kl": 45.5625, "learning_rate": 6.320621827010595e-07, "loss": 0.8069, "num_tokens": 437517676.0, "reward": 0.2175000160932541, "reward_std": 0.19857631623744965, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.2183801233768463, "step": 2305 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3715.0, "completions/mean_length": 724.4750366210938, "completions/mean_terminated_length": 681.7974853515625, "completions/min_length": 43.0, "completions/min_terminated_length": 43.0, "epoch": 0.8944918541505043, "grad_norm": 9.50647258758545, "kl": 41.5625, "learning_rate": 6.311094400620305e-07, "loss": 0.5976, "num_tokens": 437627522.0, "reward": 0.11250000447034836, "reward_std": 0.10884843021631241, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.14300128817558289, "step": 2306 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3837.0, "completions/mean_length": 759.1625366210938, "completions/mean_terminated_length": 673.6026000976562, "completions/min_length": 78.0, "completions/min_terminated_length": 78.0, "epoch": 0.8948797517455391, "grad_norm": 1.753281593322754, "kl": 36.84375, "learning_rate": 6.301600434404167e-07, "loss": 0.6385, "num_tokens": 437736279.0, "reward": 0.18000000715255737, "reward_std": 0.1938433051109314, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.20769988000392914, "step": 2307 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3994.0, "completions/mean_length": 810.9000244140625, "completions/mean_terminated_length": 638.0, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.8952676493405741, "grad_norm": 5.998199939727783, "kl": 33.9375, "learning_rate": 6.292139943354446e-07, "loss": 0.5012, "num_tokens": 437842959.0, "reward": 0.14000001549720764, "reward_std": 0.17198175191879272, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.19251829385757446, "step": 2308 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3207.0, "completions/mean_length": 785.9500122070312, "completions/mean_terminated_length": 656.9869995117188, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.895655546935609, "grad_norm": 9.712589263916016, "kl": 43.75, "learning_rate": 6.282712942410549e-07, "loss": 0.7486, "num_tokens": 437945331.0, "reward": 0.12375000864267349, "reward_std": 0.20294106006622314, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.22387553751468658, "step": 2309 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2300.0, "completions/mean_length": 609.0499877929688, "completions/mean_terminated_length": 564.911376953125, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "epoch": 0.8960434445306439, "grad_norm": 11.934174537658691, "kl": 26.84375, "learning_rate": 6.273319446459002e-07, "loss": 0.5499, "num_tokens": 438038679.0, "reward": 0.2487500160932541, "reward_std": 0.25010496377944946, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.262401282787323, "step": 2310 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3558.0, "completions/max_terminated_length": 3558.0, "completions/mean_length": 605.7999877929688, "completions/mean_terminated_length": 605.7999877929688, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "epoch": 0.8964313421256789, "grad_norm": 3.6841635704040527, "kl": 29.53125, "learning_rate": 6.26395947033341e-07, "loss": 0.5245, "num_tokens": 438134247.0, "reward": 0.23875001072883606, "reward_std": 0.230800062417984, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.31015512347221375, "step": 2311 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3942.0, "completions/max_terminated_length": 3942.0, "completions/mean_length": 531.2374877929688, "completions/mean_terminated_length": 531.2374877929688, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.8968192397207138, "grad_norm": 6.38387393951416, "kl": 22.9453125, "learning_rate": 6.254633028814459e-07, "loss": 0.4051, "num_tokens": 438222514.0, "reward": 0.21000002324581146, "reward_std": 0.160136416554451, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.14000001549720764, "rewards/unicoder_reward_fn/std": 0.19974668323993683, "step": 2312 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3096.0, "completions/mean_length": 504.45001220703125, "completions/mean_terminated_length": 458.98736572265625, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.8972071373157486, "grad_norm": 8.64488697052002, "kl": 20.78125, "learning_rate": 6.245340136629872e-07, "loss": 0.4153, "num_tokens": 438310158.0, "reward": 0.19749999046325684, "reward_std": 0.20447149872779846, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.20682960748672485, "step": 2313 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 4079.0, "completions/mean_length": 741.7374877929688, "completions/mean_terminated_length": 655.7307739257812, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.8975950349107835, "grad_norm": 1.8443647623062134, "kl": 33.1875, "learning_rate": 6.236080808454395e-07, "loss": 0.6165, "num_tokens": 438420001.0, "reward": 0.17000000178813934, "reward_std": 0.217554971575737, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.11749999225139618, "rewards/unicoder_reward_fn/std": 0.2374841868877411, "step": 2314 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3868.0, "completions/mean_length": 682.5750122070312, "completions/mean_terminated_length": 639.3671264648438, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.8979829325058185, "grad_norm": 5.7921271324157715, "kl": 25.75, "learning_rate": 6.226855058909763e-07, "loss": 0.4167, "num_tokens": 438527191.0, "reward": 0.17749999463558197, "reward_std": 0.18065960705280304, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.2580844759941101, "step": 2315 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3357.0, "completions/mean_length": 842.8250122070312, "completions/mean_terminated_length": 759.4102783203125, "completions/min_length": 53.0, "completions/min_terminated_length": 53.0, "epoch": 0.8983708301008534, "grad_norm": 3.7507452964782715, "kl": 33.53125, "learning_rate": 6.217662902564704e-07, "loss": 0.6581, "num_tokens": 438643505.0, "reward": 0.11625001579523087, "reward_std": 0.15828314423561096, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.06000000238418579, "rewards/unicoder_reward_fn/std": 0.15061058104038239, "step": 2316 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3773.0, "completions/max_terminated_length": 3773.0, "completions/mean_length": 682.9625244140625, "completions/mean_terminated_length": 682.9625244140625, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.8987587276958883, "grad_norm": 4.62839937210083, "kl": 31.25, "learning_rate": 6.208504353934885e-07, "loss": 0.5052, "num_tokens": 438743134.0, "reward": 0.14500001072883606, "reward_std": 0.17088228464126587, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.22690823674201965, "step": 2317 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3866.0, "completions/mean_length": 834.8500366210938, "completions/mean_terminated_length": 751.2307739257812, "completions/min_length": 56.0, "completions/min_terminated_length": 56.0, "epoch": 0.8991466252909232, "grad_norm": 3.640059232711792, "kl": 35.6875, "learning_rate": 6.199379427482897e-07, "loss": 0.6409, "num_tokens": 438861354.0, "reward": 0.1550000160932541, "reward_std": 0.22049008309841156, "rewards/format_reward/mean": 0.4749999940395355, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.23372329771518707, "step": 2318 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2791.0, "completions/max_terminated_length": 2791.0, "completions/mean_length": 655.4500122070312, "completions/mean_terminated_length": 655.4500122070312, "completions/min_length": 56.0, "completions/min_terminated_length": 56.0, "epoch": 0.8995345228859581, "grad_norm": 10.748713493347168, "kl": 37.125, "learning_rate": 6.190288137618256e-07, "loss": 0.5292, "num_tokens": 438962222.0, "reward": 0.08250000327825546, "reward_std": 0.114463210105896, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.14264677464962006, "step": 2319 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2917.0, "completions/max_terminated_length": 2917.0, "completions/mean_length": 756.0499877929688, "completions/mean_terminated_length": 756.0499877929688, "completions/min_length": 62.0, "completions/min_terminated_length": 62.0, "epoch": 0.899922420480993, "grad_norm": 16.965484619140625, "kl": 37.25, "learning_rate": 6.181230498697343e-07, "loss": 0.5423, "num_tokens": 439071858.0, "reward": 0.15125000476837158, "reward_std": 0.17425833642482758, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.10249999910593033, "rewards/unicoder_reward_fn/std": 0.22274184226989746, "step": 2320 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3762.0, "completions/mean_length": 850.6124877929688, "completions/mean_terminated_length": 809.5316772460938, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.9003103180760279, "grad_norm": 10.674076080322266, "kl": 41.4375, "learning_rate": 6.172206525023406e-07, "loss": 0.6729, "num_tokens": 439194507.0, "reward": 0.13625000417232513, "reward_std": 0.22048859298229218, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.21810460090637207, "step": 2321 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2370.0, "completions/mean_length": 585.3624877929688, "completions/mean_terminated_length": 495.3461608886719, "completions/min_length": 92.0, "completions/min_terminated_length": 92.0, "epoch": 0.9006982156710628, "grad_norm": 3.2991950511932373, "kl": 27.75, "learning_rate": 6.163216230846532e-07, "loss": 0.3878, "num_tokens": 439278936.0, "reward": 0.1862500011920929, "reward_std": 0.21846090257167816, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.24104204773902893, "step": 2322 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3790.0, "completions/mean_length": 776.9625244140625, "completions/mean_terminated_length": 734.9494018554688, "completions/min_length": 53.0, "completions/min_terminated_length": 53.0, "epoch": 0.9010861132660978, "grad_norm": 3.5334248542785645, "kl": 33.375, "learning_rate": 6.154259630363625e-07, "loss": 0.5765, "num_tokens": 439382901.0, "reward": 0.19249999523162842, "reward_std": 0.19815205037593842, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.13999998569488525, "rewards/unicoder_reward_fn/std": 0.2707701027393341, "step": 2323 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2442.0, "completions/max_terminated_length": 2442.0, "completions/mean_length": 502.9750061035156, "completions/mean_terminated_length": 502.9750061035156, "completions/min_length": 52.0, "completions/min_terminated_length": 52.0, "epoch": 0.9014740108611327, "grad_norm": 5.299715518951416, "kl": 25.25, "learning_rate": 6.145336737718375e-07, "loss": 0.4368, "num_tokens": 439473571.0, "reward": 0.14374999701976776, "reward_std": 0.150228351354599, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.16441303491592407, "step": 2324 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3450.0, "completions/max_terminated_length": 3450.0, "completions/mean_length": 543.9125366210938, "completions/mean_terminated_length": 543.9125366210938, "completions/min_length": 99.0, "completions/min_terminated_length": 99.0, "epoch": 0.9018619084561675, "grad_norm": 12.447772979736328, "kl": 21.875, "learning_rate": 6.136447567001252e-07, "loss": 0.4903, "num_tokens": 439563068.0, "reward": 0.1875000149011612, "reward_std": 0.17045097053050995, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.11749999225139618, "rewards/unicoder_reward_fn/std": 0.20547877252101898, "step": 2325 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3610.0, "completions/mean_length": 565.4500122070312, "completions/mean_terminated_length": 520.759521484375, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.9022498060512025, "grad_norm": 7.448423385620117, "kl": 22.78125, "learning_rate": 6.127592132249465e-07, "loss": 0.3859, "num_tokens": 439661616.0, "reward": 0.26750001311302185, "reward_std": 0.23531608283519745, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.19749999046325684, "rewards/unicoder_reward_fn/std": 0.30604666471481323, "step": 2326 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2235.0, "completions/max_terminated_length": 2235.0, "completions/mean_length": 335.3500061035156, "completions/mean_terminated_length": 335.3500061035156, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.9026377036462374, "grad_norm": 9.555004119873047, "kl": 11.578125, "learning_rate": 6.118770447446962e-07, "loss": 0.2017, "num_tokens": 439738580.0, "reward": 0.2562500238418579, "reward_std": 0.18307411670684814, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.2280905693769455, "step": 2327 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2120.0, "completions/mean_length": 521.9500122070312, "completions/mean_terminated_length": 430.3077087402344, "completions/min_length": 67.0, "completions/min_terminated_length": 67.0, "epoch": 0.9030256012412723, "grad_norm": 20.81097412109375, "kl": 21.78125, "learning_rate": 6.109982526524379e-07, "loss": 0.5681, "num_tokens": 439822728.0, "reward": 0.2550000250339508, "reward_std": 0.2076202929019928, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.22133083641529083, "step": 2328 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3486.0, "completions/mean_length": 524.2875366210938, "completions/mean_terminated_length": 432.70513916015625, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.9034134988363072, "grad_norm": 17.040302276611328, "kl": 21.96875, "learning_rate": 6.101228383359046e-07, "loss": 0.554, "num_tokens": 439911047.0, "reward": 0.28125, "reward_std": 0.18407557904720306, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.20500001311302185, "rewards/unicoder_reward_fn/std": 0.23808306455612183, "step": 2329 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2505.0, "completions/max_terminated_length": 2505.0, "completions/mean_length": 397.7375183105469, "completions/mean_terminated_length": 397.7375183105469, "completions/min_length": 166.0, "completions/min_terminated_length": 166.0, "epoch": 0.9038013964313422, "grad_norm": 9.609770774841309, "kl": 12.453125, "learning_rate": 6.092508031774954e-07, "loss": 0.2919, "num_tokens": 439992954.0, "reward": 0.257500022649765, "reward_std": 0.148403599858284, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.17000000178813934, "rewards/unicoder_reward_fn/std": 0.27020856738090515, "step": 2330 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2268.0, "completions/max_terminated_length": 2268.0, "completions/mean_length": 383.6625061035156, "completions/mean_terminated_length": 383.6625061035156, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.904189294026377, "grad_norm": 9.769867897033691, "kl": 13.671875, "learning_rate": 6.083821485542719e-07, "loss": 0.291, "num_tokens": 440074447.0, "reward": 0.21375003457069397, "reward_std": 0.07703851908445358, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.1274999976158142, "rewards/unicoder_reward_fn/std": 0.19935020804405212, "step": 2331 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2951.0, "completions/max_terminated_length": 2951.0, "completions/mean_length": 437.7124938964844, "completions/mean_terminated_length": 437.7124938964844, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.9045771916214119, "grad_norm": 7.333971977233887, "kl": 17.625, "learning_rate": 6.075168758379594e-07, "loss": 0.3377, "num_tokens": 440158936.0, "reward": 0.2237500250339508, "reward_std": 0.17205749452114105, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.2002371996641159, "step": 2332 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1619.0, "completions/max_terminated_length": 1619.0, "completions/mean_length": 355.875, "completions/mean_terminated_length": 355.875, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "epoch": 0.9049650892164468, "grad_norm": 10.89355182647705, "kl": 12.34375, "learning_rate": 6.06654986394941e-07, "loss": 0.2559, "num_tokens": 440230094.0, "reward": 0.4775000214576721, "reward_std": 0.29430100321769714, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.33280548453330994, "rewards/unicoder_reward_fn/mean": 0.39000001549720764, "rewards/unicoder_reward_fn/std": 0.40864086151123047, "step": 2333 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1762.0, "completions/max_terminated_length": 1762.0, "completions/mean_length": 297.13751220703125, "completions/mean_terminated_length": 297.13751220703125, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "epoch": 0.9053529868114818, "grad_norm": 10.938139915466309, "kl": 9.6982421875, "learning_rate": 6.057964815862579e-07, "loss": 0.1808, "num_tokens": 440302793.0, "reward": 0.5400000214576721, "reward_std": 0.19080601632595062, "rewards/format_reward/mean": 0.949999988079071, "rewards/format_reward/std": 0.21931999921798706, "rewards/unicoder_reward_fn/mean": 0.445000022649765, "rewards/unicoder_reward_fn/std": 0.32604262232780457, "step": 2334 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3779.0, "completions/mean_length": 577.9750366210938, "completions/mean_terminated_length": 533.4430541992188, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "epoch": 0.9057408844065167, "grad_norm": 3.4490935802459717, "kl": 34.0, "learning_rate": 6.049413627676064e-07, "loss": 0.6361, "num_tokens": 440397735.0, "reward": 0.2637500464916229, "reward_std": 0.17762066423892975, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.2834573984146118, "step": 2335 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2638.0, "completions/mean_length": 513.6875, "completions/mean_terminated_length": 468.341796875, "completions/min_length": 77.0, "completions/min_terminated_length": 77.0, "epoch": 0.9061287820015516, "grad_norm": 6.964608192443848, "kl": 30.71875, "learning_rate": 6.040896312893355e-07, "loss": 0.555, "num_tokens": 440484078.0, "reward": 0.43250003457069397, "reward_std": 0.25544363260269165, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.35500001907348633, "rewards/unicoder_reward_fn/std": 0.3805059790611267, "step": 2336 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2807.0, "completions/mean_length": 416.4875183105469, "completions/mean_terminated_length": 369.9114074707031, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.9065166795965865, "grad_norm": 12.998773574829102, "kl": 21.765625, "learning_rate": 6.03241288496446e-07, "loss": 0.4081, "num_tokens": 440553061.0, "reward": 0.36500000953674316, "reward_std": 0.2325562685728073, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.2824999988079071, "rewards/unicoder_reward_fn/std": 0.3271569013595581, "step": 2337 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3198.0, "completions/max_terminated_length": 3198.0, "completions/mean_length": 563.8500366210938, "completions/mean_terminated_length": 563.8500366210938, "completions/min_length": 144.0, "completions/min_terminated_length": 144.0, "epoch": 0.9069045771916214, "grad_norm": 5.750122547149658, "kl": 40.125, "learning_rate": 6.023963357285869e-07, "loss": 0.6462, "num_tokens": 440638489.0, "reward": 0.30375000834465027, "reward_std": 0.1747356504201889, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.23250000178813934, "rewards/unicoder_reward_fn/std": 0.3173366189002991, "step": 2338 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3574.0, "completions/mean_length": 621.1625366210938, "completions/mean_terminated_length": 532.0640869140625, "completions/min_length": 123.0, "completions/min_terminated_length": 123.0, "epoch": 0.9072924747866563, "grad_norm": 12.362531661987305, "kl": 47.9375, "learning_rate": 6.015547743200535e-07, "loss": 0.7701, "num_tokens": 440742374.0, "reward": 0.2212500125169754, "reward_std": 0.17289768159389496, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.19225510954856873, "step": 2339 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 4087.0, "completions/mean_length": 745.6875, "completions/mean_terminated_length": 659.7820434570312, "completions/min_length": 118.0, "completions/min_terminated_length": 118.0, "epoch": 0.9076803723816912, "grad_norm": 6.363870620727539, "kl": 54.625, "learning_rate": 6.007166055997865e-07, "loss": 0.9115, "num_tokens": 440847061.0, "reward": 0.17499999701976776, "reward_std": 0.17433665692806244, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.17957404255867004, "step": 2340 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3745.0, "completions/mean_length": 653.6500244140625, "completions/mean_terminated_length": 610.0759887695312, "completions/min_length": 195.0, "completions/min_terminated_length": 195.0, "epoch": 0.9080682699767262, "grad_norm": 8.975284576416016, "kl": 49.6875, "learning_rate": 5.998818308913685e-07, "loss": 0.789, "num_tokens": 440948545.0, "reward": 0.15125000476837158, "reward_std": 0.18445563316345215, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.09250000864267349, "rewards/unicoder_reward_fn/std": 0.22489097714424133, "step": 2341 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3775.0, "completions/mean_length": 924.1000366210938, "completions/mean_terminated_length": 800.5194702148438, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.9084561675717611, "grad_norm": 21.640600204467773, "kl": 59.625, "learning_rate": 5.990504515130232e-07, "loss": 0.8168, "num_tokens": 441067297.0, "reward": 0.257500022649765, "reward_std": 0.32377609610557556, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.32839271426200867, "step": 2342 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 2563.0, "completions/mean_length": 755.4375, "completions/mean_terminated_length": 579.618408203125, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.908844065166796, "grad_norm": 9.136724472045898, "kl": 51.4375, "learning_rate": 5.982224687776116e-07, "loss": 0.8936, "num_tokens": 441175588.0, "reward": 0.18000000715255737, "reward_std": 0.183511883020401, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.22690823674201965, "step": 2343 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2022.0, "completions/mean_length": 602.5499877929688, "completions/mean_terminated_length": 512.974365234375, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.9092319627618308, "grad_norm": 3.534237861633301, "kl": 43.625, "learning_rate": 5.973978839926314e-07, "loss": 0.7316, "num_tokens": 441284496.0, "reward": 0.20000000298023224, "reward_std": 0.21061073243618011, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.21839460730552673, "step": 2344 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3875.0, "completions/mean_length": 795.0, "completions/mean_terminated_length": 753.2152099609375, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.9096198603568658, "grad_norm": 13.268715858459473, "kl": 55.9375, "learning_rate": 5.96576698460215e-07, "loss": 0.8598, "num_tokens": 441397384.0, "reward": 0.16500000655651093, "reward_std": 0.15615446865558624, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.15253138542175293, "step": 2345 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 4055.0, "completions/mean_length": 744.7000122070312, "completions/mean_terminated_length": 614.1298828125, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.9100077579519007, "grad_norm": 7.435854911804199, "kl": 48.75, "learning_rate": 5.957589134771259e-07, "loss": 0.7789, "num_tokens": 441502576.0, "reward": 0.2525000274181366, "reward_std": 0.1922747939825058, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.24097640812397003, "step": 2346 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2878.0, "completions/mean_length": 642.125, "completions/mean_terminated_length": 553.5640869140625, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.9103956555469356, "grad_norm": 2.933926582336426, "kl": 40.21875, "learning_rate": 5.949445303347589e-07, "loss": 0.704, "num_tokens": 441599194.0, "reward": 0.18500001728534698, "reward_std": 0.17524360120296478, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.44933146238327026, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.19051894545555115, "step": 2347 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2398.0, "completions/mean_length": 675.1124877929688, "completions/mean_terminated_length": 541.8311767578125, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.9107835531419706, "grad_norm": 6.258447170257568, "kl": 40.875, "learning_rate": 5.941335503191357e-07, "loss": 0.6989, "num_tokens": 441695859.0, "reward": 0.21875, "reward_std": 0.19261495769023895, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.22556540369987488, "step": 2348 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3061.0, "completions/mean_length": 736.2374877929688, "completions/mean_terminated_length": 650.0897216796875, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.9111714507370055, "grad_norm": 6.155393123626709, "kl": 42.8125, "learning_rate": 5.933259747109043e-07, "loss": 0.6968, "num_tokens": 441806350.0, "reward": 0.20374999940395355, "reward_std": 0.17330078780651093, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.19602054357528687, "step": 2349 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3853.0, "completions/mean_length": 872.4625244140625, "completions/mean_terminated_length": 746.8701171875, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.9115593483320403, "grad_norm": 8.00830078125, "kl": 48.6875, "learning_rate": 5.925218047853371e-07, "loss": 0.7956, "num_tokens": 441923851.0, "reward": 0.18000000715255737, "reward_std": 0.1979169100522995, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.1224999874830246, "rewards/unicoder_reward_fn/std": 0.2575320899486542, "step": 2350 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3064.0, "completions/mean_length": 641.6375122070312, "completions/mean_terminated_length": 553.0640869140625, "completions/min_length": 49.0, "completions/min_terminated_length": 49.0, "epoch": 0.9119472459270752, "grad_norm": 4.3975138664245605, "kl": 33.5, "learning_rate": 5.917210418123279e-07, "loss": 0.5839, "num_tokens": 442021630.0, "reward": 0.27250000834465027, "reward_std": 0.19610154628753662, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.20000000298023224, "rewards/unicoder_reward_fn/std": 0.27374571561813354, "step": 2351 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3648.0, "completions/max_terminated_length": 3648.0, "completions/mean_length": 591.1500244140625, "completions/mean_terminated_length": 591.1500244140625, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.9123351435221102, "grad_norm": 4.789183139801025, "kl": 35.34375, "learning_rate": 5.909236870563911e-07, "loss": 0.6059, "num_tokens": 442110530.0, "reward": 0.20250001549720764, "reward_std": 0.2041792869567871, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.2376573532819748, "step": 2352 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3664.0, "completions/mean_length": 634.875, "completions/mean_terminated_length": 591.0632934570312, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.9127230411171451, "grad_norm": 2.9162757396698, "kl": 36.25, "learning_rate": 5.901297417766589e-07, "loss": 0.6576, "num_tokens": 442205832.0, "reward": 0.19375000894069672, "reward_std": 0.17498679459095, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.24076619744300842, "step": 2353 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3888.0, "completions/mean_length": 850.8624877929688, "completions/mean_terminated_length": 724.4285888671875, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "epoch": 0.91311093871218, "grad_norm": 4.0928497314453125, "kl": 34.53125, "learning_rate": 5.893392072268781e-07, "loss": 0.5466, "num_tokens": 442329509.0, "reward": 0.2862500250339508, "reward_std": 0.21216832101345062, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.21250000596046448, "rewards/unicoder_reward_fn/std": 0.2668973207473755, "step": 2354 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3553.0, "completions/mean_length": 809.9500122070312, "completions/mean_terminated_length": 725.6923217773438, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.9134988363072148, "grad_norm": 3.4246251583099365, "kl": 38.0625, "learning_rate": 5.88552084655412e-07, "loss": 0.6088, "num_tokens": 442437417.0, "reward": 0.1887500137090683, "reward_std": 0.22447870671749115, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.2431335747241974, "step": 2355 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 2710.0, "completions/mean_length": 952.0, "completions/mean_terminated_length": 786.5263061523438, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.9138867339022498, "grad_norm": 14.636723518371582, "kl": 43.9375, "learning_rate": 5.877683753052336e-07, "loss": 0.6615, "num_tokens": 442562393.0, "reward": 0.1612500101327896, "reward_std": 0.1468632072210312, "rewards/format_reward/mean": 0.512499988079071, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.2132331132888794, "step": 2356 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2791.0, "completions/max_terminated_length": 2791.0, "completions/mean_length": 540.5750122070312, "completions/mean_terminated_length": 540.5750122070312, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "epoch": 0.9142746314972847, "grad_norm": 13.591578483581543, "kl": 31.3125, "learning_rate": 5.869880804139277e-07, "loss": 0.6924, "num_tokens": 442649183.0, "reward": 0.2537499964237213, "reward_std": 0.18405954539775848, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.1824999898672104, "rewards/unicoder_reward_fn/std": 0.24483266472816467, "step": 2357 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3966.0, "completions/max_terminated_length": 3966.0, "completions/mean_length": 646.7000122070312, "completions/mean_terminated_length": 646.7000122070312, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.9146625290923196, "grad_norm": 7.47510290145874, "kl": 27.78125, "learning_rate": 5.86211201213686e-07, "loss": 0.5453, "num_tokens": 442745487.0, "reward": 0.2600000202655792, "reward_std": 0.225098118185997, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.28541553020477295, "step": 2358 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3759.0, "completions/max_terminated_length": 3759.0, "completions/mean_length": 547.7750244140625, "completions/mean_terminated_length": 547.7750244140625, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.9150504266873546, "grad_norm": 16.441089630126953, "kl": 20.71875, "learning_rate": 5.854377389313067e-07, "loss": 0.4133, "num_tokens": 442843125.0, "reward": 0.4000000059604645, "reward_std": 0.3142605721950531, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.3199999928474426, "rewards/unicoder_reward_fn/std": 0.35235196352005005, "step": 2359 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2921.0, "completions/mean_length": 705.1375122070312, "completions/mean_terminated_length": 662.2152099609375, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.9154383242823895, "grad_norm": 5.866003513336182, "kl": 38.375, "learning_rate": 5.846676947881928e-07, "loss": 0.6349, "num_tokens": 442948992.0, "reward": 0.17625001072883606, "reward_std": 0.22572050988674164, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.11749999225139618, "rewards/unicoder_reward_fn/std": 0.24586454033851624, "step": 2360 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2559.0, "completions/mean_length": 793.1749877929688, "completions/mean_terminated_length": 751.3671264648438, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.9158262218774244, "grad_norm": 9.775262832641602, "kl": 37.78125, "learning_rate": 5.839010700003488e-07, "loss": 0.5866, "num_tokens": 443055582.0, "reward": 0.13875000178813934, "reward_std": 0.1582740992307663, "rewards/format_reward/mean": 0.512499988079071, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.1823718547821045, "step": 2361 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3504.0, "completions/max_terminated_length": 3504.0, "completions/mean_length": 785.2625122070312, "completions/mean_terminated_length": 785.2625122070312, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.9162141194724592, "grad_norm": 4.066385746002197, "kl": 38.59375, "learning_rate": 5.8313786577838e-07, "loss": 0.668, "num_tokens": 443166547.0, "reward": 0.20125003159046173, "reward_std": 0.23653554916381836, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2916127145290375, "step": 2362 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3042.0, "completions/mean_length": 713.9000244140625, "completions/mean_terminated_length": 627.1795043945312, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.9166020170674942, "grad_norm": 5.815977096557617, "kl": 34.3125, "learning_rate": 5.823780833274894e-07, "loss": 0.5485, "num_tokens": 443276675.0, "reward": 0.11000000685453415, "reward_std": 0.14449022710323334, "rewards/format_reward/mean": 0.4749999940395355, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.0625, "rewards/unicoder_reward_fn/std": 0.14787524938583374, "step": 2363 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2969.0, "completions/mean_length": 833.7374877929688, "completions/mean_terminated_length": 792.4430541992188, "completions/min_length": 185.0, "completions/min_terminated_length": 185.0, "epoch": 0.9169899146625291, "grad_norm": 9.933703422546387, "kl": 35.8125, "learning_rate": 5.816217238474782e-07, "loss": 0.6452, "num_tokens": 443391094.0, "reward": 0.11500000208616257, "reward_std": 0.12539666891098022, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.14917494356632233, "step": 2364 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3536.0, "completions/max_terminated_length": 3536.0, "completions/mean_length": 618.8250122070312, "completions/mean_terminated_length": 618.8250122070312, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.917377812257564, "grad_norm": 10.490452766418457, "kl": 24.625, "learning_rate": 5.80868788532741e-07, "loss": 0.4925, "num_tokens": 443495048.0, "reward": 0.16625002026557922, "reward_std": 0.14113853871822357, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.18278783559799194, "step": 2365 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2783.0, "completions/mean_length": 696.0750122070312, "completions/mean_terminated_length": 653.0379638671875, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "epoch": 0.9177657098525989, "grad_norm": 6.992780685424805, "kl": 35.125, "learning_rate": 5.801192785722648e-07, "loss": 0.5395, "num_tokens": 443596478.0, "reward": 0.13625000417232513, "reward_std": 0.1805988997220993, "rewards/format_reward/mean": 0.4625000059604645, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.09000000357627869, "rewards/unicoder_reward_fn/std": 0.2108451873064041, "step": 2366 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3269.0, "completions/max_terminated_length": 3269.0, "completions/mean_length": 446.01251220703125, "completions/mean_terminated_length": 446.01251220703125, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.9181536074476339, "grad_norm": 17.325233459472656, "kl": 15.890625, "learning_rate": 5.793731951496289e-07, "loss": 0.4634, "num_tokens": 443685527.0, "reward": 0.1600000113248825, "reward_std": 0.1398274153470993, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.17603221535682678, "step": 2367 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3806.0, "completions/max_terminated_length": 3806.0, "completions/mean_length": 659.0, "completions/mean_terminated_length": 659.0, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.9185415050426687, "grad_norm": 7.70704460144043, "kl": 22.78125, "learning_rate": 5.786305394430003e-07, "loss": 0.4753, "num_tokens": 443797583.0, "reward": 0.16375000774860382, "reward_std": 0.15992534160614014, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.19355109333992004, "step": 2368 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2092.0, "completions/max_terminated_length": 2092.0, "completions/mean_length": 543.4750366210938, "completions/mean_terminated_length": 543.4750366210938, "completions/min_length": 90.0, "completions/min_terminated_length": 90.0, "epoch": 0.9189294026377036, "grad_norm": 9.900200843811035, "kl": 22.75, "learning_rate": 5.778913126251339e-07, "loss": 0.4088, "num_tokens": 443889037.0, "reward": 0.26125001907348633, "reward_std": 0.22822605073451996, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.1925000101327896, "rewards/unicoder_reward_fn/std": 0.2906888723373413, "step": 2369 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1645.0, "completions/max_terminated_length": 1645.0, "completions/mean_length": 478.1499938964844, "completions/mean_terminated_length": 478.1499938964844, "completions/min_length": 45.0, "completions/min_terminated_length": 45.0, "epoch": 0.9193173002327386, "grad_norm": 9.06127643585205, "kl": 19.671875, "learning_rate": 5.771555158633699e-07, "loss": 0.4365, "num_tokens": 443981041.0, "reward": 0.1275000125169754, "reward_std": 0.08964955806732178, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.05249999836087227, "rewards/unicoder_reward_fn/std": 0.09409797936677933, "step": 2370 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3611.0, "completions/max_terminated_length": 3611.0, "completions/mean_length": 656.9125366210938, "completions/mean_terminated_length": 656.9125366210938, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.9197051978277735, "grad_norm": 6.468993186950684, "kl": 24.8125, "learning_rate": 5.76423150319632e-07, "loss": 0.4277, "num_tokens": 444093762.0, "reward": 0.12000000476837158, "reward_std": 0.1086517795920372, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.14755390584468842, "step": 2371 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2668.0, "completions/mean_length": 717.3250122070312, "completions/mean_terminated_length": 630.6923217773438, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.9200930954228084, "grad_norm": 8.729488372802734, "kl": 25.15625, "learning_rate": 5.756942171504247e-07, "loss": 0.4859, "num_tokens": 444207652.0, "reward": 0.13125000894069672, "reward_std": 0.13164809346199036, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.12365734577178955, "step": 2372 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3278.0, "completions/mean_length": 655.3250122070312, "completions/mean_terminated_length": 611.7721557617188, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "epoch": 0.9204809930178433, "grad_norm": 4.500354290008545, "kl": 26.3125, "learning_rate": 5.749687175068337e-07, "loss": 0.4909, "num_tokens": 444314110.0, "reward": 0.20999999344348907, "reward_std": 0.21550516784191132, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2755431830883026, "step": 2373 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3801.0, "completions/max_terminated_length": 3801.0, "completions/mean_length": 697.5375366210938, "completions/mean_terminated_length": 697.5375366210938, "completions/min_length": 165.0, "completions/min_terminated_length": 165.0, "epoch": 0.9208688906128782, "grad_norm": 3.6512038707733154, "kl": 29.3125, "learning_rate": 5.742466525345213e-07, "loss": 0.5008, "num_tokens": 444416225.0, "reward": 0.16250000894069672, "reward_std": 0.1572139710187912, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.22501757740974426, "step": 2374 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3955.0, "completions/max_terminated_length": 3955.0, "completions/mean_length": 676.1375122070312, "completions/mean_terminated_length": 676.1375122070312, "completions/min_length": 26.0, "completions/min_terminated_length": 26.0, "epoch": 0.9212567882079131, "grad_norm": 5.574721336364746, "kl": 26.1875, "learning_rate": 5.735280233737272e-07, "loss": 0.5071, "num_tokens": 444522204.0, "reward": 0.2212500125169754, "reward_std": 0.2161877602338791, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.33631473779678345, "step": 2375 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2898.0, "completions/max_terminated_length": 2898.0, "completions/mean_length": 564.3375244140625, "completions/mean_terminated_length": 564.3375244140625, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.921644685802948, "grad_norm": 4.195156097412109, "kl": 29.09375, "learning_rate": 5.728128311592648e-07, "loss": 0.5262, "num_tokens": 444611575.0, "reward": 0.13625000417232513, "reward_std": 0.1653060019016266, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.07999999821186066, "rewards/unicoder_reward_fn/std": 0.1512814462184906, "step": 2376 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3272.0, "completions/max_terminated_length": 3272.0, "completions/mean_length": 694.25, "completions/mean_terminated_length": 694.25, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.9220325833979829, "grad_norm": 14.250075340270996, "kl": 40.375, "learning_rate": 5.721010770205198e-07, "loss": 0.5866, "num_tokens": 444718979.0, "reward": 0.11000001430511475, "reward_std": 0.13107822835445404, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.0650000050663948, "rewards/unicoder_reward_fn/std": 0.1378863900899887, "step": 2377 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3073.0, "completions/mean_length": 848.5875244140625, "completions/mean_terminated_length": 807.4810180664062, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.9224204809930179, "grad_norm": 21.218570709228516, "kl": 43.5625, "learning_rate": 5.713927620814497e-07, "loss": 0.638, "num_tokens": 444832002.0, "reward": 0.09250000864267349, "reward_std": 0.16291336715221405, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.17422524094581604, "step": 2378 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3775.0, "completions/mean_length": 852.5250244140625, "completions/mean_terminated_length": 636.2933349609375, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.9228083785880528, "grad_norm": 6.240185260772705, "kl": 40.125, "learning_rate": 5.706878874605801e-07, "loss": 0.6298, "num_tokens": 444956540.0, "reward": 0.14625000953674316, "reward_std": 0.16279935836791992, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.20865756273269653, "step": 2379 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2957.0, "completions/max_terminated_length": 2957.0, "completions/mean_length": 741.1625366210938, "completions/mean_terminated_length": 741.1625366210938, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.9231962761830876, "grad_norm": 17.658245086669922, "kl": 41.75, "learning_rate": 5.699864542710045e-07, "loss": 0.5654, "num_tokens": 445066593.0, "reward": 0.09875000268220901, "reward_std": 0.16361413896083832, "rewards/format_reward/mean": 0.3375000059604645, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.06499999761581421, "rewards/unicoder_reward_fn/std": 0.19558416306972504, "step": 2380 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3740.0, "completions/max_terminated_length": 3740.0, "completions/mean_length": 683.0250244140625, "completions/mean_terminated_length": 683.0250244140625, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.9235841737781226, "grad_norm": 17.44599151611328, "kl": 43.0, "learning_rate": 5.692884636203816e-07, "loss": 0.6661, "num_tokens": 445161835.0, "reward": 0.0962500125169754, "reward_std": 0.0998762845993042, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.14317822456359863, "step": 2381 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3311.0, "completions/max_terminated_length": 3311.0, "completions/mean_length": 765.9249877929688, "completions/mean_terminated_length": 765.9249877929688, "completions/min_length": 133.0, "completions/min_terminated_length": 133.0, "epoch": 0.9239720713731575, "grad_norm": 11.494840621948242, "kl": 39.375, "learning_rate": 5.685939166109338e-07, "loss": 0.6497, "num_tokens": 445269053.0, "reward": 0.08750000596046448, "reward_std": 0.10778830200433731, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.042500000447034836, "rewards/unicoder_reward_fn/std": 0.10881990939378738, "step": 2382 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4087.0, "completions/max_terminated_length": 4087.0, "completions/mean_length": 657.3500366210938, "completions/mean_terminated_length": 657.3500366210938, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.9243599689681924, "grad_norm": 12.67678451538086, "kl": 35.21875, "learning_rate": 5.679028143394457e-07, "loss": 0.5306, "num_tokens": 445367385.0, "reward": 0.10375000536441803, "reward_std": 0.14535920321941376, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.17574433982372284, "step": 2383 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2816.0, "completions/mean_length": 738.6124877929688, "completions/mean_terminated_length": 652.525634765625, "completions/min_length": 17.0, "completions/min_terminated_length": 17.0, "epoch": 0.9247478665632273, "grad_norm": 4.409816741943359, "kl": 34.65625, "learning_rate": 5.672151578972621e-07, "loss": 0.6446, "num_tokens": 445465642.0, "reward": 0.1575000137090683, "reward_std": 0.19207309186458588, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619150161743, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.2204713076353073, "step": 2384 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2049.0, "completions/max_terminated_length": 2049.0, "completions/mean_length": 530.75, "completions/mean_terminated_length": 530.75, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.9251357641582623, "grad_norm": 7.463233470916748, "kl": 25.84375, "learning_rate": 5.665309483702863e-07, "loss": 0.4758, "num_tokens": 445556806.0, "reward": 0.17250001430511475, "reward_std": 0.18714581429958344, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.11000001430511475, "rewards/unicoder_reward_fn/std": 0.1852811723947525, "step": 2385 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3761.0, "completions/max_terminated_length": 3761.0, "completions/mean_length": 506.63751220703125, "completions/mean_terminated_length": 506.63751220703125, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.9255236617532971, "grad_norm": 12.518546104431152, "kl": 22.21875, "learning_rate": 5.658501868389786e-07, "loss": 0.3068, "num_tokens": 445642265.0, "reward": 0.24750001728534698, "reward_std": 0.23766230046749115, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.33314764499664307, "step": 2386 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2952.0, "completions/max_terminated_length": 2952.0, "completions/mean_length": 542.625, "completions/mean_terminated_length": 542.625, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.925911559348332, "grad_norm": 9.831409454345703, "kl": 24.96875, "learning_rate": 5.651728743783544e-07, "loss": 0.4982, "num_tokens": 445734699.0, "reward": 0.17374999821186066, "reward_std": 0.1947653591632843, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.24013446271419525, "step": 2387 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2850.0, "completions/max_terminated_length": 2850.0, "completions/mean_length": 524.8125, "completions/mean_terminated_length": 524.8125, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.9262994569433669, "grad_norm": 7.130321502685547, "kl": 22.53125, "learning_rate": 5.644990120579828e-07, "loss": 0.4173, "num_tokens": 445829148.0, "reward": 0.29375001788139343, "reward_std": 0.25724664330482483, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.22499999403953552, "rewards/unicoder_reward_fn/std": 0.3675095736980438, "step": 2388 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3661.0, "completions/max_terminated_length": 3661.0, "completions/mean_length": 652.7875366210938, "completions/mean_terminated_length": 652.7875366210938, "completions/min_length": 45.0, "completions/min_terminated_length": 45.0, "epoch": 0.9266873545384019, "grad_norm": 7.0350446701049805, "kl": 24.0625, "learning_rate": 5.638286009419836e-07, "loss": 0.4383, "num_tokens": 445938939.0, "reward": 0.16375000774860382, "reward_std": 0.18806330859661102, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.20124612748622894, "step": 2389 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2826.0, "completions/max_terminated_length": 2826.0, "completions/mean_length": 674.2750244140625, "completions/mean_terminated_length": 674.2750244140625, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.9270752521334368, "grad_norm": 6.340228080749512, "kl": 29.3125, "learning_rate": 5.631616420890285e-07, "loss": 0.5637, "num_tokens": 446038513.0, "reward": 0.14625000953674316, "reward_std": 0.14776186645030975, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.0924999937415123, "rewards/unicoder_reward_fn/std": 0.21095022559165955, "step": 2390 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2701.0, "completions/mean_length": 687.0499877929688, "completions/mean_terminated_length": 554.2337646484375, "completions/min_length": 44.0, "completions/min_terminated_length": 44.0, "epoch": 0.9274631497284717, "grad_norm": 6.362987995147705, "kl": 25.453125, "learning_rate": 5.624981365523359e-07, "loss": 0.508, "num_tokens": 446144037.0, "reward": 0.20624999701976776, "reward_std": 0.18335631489753723, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.24275581538677216, "step": 2391 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2242.0, "completions/max_terminated_length": 2242.0, "completions/mean_length": 436.6499938964844, "completions/mean_terminated_length": 436.6499938964844, "completions/min_length": 53.0, "completions/min_terminated_length": 53.0, "epoch": 0.9278510473235065, "grad_norm": 11.443100929260254, "kl": 19.8125, "learning_rate": 5.618380853796714e-07, "loss": 0.3519, "num_tokens": 446230929.0, "reward": 0.2474999874830246, "reward_std": 0.2357218712568283, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.28638955950737, "step": 2392 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1981.0, "completions/mean_length": 650.5, "completions/mean_terminated_length": 606.8861083984375, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.9282389449185415, "grad_norm": 6.034416198730469, "kl": 32.6875, "learning_rate": 5.611814896133466e-07, "loss": 0.5429, "num_tokens": 446329857.0, "reward": 0.17375001311302185, "reward_std": 0.18553689122200012, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.22466571629047394, "step": 2393 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2486.0, "completions/max_terminated_length": 2486.0, "completions/mean_length": 543.2374877929688, "completions/mean_terminated_length": 543.2374877929688, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.9286268425135764, "grad_norm": 4.240497589111328, "kl": 28.78125, "learning_rate": 5.605283502902152e-07, "loss": 0.4976, "num_tokens": 446424988.0, "reward": 0.17625002562999725, "reward_std": 0.17791152000427246, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.233628511428833, "step": 2394 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3505.0, "completions/mean_length": 604.0625, "completions/mean_terminated_length": 559.8607788085938, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.9290147401086113, "grad_norm": 13.641584396362305, "kl": 29.8125, "learning_rate": 5.598786684416738e-07, "loss": 0.6466, "num_tokens": 446529473.0, "reward": 0.20875000953674316, "reward_std": 0.19085288047790527, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.1935020387172699, "step": 2395 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3220.0, "completions/max_terminated_length": 3220.0, "completions/mean_length": 659.4500122070312, "completions/mean_terminated_length": 659.4500122070312, "completions/min_length": 56.0, "completions/min_terminated_length": 56.0, "epoch": 0.9294026377036463, "grad_norm": 5.260269641876221, "kl": 39.875, "learning_rate": 5.592324450936587e-07, "loss": 0.8221, "num_tokens": 446620573.0, "reward": 0.20000003278255463, "reward_std": 0.17312245070934296, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.20768465101718903, "step": 2396 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2902.0, "completions/mean_length": 552.2000122070312, "completions/mean_terminated_length": 414.1298828125, "completions/min_length": 57.0, "completions/min_terminated_length": 57.0, "epoch": 0.9297905352986812, "grad_norm": 11.10726547241211, "kl": 24.625, "learning_rate": 5.585896812666443e-07, "loss": 0.5018, "num_tokens": 446713685.0, "reward": 0.2587500214576721, "reward_std": 0.20149722695350647, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.2525690793991089, "step": 2397 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2777.0, "completions/max_terminated_length": 2777.0, "completions/mean_length": 520.4625244140625, "completions/mean_terminated_length": 520.4625244140625, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "epoch": 0.930178432893716, "grad_norm": 11.076870918273926, "kl": 31.90625, "learning_rate": 5.57950377975643e-07, "loss": 0.6026, "num_tokens": 446802202.0, "reward": 0.2525000274181366, "reward_std": 0.18564820289611816, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.24360166490077972, "step": 2398 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1605.0, "completions/max_terminated_length": 1605.0, "completions/mean_length": 322.7749938964844, "completions/mean_terminated_length": 322.7749938964844, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.9305663304887509, "grad_norm": 8.47218132019043, "kl": 16.015625, "learning_rate": 5.573145362302013e-07, "loss": 0.2531, "num_tokens": 446875536.0, "reward": 0.4100000560283661, "reward_std": 0.26207754015922546, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.32499998807907104, "rewards/unicoder_reward_fn/std": 0.30585014820098877, "step": 2399 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2213.0, "completions/mean_length": 553.5125122070312, "completions/mean_terminated_length": 508.6708984375, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.9309542280837859, "grad_norm": 7.626381874084473, "kl": 34.5625, "learning_rate": 5.56682157034401e-07, "loss": 0.6538, "num_tokens": 446968593.0, "reward": 0.26875001192092896, "reward_std": 0.15704984962940216, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.19750000536441803, "rewards/unicoder_reward_fn/std": 0.22387553751468658, "step": 2400 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 4040.0, "completions/mean_length": 671.4249877929688, "completions/mean_terminated_length": 583.6154174804688, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.9313421256788208, "grad_norm": 4.090156078338623, "kl": 35.90625, "learning_rate": 5.560532413868545e-07, "loss": 0.5991, "num_tokens": 447072219.0, "reward": 0.23625002801418304, "reward_std": 0.17089568078517914, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.16500000655651093, "rewards/unicoder_reward_fn/std": 0.26003894209861755, "step": 2401 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3976.0, "completions/mean_length": 630.7625122070312, "completions/mean_terminated_length": 586.8987426757812, "completions/min_length": 72.0, "completions/min_terminated_length": 72.0, "epoch": 0.9317300232738557, "grad_norm": 7.687890529632568, "kl": 43.03125, "learning_rate": 5.554277902807053e-07, "loss": 0.6921, "num_tokens": 447173744.0, "reward": 0.18250001966953278, "reward_std": 0.18224100768566132, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.21955075860023499, "step": 2402 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3681.0, "completions/mean_length": 594.1375122070312, "completions/mean_terminated_length": 504.3461608886719, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.9321179208688906, "grad_norm": 3.8669216632843018, "kl": 40.6875, "learning_rate": 5.548058047036266e-07, "loss": 0.7025, "num_tokens": 447269747.0, "reward": 0.23375001549720764, "reward_std": 0.2118714600801468, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.16499999165534973, "rewards/unicoder_reward_fn/std": 0.23551718890666962, "step": 2403 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2275.0, "completions/max_terminated_length": 2275.0, "completions/mean_length": 410.125, "completions/mean_terminated_length": 410.125, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.9325058184639256, "grad_norm": 6.431021213531494, "kl": 28.953125, "learning_rate": 5.541872856378183e-07, "loss": 0.5557, "num_tokens": 447351885.0, "reward": 0.2512500286102295, "reward_std": 0.17230771481990814, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.17249999940395355, "rewards/unicoder_reward_fn/std": 0.24390028417110443, "step": 2404 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 4044.0, "completions/mean_length": 605.2374877929688, "completions/mean_terminated_length": 515.7307739257812, "completions/min_length": 77.0, "completions/min_terminated_length": 77.0, "epoch": 0.9328937160589604, "grad_norm": 3.9047927856445312, "kl": 45.375, "learning_rate": 5.535722340600065e-07, "loss": 0.877, "num_tokens": 447441800.0, "reward": 0.2550000250339508, "reward_std": 0.22428198158740997, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.18500001728534698, "rewards/unicoder_reward_fn/std": 0.25011390447616577, "step": 2405 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3319.0, "completions/mean_length": 704.625, "completions/mean_terminated_length": 661.6962280273438, "completions/min_length": 80.0, "completions/min_terminated_length": 80.0, "epoch": 0.9332816136539953, "grad_norm": 11.836977005004883, "kl": 49.53125, "learning_rate": 5.529606509414416e-07, "loss": 0.7499, "num_tokens": 447541210.0, "reward": 0.23375001549720764, "reward_std": 0.22158494591712952, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.16750000417232513, "rewards/unicoder_reward_fn/std": 0.26133784651756287, "step": 2406 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2959.0, "completions/max_terminated_length": 2959.0, "completions/mean_length": 563.1000366210938, "completions/mean_terminated_length": 563.1000366210938, "completions/min_length": 158.0, "completions/min_terminated_length": 158.0, "epoch": 0.9336695112490303, "grad_norm": 4.837055683135986, "kl": 42.8125, "learning_rate": 5.523525372478966e-07, "loss": 0.7379, "num_tokens": 447640874.0, "reward": 0.1899999976158142, "reward_std": 0.16887304186820984, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.20812606811523438, "step": 2407 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4042.0, "completions/max_terminated_length": 4042.0, "completions/mean_length": 478.20001220703125, "completions/mean_terminated_length": 478.20001220703125, "completions/min_length": 56.0, "completions/min_terminated_length": 56.0, "epoch": 0.9340574088440652, "grad_norm": 3.274435520172119, "kl": 34.4375, "learning_rate": 5.517478939396662e-07, "loss": 0.4909, "num_tokens": 447727978.0, "reward": 0.2537500262260437, "reward_std": 0.21886157989501953, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.17749999463558197, "rewards/unicoder_reward_fn/std": 0.24023984372615814, "step": 2408 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3468.0, "completions/max_terminated_length": 3468.0, "completions/mean_length": 641.1375122070312, "completions/mean_terminated_length": 641.1375122070312, "completions/min_length": 147.0, "completions/min_terminated_length": 147.0, "epoch": 0.9344453064391001, "grad_norm": 11.467655181884766, "kl": 47.375, "learning_rate": 5.511467219715648e-07, "loss": 0.727, "num_tokens": 447816133.0, "reward": 0.2587500214576721, "reward_std": 0.17673666775226593, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.18500001728534698, "rewards/unicoder_reward_fn/std": 0.2376573532819748, "step": 2409 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3763.0, "completions/max_terminated_length": 3763.0, "completions/mean_length": 607.9375, "completions/mean_terminated_length": 607.9375, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.934833204034135, "grad_norm": 8.826173782348633, "kl": 46.5625, "learning_rate": 5.505490222929247e-07, "loss": 0.7525, "num_tokens": 447906864.0, "reward": 0.2525000274181366, "reward_std": 0.2349713146686554, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.255261093378067, "step": 2410 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2783.0, "completions/max_terminated_length": 2783.0, "completions/mean_length": 505.6750183105469, "completions/mean_terminated_length": 505.6750183105469, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.9352211016291699, "grad_norm": 13.110821723937988, "kl": 31.65625, "learning_rate": 5.499547958475956e-07, "loss": 0.4091, "num_tokens": 447993294.0, "reward": 0.1600000113248825, "reward_std": 0.12144718319177628, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.08000000566244125, "rewards/unicoder_reward_fn/std": 0.15783344209194183, "step": 2411 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3861.0, "completions/mean_length": 686.4125366210938, "completions/mean_terminated_length": 643.253173828125, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 0.9356089992242048, "grad_norm": 28.502845764160156, "kl": 54.6875, "learning_rate": 5.493640435739421e-07, "loss": 0.7833, "num_tokens": 448095335.0, "reward": 0.1600000113248825, "reward_std": 0.1576678305864334, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.20126184821128845, "step": 2412 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3596.0, "completions/max_terminated_length": 3596.0, "completions/mean_length": 501.8000183105469, "completions/mean_terminated_length": 501.8000183105469, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.9359968968192397, "grad_norm": 2.3749876022338867, "kl": 34.21875, "learning_rate": 5.487767664048427e-07, "loss": 0.6508, "num_tokens": 448187623.0, "reward": 0.257500022649765, "reward_std": 0.17268793284893036, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.17500001192092896, "rewards/unicoder_reward_fn/std": 0.22809059917926788, "step": 2413 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3942.0, "completions/mean_length": 711.3875122070312, "completions/mean_terminated_length": 624.6026000976562, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.9363847944142746, "grad_norm": 3.569035291671753, "kl": 39.4375, "learning_rate": 5.481929652676879e-07, "loss": 0.7294, "num_tokens": 448297198.0, "reward": 0.3362500071525574, "reward_std": 0.23677395284175873, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.26249998807907104, "rewards/unicoder_reward_fn/std": 0.3509030044078827, "step": 2414 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3879.0, "completions/max_terminated_length": 3879.0, "completions/mean_length": 539.4000244140625, "completions/mean_terminated_length": 539.4000244140625, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "epoch": 0.9367726920093096, "grad_norm": 11.268847465515137, "kl": 29.4375, "learning_rate": 5.476126410843797e-07, "loss": 0.6045, "num_tokens": 448390886.0, "reward": 0.3400000035762787, "reward_std": 0.1837848275899887, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.2600000202655792, "rewards/unicoder_reward_fn/std": 0.30423176288604736, "step": 2415 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3801.0, "completions/max_terminated_length": 3801.0, "completions/mean_length": 544.3624877929688, "completions/mean_terminated_length": 544.3624877929688, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.9371605896043445, "grad_norm": 12.741057395935059, "kl": 47.1875, "learning_rate": 5.470357947713292e-07, "loss": 0.7396, "num_tokens": 448484963.0, "reward": 0.21000002324581146, "reward_std": 0.20864133536815643, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.4713299572467804, "rewards/unicoder_reward_fn/mean": 0.14250001311302185, "rewards/unicoder_reward_fn/std": 0.25694161653518677, "step": 2416 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3529.0, "completions/mean_length": 569.4750366210938, "completions/mean_terminated_length": 479.0513000488281, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.9375484871993793, "grad_norm": 4.683617115020752, "kl": 35.5625, "learning_rate": 5.464624272394551e-07, "loss": 0.6269, "num_tokens": 448574561.0, "reward": 0.29625001549720764, "reward_std": 0.17005054652690887, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.2199999988079071, "rewards/unicoder_reward_fn/std": 0.2803252935409546, "step": 2417 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3688.0, "completions/mean_length": 559.6000366210938, "completions/mean_terminated_length": 468.923095703125, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "epoch": 0.9379363847944143, "grad_norm": 9.3636474609375, "kl": 28.734375, "learning_rate": 5.458925393941834e-07, "loss": 0.5657, "num_tokens": 448668321.0, "reward": 0.30250000953674316, "reward_std": 0.1929663121700287, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.2199999988079071, "rewards/unicoder_reward_fn/std": 0.23940858244895935, "step": 2418 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2170.0, "completions/mean_length": 566.8250122070312, "completions/mean_terminated_length": 522.1519165039062, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.9383242823894492, "grad_norm": 4.946855068206787, "kl": 32.515625, "learning_rate": 5.453261321354446e-07, "loss": 0.6026, "num_tokens": 448762643.0, "reward": 0.257500022649765, "reward_std": 0.20214013755321503, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.17750000953674316, "rewards/unicoder_reward_fn/std": 0.26812759041786194, "step": 2419 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3575.0, "completions/mean_length": 718.4874877929688, "completions/mean_terminated_length": 675.7341918945312, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.9387121799844841, "grad_norm": 3.8794236183166504, "kl": 38.4375, "learning_rate": 5.447632063576734e-07, "loss": 0.7207, "num_tokens": 448870282.0, "reward": 0.29250001907348633, "reward_std": 0.23689019680023193, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.2175000011920929, "rewards/unicoder_reward_fn/std": 0.310481458902359, "step": 2420 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 3503.0, "completions/mean_length": 671.9625244140625, "completions/mean_terminated_length": 538.5584106445312, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.939100077579519, "grad_norm": 4.829666614532471, "kl": 34.78125, "learning_rate": 5.442037629498059e-07, "loss": 0.6111, "num_tokens": 448976919.0, "reward": 0.27000001072883606, "reward_std": 0.20775914192199707, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.19249999523162842, "rewards/unicoder_reward_fn/std": 0.2390778660774231, "step": 2421 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2101.0, "completions/mean_length": 573.5499877929688, "completions/mean_terminated_length": 528.9620361328125, "completions/min_length": 191.0, "completions/min_terminated_length": 191.0, "epoch": 0.939487975174554, "grad_norm": 3.4694466590881348, "kl": 36.71875, "learning_rate": 5.436478027952798e-07, "loss": 0.593, "num_tokens": 449075683.0, "reward": 0.3150000274181366, "reward_std": 0.2770536541938782, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.24750001728534698, "rewards/unicoder_reward_fn/std": 0.32490506768226624, "step": 2422 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.050000000000000044, "completions/max_length": 4096.0, "completions/max_terminated_length": 3371.0, "completions/mean_length": 907.4125366210938, "completions/mean_terminated_length": 739.5921020507812, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.9398758727695888, "grad_norm": 6.594700336456299, "kl": 46.9375, "learning_rate": 5.430953267720325e-07, "loss": 0.7257, "num_tokens": 449199756.0, "reward": 0.21375000476837158, "reward_std": 0.18513445556163788, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.21404778957366943, "step": 2423 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2961.0, "completions/mean_length": 732.0, "completions/mean_terminated_length": 689.417724609375, "completions/min_length": 56.0, "completions/min_terminated_length": 56.0, "epoch": 0.9402637703646237, "grad_norm": 11.997736930847168, "kl": 49.125, "learning_rate": 5.425463357524986e-07, "loss": 0.7005, "num_tokens": 449298284.0, "reward": 0.20499999821186066, "reward_std": 0.21587203443050385, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2661968469619751, "step": 2424 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3197.0, "completions/mean_length": 754.0499877929688, "completions/mean_terminated_length": 668.3590087890625, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "epoch": 0.9406516679596586, "grad_norm": 9.253593444824219, "kl": 37.1875, "learning_rate": 5.420008306036098e-07, "loss": 0.5393, "num_tokens": 449409656.0, "reward": 0.18125002086162567, "reward_std": 0.1364014744758606, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.2132331132888794, "step": 2425 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3894.0, "completions/max_terminated_length": 3894.0, "completions/mean_length": 795.0750122070312, "completions/mean_terminated_length": 795.0750122070312, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.9410395655546936, "grad_norm": 21.857440948486328, "kl": 48.9375, "learning_rate": 5.414588121867933e-07, "loss": 0.5902, "num_tokens": 449529958.0, "reward": 0.09875000268220901, "reward_std": 0.10121400654315948, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.03500000387430191, "rewards/unicoder_reward_fn/std": 0.08282969892024994, "step": 2426 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3688.0, "completions/mean_length": 1027.0750732421875, "completions/mean_terminated_length": 988.2278442382812, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.9414274631497285, "grad_norm": 36.68267059326172, "kl": 64.1875, "learning_rate": 5.409202813579703e-07, "loss": 0.7984, "num_tokens": 449667188.0, "reward": 0.09375, "reward_std": 0.14527182281017303, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.05000000074505806, "rewards/unicoder_reward_fn/std": 0.15751230716705322, "step": 2427 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3653.0, "completions/max_terminated_length": 3653.0, "completions/mean_length": 761.7000122070312, "completions/mean_terminated_length": 761.7000122070312, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.9418153607447634, "grad_norm": 21.366260528564453, "kl": 48.6875, "learning_rate": 5.403852389675544e-07, "loss": 0.618, "num_tokens": 449772068.0, "reward": 0.11375000327825546, "reward_std": 0.11564024537801743, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.05250000208616257, "rewards/unicoder_reward_fn/std": 0.126265749335289, "step": 2428 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3909.0, "completions/max_terminated_length": 3909.0, "completions/mean_length": 707.0250244140625, "completions/mean_terminated_length": 707.0250244140625, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.9422032583397983, "grad_norm": 4.016982078552246, "kl": 36.625, "learning_rate": 5.398536858604507e-07, "loss": 0.5823, "num_tokens": 449864502.0, "reward": 0.23750002682209015, "reward_std": 0.19048067927360535, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.2683163583278656, "step": 2429 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3245.0, "completions/max_terminated_length": 3245.0, "completions/mean_length": 609.1875, "completions/mean_terminated_length": 609.1875, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.9425911559348332, "grad_norm": 4.031374454498291, "kl": 32.625, "learning_rate": 5.393256228760536e-07, "loss": 0.4831, "num_tokens": 449957485.0, "reward": 0.22750000655651093, "reward_std": 0.17507247626781464, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.23808306455612183, "step": 2430 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2937.0, "completions/mean_length": 776.5750122070312, "completions/mean_terminated_length": 734.5570068359375, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.9429790535298681, "grad_norm": 10.453817367553711, "kl": 42.8125, "learning_rate": 5.388010508482474e-07, "loss": 0.6285, "num_tokens": 450062611.0, "reward": 0.15000000596046448, "reward_std": 0.15179458260536194, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.19297799468040466, "step": 2431 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3568.0, "completions/mean_length": 695.8250122070312, "completions/mean_terminated_length": 652.7848510742188, "completions/min_length": 43.0, "completions/min_terminated_length": 43.0, "epoch": 0.943366951124903, "grad_norm": 3.851907730102539, "kl": 39.0, "learning_rate": 5.382799706054024e-07, "loss": 0.7485, "num_tokens": 450167333.0, "reward": 0.17750000953674316, "reward_std": 0.17486055195331573, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.22024153172969818, "step": 2432 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3851.0, "completions/max_terminated_length": 3851.0, "completions/mean_length": 634.3875122070312, "completions/mean_terminated_length": 634.3875122070312, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.943754848719938, "grad_norm": 6.401754379272461, "kl": 39.8125, "learning_rate": 5.377623829703754e-07, "loss": 0.6843, "num_tokens": 450273436.0, "reward": 0.16250000894069672, "reward_std": 0.12391306459903717, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.16226248443126678, "step": 2433 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3368.0, "completions/max_terminated_length": 3368.0, "completions/mean_length": 695.9750366210938, "completions/mean_terminated_length": 695.9750366210938, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.9441427463149729, "grad_norm": 10.92895793914795, "kl": 40.875, "learning_rate": 5.372482887605084e-07, "loss": 0.6543, "num_tokens": 450377370.0, "reward": 0.2162499874830246, "reward_std": 0.1780303716659546, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.3030540645122528, "step": 2434 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2656.0, "completions/max_terminated_length": 2656.0, "completions/mean_length": 505.8999938964844, "completions/mean_terminated_length": 505.8999938964844, "completions/min_length": 138.0, "completions/min_terminated_length": 138.0, "epoch": 0.9445306439100077, "grad_norm": 8.318629264831543, "kl": 23.75, "learning_rate": 5.367376887876257e-07, "loss": 0.4615, "num_tokens": 450473074.0, "reward": 0.1912500113248825, "reward_std": 0.16846472024917603, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.17957404255867004, "step": 2435 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3870.0, "completions/max_terminated_length": 3870.0, "completions/mean_length": 530.5750122070312, "completions/mean_terminated_length": 530.5750122070312, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.9449185415050426, "grad_norm": 6.824469566345215, "kl": 24.8125, "learning_rate": 5.362305838580351e-07, "loss": 0.4636, "num_tokens": 450567080.0, "reward": 0.33625003695487976, "reward_std": 0.17584948241710663, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.2574999928474426, "rewards/unicoder_reward_fn/std": 0.298848032951355, "step": 2436 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2350.0, "completions/max_terminated_length": 2350.0, "completions/mean_length": 521.1749877929688, "completions/mean_terminated_length": 521.1749877929688, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.9453064391000776, "grad_norm": 48.468650817871094, "kl": 27.765625, "learning_rate": 5.357269747725235e-07, "loss": 0.5212, "num_tokens": 450644910.0, "reward": 0.3074999749660492, "reward_std": 0.22503240406513214, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.23249998688697815, "rewards/unicoder_reward_fn/std": 0.3205118477344513, "step": 2437 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3734.0, "completions/max_terminated_length": 3734.0, "completions/mean_length": 444.25, "completions/mean_terminated_length": 444.25, "completions/min_length": 44.0, "completions/min_terminated_length": 44.0, "epoch": 0.9456943366951125, "grad_norm": 10.890642166137695, "kl": 17.0625, "learning_rate": 5.352268623263595e-07, "loss": 0.2886, "num_tokens": 450733394.0, "reward": 0.3112500309944153, "reward_std": 0.191035196185112, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.23499998450279236, "rewards/unicoder_reward_fn/std": 0.2619788348674774, "step": 2438 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2475.0, "completions/max_terminated_length": 2475.0, "completions/mean_length": 571.7875366210938, "completions/mean_terminated_length": 571.7875366210938, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.9460822342901474, "grad_norm": 5.590626239776611, "kl": 29.59375, "learning_rate": 5.347302473092881e-07, "loss": 0.5908, "num_tokens": 450837609.0, "reward": 0.1599999964237213, "reward_std": 0.145721435546875, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.2107100784778595, "step": 2439 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3262.0, "completions/mean_length": 719.0250244140625, "completions/mean_terminated_length": 632.4359130859375, "completions/min_length": 77.0, "completions/min_terminated_length": 77.0, "epoch": 0.9464701318851824, "grad_norm": 7.837053298950195, "kl": 35.109375, "learning_rate": 5.342371305055323e-07, "loss": 0.7006, "num_tokens": 450947587.0, "reward": 0.2537500262260437, "reward_std": 0.2323053926229477, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.1875, "rewards/unicoder_reward_fn/std": 0.24719950556755066, "step": 2440 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3654.0, "completions/max_terminated_length": 3654.0, "completions/mean_length": 526.7625122070312, "completions/mean_terminated_length": 526.7625122070312, "completions/min_length": 22.0, "completions/min_terminated_length": 22.0, "epoch": 0.9468580294802172, "grad_norm": 12.289403915405273, "kl": 24.25, "learning_rate": 5.337475126937913e-07, "loss": 0.5722, "num_tokens": 451040528.0, "reward": 0.22875002026557922, "reward_std": 0.16356006264686584, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.23379100859165192, "step": 2441 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4056.0, "completions/max_terminated_length": 4056.0, "completions/mean_length": 496.3374938964844, "completions/mean_terminated_length": 496.3374938964844, "completions/min_length": 67.0, "completions/min_terminated_length": 67.0, "epoch": 0.9472459270752521, "grad_norm": 11.737290382385254, "kl": 25.25, "learning_rate": 5.332613946472381e-07, "loss": 0.4792, "num_tokens": 451124035.0, "reward": 0.2212500125169754, "reward_std": 0.17715804278850555, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.20745596289634705, "step": 2442 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2272.0, "completions/max_terminated_length": 2272.0, "completions/mean_length": 523.4500122070312, "completions/mean_terminated_length": 523.4500122070312, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.947633824670287, "grad_norm": 4.381513595581055, "kl": 30.796875, "learning_rate": 5.327787771335197e-07, "loss": 0.5425, "num_tokens": 451209087.0, "reward": 0.1562500149011612, "reward_std": 0.1419380158185959, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.08749999850988388, "rewards/unicoder_reward_fn/std": 0.17673194408416748, "step": 2443 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2523.0, "completions/mean_length": 675.0875244140625, "completions/mean_terminated_length": 587.371826171875, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.948021722265322, "grad_norm": 3.1611080169677734, "kl": 36.09375, "learning_rate": 5.322996609147545e-07, "loss": 0.6114, "num_tokens": 451320150.0, "reward": 0.18250000476837158, "reward_std": 0.21590550243854523, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.22432741522789001, "step": 2444 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3574.0, "completions/max_terminated_length": 3574.0, "completions/mean_length": 772.8125, "completions/mean_terminated_length": 772.8125, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.9484096198603569, "grad_norm": 5.066558837890625, "kl": 39.25, "learning_rate": 5.318240467475331e-07, "loss": 0.7078, "num_tokens": 451436959.0, "reward": 0.19875001907348633, "reward_std": 0.19191761314868927, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.2456585019826889, "step": 2445 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3836.0, "completions/mean_length": 788.1875, "completions/mean_terminated_length": 746.3164672851562, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "epoch": 0.9487975174553918, "grad_norm": 8.849461555480957, "kl": 46.9375, "learning_rate": 5.313519353829145e-07, "loss": 0.7593, "num_tokens": 451556390.0, "reward": 0.1887500137090683, "reward_std": 0.2269369661808014, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.14000000059604645, "rewards/unicoder_reward_fn/std": 0.24935360252857208, "step": 2446 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3416.0, "completions/mean_length": 768.8125, "completions/mean_terminated_length": 683.5, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.9491854150504266, "grad_norm": 5.157288551330566, "kl": 46.4375, "learning_rate": 5.308833275664274e-07, "loss": 0.7885, "num_tokens": 451659519.0, "reward": 0.24000000953674316, "reward_std": 0.2831694185733795, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.3248271346092224, "step": 2447 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3214.0, "completions/mean_length": 662.4375, "completions/mean_terminated_length": 618.9746704101562, "completions/min_length": 45.0, "completions/min_terminated_length": 45.0, "epoch": 0.9495733126454616, "grad_norm": 10.622920036315918, "kl": 41.4375, "learning_rate": 5.304182240380677e-07, "loss": 0.5479, "num_tokens": 451767786.0, "reward": 0.16250000894069672, "reward_std": 0.18828986585140228, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.20621667802333832, "step": 2448 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3253.0, "completions/mean_length": 790.5499877929688, "completions/mean_terminated_length": 705.7948608398438, "completions/min_length": 51.0, "completions/min_terminated_length": 51.0, "epoch": 0.9499612102404965, "grad_norm": 13.381734848022461, "kl": 50.3125, "learning_rate": 5.29956625532297e-07, "loss": 0.8032, "num_tokens": 451881478.0, "reward": 0.1262500137090683, "reward_std": 0.1370772421360016, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.1566864401102066, "step": 2449 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3225.0, "completions/mean_length": 643.9375, "completions/mean_terminated_length": 600.2405395507812, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.9503491078355314, "grad_norm": 8.426115989685059, "kl": 44.15625, "learning_rate": 5.294985327780424e-07, "loss": 0.6789, "num_tokens": 451969129.0, "reward": 0.23000001907348633, "reward_std": 0.2544630467891693, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.2906888425350189, "step": 2450 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2820.0, "completions/max_terminated_length": 2820.0, "completions/mean_length": 627.6124877929688, "completions/mean_terminated_length": 627.6124877929688, "completions/min_length": 63.0, "completions/min_terminated_length": 63.0, "epoch": 0.9507370054305664, "grad_norm": 2.824716567993164, "kl": 44.0625, "learning_rate": 5.290439464986948e-07, "loss": 0.788, "num_tokens": 452060314.0, "reward": 0.15625, "reward_std": 0.20187316834926605, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.23165613412857056, "step": 2451 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2533.0, "completions/max_terminated_length": 2533.0, "completions/mean_length": 616.4125366210938, "completions/mean_terminated_length": 616.4125366210938, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.9511249030256013, "grad_norm": 16.926069259643555, "kl": 43.875, "learning_rate": 5.285928674121083e-07, "loss": 0.5914, "num_tokens": 452161619.0, "reward": 0.1550000160932541, "reward_std": 0.2025465965270996, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.2247924506664276, "step": 2452 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2567.0, "completions/max_terminated_length": 2567.0, "completions/mean_length": 584.4750366210938, "completions/mean_terminated_length": 584.4750366210938, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.9515128006206361, "grad_norm": 2.8821985721588135, "kl": 39.9375, "learning_rate": 5.281452962305979e-07, "loss": 0.692, "num_tokens": 452262137.0, "reward": 0.18125000596046448, "reward_std": 0.19955120980739594, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.22657325863838196, "step": 2453 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3276.0, "completions/mean_length": 669.7374877929688, "completions/mean_terminated_length": 626.3671264648438, "completions/min_length": 45.0, "completions/min_terminated_length": 45.0, "epoch": 0.951900698215671, "grad_norm": 5.732163906097412, "kl": 42.4375, "learning_rate": 5.277012336609404e-07, "loss": 0.7265, "num_tokens": 452359308.0, "reward": 0.14249999821186066, "reward_std": 0.16741079092025757, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.08249999582767487, "rewards/unicoder_reward_fn/std": 0.21510964632034302, "step": 2454 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3741.0, "completions/max_terminated_length": 3741.0, "completions/mean_length": 603.6000366210938, "completions/mean_terminated_length": 603.6000366210938, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.952288595810706, "grad_norm": 5.508213043212891, "kl": 34.375, "learning_rate": 5.272606804043701e-07, "loss": 0.607, "num_tokens": 452455428.0, "reward": 0.17000000178813934, "reward_std": 0.16784702241420746, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.17413440346717834, "step": 2455 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3954.0, "completions/mean_length": 591.1875, "completions/mean_terminated_length": 546.8228149414062, "completions/min_length": 62.0, "completions/min_terminated_length": 62.0, "epoch": 0.9526764934057409, "grad_norm": 2.8818578720092773, "kl": 37.0625, "learning_rate": 5.268236371565816e-07, "loss": 0.6386, "num_tokens": 452550795.0, "reward": 0.2762500047683716, "reward_std": 0.29234200716018677, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.2175000011920929, "rewards/unicoder_reward_fn/std": 0.31210798025131226, "step": 2456 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3815.0, "completions/max_terminated_length": 3815.0, "completions/mean_length": 726.7125244140625, "completions/mean_terminated_length": 726.7125244140625, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.9530643910007758, "grad_norm": 11.07554817199707, "kl": 43.5, "learning_rate": 5.263901046077253e-07, "loss": 0.6563, "num_tokens": 452668332.0, "reward": 0.08999999612569809, "reward_std": 0.12304003536701202, "rewards/format_reward/mean": 0.42500001192092896, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.04750000312924385, "rewards/unicoder_reward_fn/std": 0.14317820966243744, "step": 2457 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2947.0, "completions/max_terminated_length": 2947.0, "completions/mean_length": 539.7999877929688, "completions/mean_terminated_length": 539.7999877929688, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.9534522885958107, "grad_norm": 11.098772048950195, "kl": 30.75, "learning_rate": 5.259600834424084e-07, "loss": 0.6474, "num_tokens": 452763996.0, "reward": 0.24250002205371857, "reward_std": 0.20012226700782776, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.17250001430511475, "rewards/unicoder_reward_fn/std": 0.2500506341457367, "step": 2458 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2780.0, "completions/max_terminated_length": 2780.0, "completions/mean_length": 559.125, "completions/mean_terminated_length": 559.125, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.9538401861908457, "grad_norm": 7.565894603729248, "kl": 30.84375, "learning_rate": 5.255335743396937e-07, "loss": 0.4673, "num_tokens": 452858614.0, "reward": 0.11749999970197678, "reward_std": 0.1117108091711998, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.054999999701976776, "rewards/unicoder_reward_fn/std": 0.10050506144762039, "step": 2459 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3145.0, "completions/mean_length": 599.9249877929688, "completions/mean_terminated_length": 555.6708984375, "completions/min_length": 62.0, "completions/min_terminated_length": 62.0, "epoch": 0.9542280837858805, "grad_norm": 6.5056586265563965, "kl": 31.1875, "learning_rate": 5.251105779730963e-07, "loss": 0.5454, "num_tokens": 452952128.0, "reward": 0.1900000125169754, "reward_std": 0.1766284555196762, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.18987004458904266, "step": 2460 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2871.0, "completions/max_terminated_length": 2871.0, "completions/mean_length": 460.4875183105469, "completions/mean_terminated_length": 460.4875183105469, "completions/min_length": 25.0, "completions/min_terminated_length": 25.0, "epoch": 0.9546159813809154, "grad_norm": 5.784087181091309, "kl": 22.78125, "learning_rate": 5.246910950105857e-07, "loss": 0.325, "num_tokens": 453043647.0, "reward": 0.1887500137090683, "reward_std": 0.14700666069984436, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.2010101079940796, "step": 2461 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2576.0, "completions/max_terminated_length": 2576.0, "completions/mean_length": 605.6375122070312, "completions/mean_terminated_length": 605.6375122070312, "completions/min_length": 56.0, "completions/min_terminated_length": 56.0, "epoch": 0.9550038789759504, "grad_norm": 9.974701881408691, "kl": 33.96875, "learning_rate": 5.242751261145825e-07, "loss": 0.4747, "num_tokens": 453126754.0, "reward": 0.09624999761581421, "reward_std": 0.10648422688245773, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.10744442045688629, "step": 2462 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2023.0, "completions/mean_length": 530.0750122070312, "completions/mean_terminated_length": 484.93670654296875, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.9553917765709853, "grad_norm": 5.0624680519104, "kl": 28.71875, "learning_rate": 5.238626719419587e-07, "loss": 0.5174, "num_tokens": 453218032.0, "reward": 0.21500001847743988, "reward_std": 0.20562846958637238, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.2565101981163025, "step": 2463 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2810.0, "completions/max_terminated_length": 2810.0, "completions/mean_length": 601.6124877929688, "completions/mean_terminated_length": 601.6124877929688, "completions/min_length": 181.0, "completions/min_terminated_length": 181.0, "epoch": 0.9557796741660202, "grad_norm": 2.9959704875946045, "kl": 30.875, "learning_rate": 5.234537331440354e-07, "loss": 0.5472, "num_tokens": 453324177.0, "reward": 0.19500000774860382, "reward_std": 0.18168798089027405, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.22585982084274292, "step": 2464 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2857.0, "completions/max_terminated_length": 2857.0, "completions/mean_length": 538.4625244140625, "completions/mean_terminated_length": 538.4625244140625, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.956167571761055, "grad_norm": 3.050632953643799, "kl": 27.625, "learning_rate": 5.23048310366583e-07, "loss": 0.4562, "num_tokens": 453402822.0, "reward": 0.1912500113248825, "reward_std": 0.2132166177034378, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.23248793184757233, "step": 2465 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3901.0, "completions/max_terminated_length": 3901.0, "completions/mean_length": 480.38751220703125, "completions/mean_terminated_length": 480.38751220703125, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "epoch": 0.95655546935609, "grad_norm": 7.290006160736084, "kl": 24.125, "learning_rate": 5.226464042498194e-07, "loss": 0.443, "num_tokens": 453491837.0, "reward": 0.23250003159046173, "reward_std": 0.19815325736999512, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.16249999403953552, "rewards/unicoder_reward_fn/std": 0.23620149493217468, "step": 2466 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2088.0, "completions/max_terminated_length": 2088.0, "completions/mean_length": 490.5, "completions/mean_terminated_length": 490.5, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.9569433669511249, "grad_norm": 3.8375730514526367, "kl": 26.9375, "learning_rate": 5.222480154284092e-07, "loss": 0.4511, "num_tokens": 453576725.0, "reward": 0.20750001072883606, "reward_std": 0.17982463538646698, "rewards/format_reward/mean": 0.675000011920929, "rewards/format_reward/std": 0.471329927444458, "rewards/unicoder_reward_fn/mean": 0.14000000059604645, "rewards/unicoder_reward_fn/std": 0.22138802707195282, "step": 2467 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2701.0, "completions/max_terminated_length": 2701.0, "completions/mean_length": 493.76251220703125, "completions/mean_terminated_length": 493.76251220703125, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.9573312645461598, "grad_norm": 17.282752990722656, "kl": 28.53125, "learning_rate": 5.218531445314628e-07, "loss": 0.4228, "num_tokens": 453669946.0, "reward": 0.11000001430511475, "reward_std": 0.11252196133136749, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.044999998062849045, "rewards/unicoder_reward_fn/std": 0.10542260110378265, "step": 2468 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3559.0, "completions/mean_length": 625.8125, "completions/mean_terminated_length": 581.8861083984375, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.9577191621411947, "grad_norm": 3.4249091148376465, "kl": 36.3125, "learning_rate": 5.214617921825349e-07, "loss": 0.5984, "num_tokens": 453772699.0, "reward": 0.17499999701976776, "reward_std": 0.15562301874160767, "rewards/format_reward/mean": 0.574999988079071, "rewards/format_reward/std": 0.4974619448184967, "rewards/unicoder_reward_fn/mean": 0.11749999225139618, "rewards/unicoder_reward_fn/std": 0.20048992335796356, "step": 2469 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2164.0, "completions/max_terminated_length": 2164.0, "completions/mean_length": 531.3375244140625, "completions/mean_terminated_length": 531.3375244140625, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.9581070597362297, "grad_norm": 5.102898597717285, "kl": 31.46875, "learning_rate": 5.210739589996246e-07, "loss": 0.5939, "num_tokens": 453867518.0, "reward": 0.17875000834465027, "reward_std": 0.17872023582458496, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.21793043613433838, "step": 2470 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3413.0, "completions/max_terminated_length": 3413.0, "completions/mean_length": 647.6749877929688, "completions/mean_terminated_length": 647.6749877929688, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.9584949573312646, "grad_norm": 7.312061786651611, "kl": 37.6875, "learning_rate": 5.206896455951735e-07, "loss": 0.596, "num_tokens": 453961972.0, "reward": 0.09874999523162842, "reward_std": 0.103340744972229, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.03999999910593033, "rewards/unicoder_reward_fn/std": 0.09756530076265335, "step": 2471 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3452.0, "completions/mean_length": 560.9625244140625, "completions/mean_terminated_length": 516.2152099609375, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.9588828549262994, "grad_norm": 3.1462254524230957, "kl": 27.8125, "learning_rate": 5.203088525760644e-07, "loss": 0.4625, "num_tokens": 454054161.0, "reward": 0.24375000596046448, "reward_std": 0.2323511689901352, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.17249999940395355, "rewards/unicoder_reward_fn/std": 0.2751179337501526, "step": 2472 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3013.0, "completions/max_terminated_length": 3013.0, "completions/mean_length": 481.3125, "completions/mean_terminated_length": 481.3125, "completions/min_length": 64.0, "completions/min_terminated_length": 64.0, "epoch": 0.9592707525213343, "grad_norm": 8.674196243286133, "kl": 29.8125, "learning_rate": 5.199315805436222e-07, "loss": 0.4974, "num_tokens": 454139834.0, "reward": 0.257500022649765, "reward_std": 0.2698681056499481, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.1925000250339508, "rewards/unicoder_reward_fn/std": 0.26897603273391724, "step": 2473 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1991.0, "completions/max_terminated_length": 1991.0, "completions/mean_length": 484.01251220703125, "completions/mean_terminated_length": 484.01251220703125, "completions/min_length": 178.0, "completions/min_terminated_length": 178.0, "epoch": 0.9596586501163693, "grad_norm": 5.203693389892578, "kl": 27.75, "learning_rate": 5.195578300936105e-07, "loss": 0.4868, "num_tokens": 454221411.0, "reward": 0.17750000953674316, "reward_std": 0.1449219286441803, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.1652769148349762, "step": 2474 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2918.0, "completions/max_terminated_length": 2918.0, "completions/mean_length": 603.5875244140625, "completions/mean_terminated_length": 603.5875244140625, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.9600465477114042, "grad_norm": 6.22816276550293, "kl": 39.21875, "learning_rate": 5.19187601816232e-07, "loss": 0.6209, "num_tokens": 454311706.0, "reward": 0.1912499964237213, "reward_std": 0.18746331334114075, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.23296384513378143, "step": 2475 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2647.0, "completions/max_terminated_length": 2647.0, "completions/mean_length": 612.375, "completions/mean_terminated_length": 612.375, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.9604344453064391, "grad_norm": 17.176733016967773, "kl": 41.75, "learning_rate": 5.188208962961283e-07, "loss": 0.6312, "num_tokens": 454410592.0, "reward": 0.07625000923871994, "reward_std": 0.08454542607069016, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.027499999850988388, "rewards/unicoder_reward_fn/std": 0.07626569271087646, "step": 2476 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3528.0, "completions/max_terminated_length": 3528.0, "completions/mean_length": 664.6375122070312, "completions/mean_terminated_length": 664.6375122070312, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.9608223429014741, "grad_norm": 16.626869201660156, "kl": 44.6875, "learning_rate": 5.184577141123771e-07, "loss": 0.6694, "num_tokens": 454508339.0, "reward": 0.1275000125169754, "reward_std": 0.15444009006023407, "rewards/format_reward/mean": 0.44999998807907104, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.08250000327825546, "rewards/unicoder_reward_fn/std": 0.20299966633319855, "step": 2477 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3275.0, "completions/max_terminated_length": 3275.0, "completions/mean_length": 656.0375366210938, "completions/mean_terminated_length": 656.0375366210938, "completions/min_length": 75.0, "completions/min_terminated_length": 75.0, "epoch": 0.9612102404965089, "grad_norm": 11.664070129394531, "kl": 40.34375, "learning_rate": 5.18098055838493e-07, "loss": 0.5715, "num_tokens": 454613390.0, "reward": 0.1275000125169754, "reward_std": 0.1604710817337036, "rewards/format_reward/mean": 0.5249999761581421, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.07500000298023224, "rewards/unicoder_reward_fn/std": 0.1724727302789688, "step": 2478 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3492.0, "completions/max_terminated_length": 3492.0, "completions/mean_length": 550.7125244140625, "completions/mean_terminated_length": 550.7125244140625, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.9615981380915438, "grad_norm": 6.6576056480407715, "kl": 38.0, "learning_rate": 5.177419220424251e-07, "loss": 0.5877, "num_tokens": 454704463.0, "reward": 0.20624999701976776, "reward_std": 0.19374997913837433, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.2555708587169647, "step": 2479 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3504.0, "completions/max_terminated_length": 3504.0, "completions/mean_length": 815.7000122070312, "completions/mean_terminated_length": 815.7000122070312, "completions/min_length": 93.0, "completions/min_terminated_length": 93.0, "epoch": 0.9619860356865787, "grad_norm": 25.897899627685547, "kl": 43.375, "learning_rate": 5.173893132865572e-07, "loss": 0.5474, "num_tokens": 454820807.0, "reward": 0.07875000685453415, "reward_std": 0.11875323206186295, "rewards/format_reward/mean": 0.4124999940395355, "rewards/format_reward/std": 0.4953900873661041, "rewards/unicoder_reward_fn/mean": 0.03750000149011612, "rewards/unicoder_reward_fn/std": 0.11946611851453781, "step": 2480 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3279.0, "completions/max_terminated_length": 3279.0, "completions/mean_length": 536.1625366210938, "completions/mean_terminated_length": 536.1625366210938, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "epoch": 0.9623739332816137, "grad_norm": 10.410140991210938, "kl": 29.21875, "learning_rate": 5.170402301277072e-07, "loss": 0.4313, "num_tokens": 454915460.0, "reward": 0.11124999821186066, "reward_std": 0.1147679015994072, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.04749999940395355, "rewards/unicoder_reward_fn/std": 0.14667193591594696, "step": 2481 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4051.0, "completions/max_terminated_length": 4051.0, "completions/mean_length": 584.75, "completions/mean_terminated_length": 584.75, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.9627618308766486, "grad_norm": 2.499032497406006, "kl": 27.421875, "learning_rate": 5.166946731171242e-07, "loss": 0.45, "num_tokens": 455006104.0, "reward": 0.1912500113248825, "reward_std": 0.1847236007452011, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.2119828164577484, "step": 2482 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2907.0, "completions/max_terminated_length": 2907.0, "completions/mean_length": 496.38751220703125, "completions/mean_terminated_length": 496.38751220703125, "completions/min_length": 121.0, "completions/min_terminated_length": 121.0, "epoch": 0.9631497284716835, "grad_norm": 7.906453609466553, "kl": 23.09375, "learning_rate": 5.16352642800491e-07, "loss": 0.4344, "num_tokens": 455089919.0, "reward": 0.1850000023841858, "reward_std": 0.15044979751110077, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.23117750883102417, "step": 2483 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3331.0, "completions/max_terminated_length": 3331.0, "completions/mean_length": 425.26251220703125, "completions/mean_terminated_length": 425.26251220703125, "completions/min_length": 118.0, "completions/min_terminated_length": 118.0, "epoch": 0.9635376260667183, "grad_norm": 3.0573272705078125, "kl": 14.890625, "learning_rate": 5.160141397179195e-07, "loss": 0.2544, "num_tokens": 455180876.0, "reward": 0.29249998927116394, "reward_std": 0.14654141664505005, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.21250000596046448, "rewards/unicoder_reward_fn/std": 0.27438077330589294, "step": 2484 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2860.0, "completions/max_terminated_length": 2860.0, "completions/mean_length": 479.3500061035156, "completions/mean_terminated_length": 479.3500061035156, "completions/min_length": 71.0, "completions/min_terminated_length": 71.0, "epoch": 0.9639255236617533, "grad_norm": 13.089987754821777, "kl": 20.59375, "learning_rate": 5.15679164403952e-07, "loss": 0.4481, "num_tokens": 455275600.0, "reward": 0.17875002324581146, "reward_std": 0.12628868222236633, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.18819597363471985, "step": 2485 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2040.0, "completions/max_terminated_length": 2040.0, "completions/mean_length": 410.38751220703125, "completions/mean_terminated_length": 410.38751220703125, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.9643134212567882, "grad_norm": 13.517622947692871, "kl": 10.97265625, "learning_rate": 5.153477173875609e-07, "loss": 0.2079, "num_tokens": 455356359.0, "reward": 0.27000001072883606, "reward_std": 0.17700964212417603, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.18000000715255737, "rewards/unicoder_reward_fn/std": 0.24975939095020294, "step": 2486 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3143.0, "completions/max_terminated_length": 3143.0, "completions/mean_length": 533.0, "completions/mean_terminated_length": 533.0, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.9647013188518231, "grad_norm": 8.874627113342285, "kl": 22.609375, "learning_rate": 5.150197991921457e-07, "loss": 0.4558, "num_tokens": 455446863.0, "reward": 0.19500000774860382, "reward_std": 0.16980181634426117, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.1996833086013794, "step": 2487 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2809.0, "completions/max_terminated_length": 2809.0, "completions/mean_length": 405.9875183105469, "completions/mean_terminated_length": 405.9875183105469, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.9650892164468581, "grad_norm": 17.197423934936523, "kl": 11.515625, "learning_rate": 5.146954103355345e-07, "loss": 0.3008, "num_tokens": 455527214.0, "reward": 0.3062500059604645, "reward_std": 0.13645033538341522, "rewards/format_reward/mean": 0.9125000238418579, "rewards/format_reward/std": 0.28434911370277405, "rewards/unicoder_reward_fn/mean": 0.2150000035762787, "rewards/unicoder_reward_fn/std": 0.20567886531352997, "step": 2488 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2461.0, "completions/max_terminated_length": 2461.0, "completions/mean_length": 384.625, "completions/mean_terminated_length": 384.625, "completions/min_length": 160.0, "completions/min_terminated_length": 160.0, "epoch": 0.965477114041893, "grad_norm": 18.88876724243164, "kl": 15.09375, "learning_rate": 5.143745513299812e-07, "loss": 0.3347, "num_tokens": 455602592.0, "reward": 0.3762500286102295, "reward_std": 0.2407073974609375, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.2900000214576721, "rewards/unicoder_reward_fn/std": 0.30004221200942993, "step": 2489 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2417.0, "completions/max_terminated_length": 2417.0, "completions/mean_length": 427.2250061035156, "completions/mean_terminated_length": 427.2250061035156, "completions/min_length": 131.0, "completions/min_terminated_length": 131.0, "epoch": 0.9658650116369278, "grad_norm": 3.8572895526885986, "kl": 16.03125, "learning_rate": 5.14057222682166e-07, "loss": 0.3091, "num_tokens": 455689522.0, "reward": 0.20624999701976776, "reward_std": 0.13570187985897064, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.22411571443080902, "step": 2490 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1958.0, "completions/max_terminated_length": 1958.0, "completions/mean_length": 360.2875061035156, "completions/mean_terminated_length": 360.2875061035156, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.9662529092319627, "grad_norm": 14.92352294921875, "kl": 15.921875, "learning_rate": 5.137434248931945e-07, "loss": 0.3986, "num_tokens": 455758513.0, "reward": 0.29625001549720764, "reward_std": 0.20104114711284637, "rewards/format_reward/mean": 0.862500011920929, "rewards/format_reward/std": 0.3465471565723419, "rewards/unicoder_reward_fn/mean": 0.20999999344348907, "rewards/unicoder_reward_fn/std": 0.30172500014305115, "step": 2491 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2653.0, "completions/max_terminated_length": 2653.0, "completions/mean_length": 451.625, "completions/mean_terminated_length": 451.625, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.9666408068269977, "grad_norm": 15.334178924560547, "kl": 19.5, "learning_rate": 5.134331584585961e-07, "loss": 0.4068, "num_tokens": 455846019.0, "reward": 0.33125001192092896, "reward_std": 0.22771234810352325, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.2474999874830246, "rewards/unicoder_reward_fn/std": 0.2955256998538971, "step": 2492 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3961.0, "completions/max_terminated_length": 3961.0, "completions/mean_length": 535.4125366210938, "completions/mean_terminated_length": 535.4125366210938, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.9670287044220326, "grad_norm": 2.4452333450317383, "kl": 25.5, "learning_rate": 5.131264238683234e-07, "loss": 0.3937, "num_tokens": 455943388.0, "reward": 0.20125003159046173, "reward_std": 0.20145557820796967, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.2428600788116455, "step": 2493 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1985.0, "completions/max_terminated_length": 1985.0, "completions/mean_length": 502.9375, "completions/mean_terminated_length": 502.9375, "completions/min_length": 57.0, "completions/min_terminated_length": 57.0, "epoch": 0.9674166020170675, "grad_norm": 4.464657306671143, "kl": 28.5, "learning_rate": 5.128232216067525e-07, "loss": 0.4706, "num_tokens": 456030479.0, "reward": 0.21125002205371857, "reward_std": 0.21718259155750275, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.14000001549720764, "rewards/unicoder_reward_fn/std": 0.25537264347076416, "step": 2494 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3273.0, "completions/mean_length": 553.1500244140625, "completions/mean_terminated_length": 508.3038024902344, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.9678044996121024, "grad_norm": 7.946094989776611, "kl": 34.9375, "learning_rate": 5.125235521526815e-07, "loss": 0.6769, "num_tokens": 456123155.0, "reward": 0.40125003457069397, "reward_std": 0.24890866875648499, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.33249998092651367, "rewards/unicoder_reward_fn/std": 0.36135295033454895, "step": 2495 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3622.0, "completions/max_terminated_length": 3622.0, "completions/mean_length": 497.13751220703125, "completions/mean_terminated_length": 497.13751220703125, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "epoch": 0.9681923972071373, "grad_norm": 9.269832611083984, "kl": 24.375, "learning_rate": 5.122274159793289e-07, "loss": 0.43, "num_tokens": 456206782.0, "reward": 0.3075000047683716, "reward_std": 0.25229430198669434, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.2224999964237213, "rewards/unicoder_reward_fn/std": 0.2968079149723053, "step": 2496 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3486.0, "completions/mean_length": 574.625, "completions/mean_terminated_length": 530.0506591796875, "completions/min_length": 101.0, "completions/min_terminated_length": 101.0, "epoch": 0.9685802948021722, "grad_norm": 3.300534248352051, "kl": 28.9375, "learning_rate": 5.119348135543341e-07, "loss": 0.4531, "num_tokens": 456302592.0, "reward": 0.17625001072883606, "reward_std": 0.12520204484462738, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.09750000387430191, "rewards/unicoder_reward_fn/std": 0.14580635726451874, "step": 2497 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1743.0, "completions/max_terminated_length": 1743.0, "completions/mean_length": 559.4000244140625, "completions/mean_terminated_length": 559.4000244140625, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.9689681923972071, "grad_norm": 16.109912872314453, "kl": 39.3125, "learning_rate": 5.116457453397568e-07, "loss": 0.5221, "num_tokens": 456397632.0, "reward": 0.18000000715255737, "reward_std": 0.16583342850208282, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.10999999940395355, "rewards/unicoder_reward_fn/std": 0.2314511239528656, "step": 2498 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2402.0, "completions/max_terminated_length": 2402.0, "completions/mean_length": 515.7250366210938, "completions/mean_terminated_length": 515.7250366210938, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.9693560899922421, "grad_norm": 2.929657459259033, "kl": 30.21875, "learning_rate": 5.113602117920747e-07, "loss": 0.5487, "num_tokens": 456490786.0, "reward": 0.22500000894069672, "reward_std": 0.14892695844173431, "rewards/format_reward/mean": 0.8999999761581421, "rewards/format_reward/std": 0.3018927574157715, "rewards/unicoder_reward_fn/mean": 0.13500000536441803, "rewards/unicoder_reward_fn/std": 0.24603182077407837, "step": 2499 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1687.0, "completions/mean_length": 559.1124877929688, "completions/mean_terminated_length": 514.341796875, "completions/min_length": 47.0, "completions/min_terminated_length": 47.0, "epoch": 0.969743987587277, "grad_norm": 11.895733833312988, "kl": 37.40625, "learning_rate": 5.110782133621843e-07, "loss": 0.6061, "num_tokens": 456586155.0, "reward": 0.16875000298023224, "reward_std": 0.1655917465686798, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.0950000062584877, "rewards/unicoder_reward_fn/std": 0.19610124826431274, "step": 2500 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3900.0, "completions/mean_length": 783.5125122070312, "completions/mean_terminated_length": 741.582275390625, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.9701318851823119, "grad_norm": 22.112550735473633, "kl": 49.25, "learning_rate": 5.107997504953999e-07, "loss": 0.6425, "num_tokens": 456706780.0, "reward": 0.12250000238418579, "reward_std": 0.12484605610370636, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.1620868742465973, "step": 2501 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4095.0, "completions/max_terminated_length": 4095.0, "completions/mean_length": 661.1625366210938, "completions/mean_terminated_length": 661.1625366210938, "completions/min_length": 35.0, "completions/min_terminated_length": 35.0, "epoch": 0.9705197827773467, "grad_norm": 3.2800285816192627, "kl": 39.09375, "learning_rate": 5.105248236314521e-07, "loss": 0.6461, "num_tokens": 456814465.0, "reward": 0.22750000655651093, "reward_std": 0.22901320457458496, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.30684182047843933, "step": 2502 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2592.0, "completions/max_terminated_length": 2592.0, "completions/mean_length": 639.0499877929688, "completions/mean_terminated_length": 639.0499877929688, "completions/min_length": 109.0, "completions/min_terminated_length": 109.0, "epoch": 0.9709076803723817, "grad_norm": 15.09201431274414, "kl": 44.5, "learning_rate": 5.10253433204488e-07, "loss": 0.6766, "num_tokens": 456909573.0, "reward": 0.20125000178813934, "reward_std": 0.21548961102962494, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.27760276198387146, "step": 2503 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3900.0, "completions/mean_length": 919.7625122070312, "completions/mean_terminated_length": 838.3204956054688, "completions/min_length": 126.0, "completions/min_terminated_length": 126.0, "epoch": 0.9712955779674166, "grad_norm": 35.62653732299805, "kl": 55.375, "learning_rate": 5.0998557964307e-07, "loss": 0.7126, "num_tokens": 457031154.0, "reward": 0.06499999761581421, "reward_std": 0.081800177693367, "rewards/format_reward/mean": 0.4749999940395355, "rewards/format_reward/std": 0.502525269985199, "rewards/unicoder_reward_fn/mean": 0.017500001937150955, "rewards/unicoder_reward_fn/std": 0.07919691503047943, "step": 2504 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3623.0, "completions/max_terminated_length": 3623.0, "completions/mean_length": 775.5499877929688, "completions/mean_terminated_length": 775.5499877929688, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.9716834755624515, "grad_norm": 18.35997772216797, "kl": 50.0, "learning_rate": 5.097212633701758e-07, "loss": 0.7946, "num_tokens": 457140702.0, "reward": 0.14625000953674316, "reward_std": 0.16993103921413422, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.08499999344348907, "rewards/unicoder_reward_fn/std": 0.17070220410823822, "step": 2505 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3762.0, "completions/max_terminated_length": 3762.0, "completions/mean_length": 627.1375122070312, "completions/mean_terminated_length": 627.1375122070312, "completions/min_length": 89.0, "completions/min_terminated_length": 89.0, "epoch": 0.9720713731574864, "grad_norm": 11.1488676071167, "kl": 39.6875, "learning_rate": 5.094604848031964e-07, "loss": 0.6532, "num_tokens": 457239985.0, "reward": 0.1237499937415123, "reward_std": 0.12782417237758636, "rewards/format_reward/mean": 0.6625000238418579, "rewards/format_reward/std": 0.47584035992622375, "rewards/unicoder_reward_fn/mean": 0.05750000476837158, "rewards/unicoder_reward_fn/std": 0.14300128817558289, "step": 2506 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3689.0, "completions/max_terminated_length": 3689.0, "completions/mean_length": 757.875, "completions/mean_terminated_length": 757.875, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.9724592707525214, "grad_norm": 16.874006271362305, "kl": 44.0625, "learning_rate": 5.092032443539371e-07, "loss": 0.6314, "num_tokens": 457346239.0, "reward": 0.17999999225139618, "reward_std": 0.21773076057434082, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.4871773421764374, "rewards/unicoder_reward_fn/mean": 0.11749999225139618, "rewards/unicoder_reward_fn/std": 0.25792503356933594, "step": 2507 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03749999999999998, "completions/max_length": 4096.0, "completions/max_terminated_length": 2552.0, "completions/mean_length": 665.8500366210938, "completions/mean_terminated_length": 532.207763671875, "completions/min_length": 42.0, "completions/min_terminated_length": 42.0, "epoch": 0.9728471683475562, "grad_norm": 6.703413009643555, "kl": 42.6875, "learning_rate": 5.089495424286158e-07, "loss": 0.684, "num_tokens": 457450083.0, "reward": 0.23125000298023224, "reward_std": 0.26222261786460876, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.17499999701976776, "rewards/unicoder_reward_fn/std": 0.33128058910369873, "step": 2508 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2323.0, "completions/max_terminated_length": 2323.0, "completions/mean_length": 512.5875244140625, "completions/mean_terminated_length": 512.5875244140625, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.9732350659425911, "grad_norm": 8.495894432067871, "kl": 31.0625, "learning_rate": 5.086993794278619e-07, "loss": 0.46, "num_tokens": 457540178.0, "reward": 0.18250000476837158, "reward_std": 0.1657661497592926, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.10499999672174454, "rewards/unicoder_reward_fn/std": 0.18548601865768433, "step": 2509 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3701.0, "completions/max_terminated_length": 3701.0, "completions/mean_length": 569.1124877929688, "completions/mean_terminated_length": 569.1124877929688, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.9736229635376261, "grad_norm": 3.9286208152770996, "kl": 33.40625, "learning_rate": 5.084527557467176e-07, "loss": 0.6345, "num_tokens": 457629955.0, "reward": 0.21875, "reward_std": 0.20637838542461395, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.442768931388855, "rewards/unicoder_reward_fn/mean": 0.14500001072883606, "rewards/unicoder_reward_fn/std": 0.24644306302070618, "step": 2510 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4076.0, "completions/max_terminated_length": 4076.0, "completions/mean_length": 546.0499877929688, "completions/mean_terminated_length": 546.0499877929688, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.974010861132661, "grad_norm": 3.5894434452056885, "kl": 32.03125, "learning_rate": 5.08209671774635e-07, "loss": 0.547, "num_tokens": 457722351.0, "reward": 0.23000001907348633, "reward_std": 0.19232778251171112, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.25051844120025635, "step": 2511 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1995.0, "completions/max_terminated_length": 1995.0, "completions/mean_length": 659.4249877929688, "completions/mean_terminated_length": 659.4249877929688, "completions/min_length": 131.0, "completions/min_terminated_length": 131.0, "epoch": 0.9743987587276959, "grad_norm": 12.67935848236084, "kl": 42.8125, "learning_rate": 5.079701278954768e-07, "loss": 0.7159, "num_tokens": 457828633.0, "reward": 0.13125000894069672, "reward_std": 0.14301076531410217, "rewards/format_reward/mean": 0.5375000238418579, "rewards/format_reward/std": 0.5017374753952026, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.19998417794704437, "step": 2512 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2237.0, "completions/max_terminated_length": 2237.0, "completions/mean_length": 545.1875, "completions/mean_terminated_length": 545.1875, "completions/min_length": 89.0, "completions/min_terminated_length": 89.0, "epoch": 0.9747866563227308, "grad_norm": 2.6392877101898193, "kl": 32.8125, "learning_rate": 5.077341244875155e-07, "loss": 0.5803, "num_tokens": 457924224.0, "reward": 0.2200000137090683, "reward_std": 0.21402306854724884, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.1550000011920929, "rewards/unicoder_reward_fn/std": 0.2584765553474426, "step": 2513 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2099.0, "completions/max_terminated_length": 2099.0, "completions/mean_length": 548.6625366210938, "completions/mean_terminated_length": 548.6625366210938, "completions/min_length": 41.0, "completions/min_terminated_length": 41.0, "epoch": 0.9751745539177658, "grad_norm": 6.207053184509277, "kl": 31.1875, "learning_rate": 5.075016619234329e-07, "loss": 0.4948, "num_tokens": 458012421.0, "reward": 0.1600000113248825, "reward_std": 0.14203238487243652, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.08750000596046448, "rewards/unicoder_reward_fn/std": 0.185127392411232, "step": 2514 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2323.0, "completions/max_terminated_length": 2323.0, "completions/mean_length": 436.9250183105469, "completions/mean_terminated_length": 436.9250183105469, "completions/min_length": 50.0, "completions/min_terminated_length": 50.0, "epoch": 0.9755624515128006, "grad_norm": 16.841571807861328, "kl": 26.28125, "learning_rate": 5.072727405703189e-07, "loss": 0.5956, "num_tokens": 458102415.0, "reward": 0.3712500035762787, "reward_std": 0.27832096815109253, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.29499998688697815, "rewards/unicoder_reward_fn/std": 0.3001687228679657, "step": 2515 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3037.0, "completions/max_terminated_length": 3037.0, "completions/mean_length": 564.0875244140625, "completions/mean_terminated_length": 564.0875244140625, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "epoch": 0.9759503491078355, "grad_norm": 5.292913436889648, "kl": 27.65625, "learning_rate": 5.070473607896718e-07, "loss": 0.519, "num_tokens": 458200950.0, "reward": 0.15000000596046448, "reward_std": 0.11359250545501709, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.18958650529384613, "step": 2516 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3551.0, "completions/max_terminated_length": 3551.0, "completions/mean_length": 608.9375, "completions/mean_terminated_length": 608.9375, "completions/min_length": 45.0, "completions/min_terminated_length": 45.0, "epoch": 0.9763382467028704, "grad_norm": 17.396656036376953, "kl": 32.90625, "learning_rate": 5.06825522937397e-07, "loss": 0.4282, "num_tokens": 458296337.0, "reward": 0.13375000655651093, "reward_std": 0.15838117897510529, "rewards/format_reward/mean": 0.637499988079071, "rewards/format_reward/std": 0.48375508189201355, "rewards/unicoder_reward_fn/mean": 0.07000000774860382, "rewards/unicoder_reward_fn/std": 0.20401042699813843, "step": 2517 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2622.0, "completions/mean_length": 506.3500061035156, "completions/mean_terminated_length": 460.9114074707031, "completions/min_length": 93.0, "completions/min_terminated_length": 93.0, "epoch": 0.9767261442979054, "grad_norm": 3.71706485748291, "kl": 22.0625, "learning_rate": 5.066072273638063e-07, "loss": 0.4082, "num_tokens": 458387165.0, "reward": 0.20000003278255463, "reward_std": 0.14653228223323822, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.19794833660125732, "step": 2518 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2832.0, "completions/max_terminated_length": 2832.0, "completions/mean_length": 416.5249938964844, "completions/mean_terminated_length": 416.5249938964844, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.9771140418929403, "grad_norm": 20.549068450927734, "kl": 21.84375, "learning_rate": 5.063924744136187e-07, "loss": 0.5317, "num_tokens": 458467071.0, "reward": 0.22750000655651093, "reward_std": 0.16745850443840027, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.14750000834465027, "rewards/unicoder_reward_fn/std": 0.17644521594047546, "step": 2519 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1962.0, "completions/max_terminated_length": 1962.0, "completions/mean_length": 458.95001220703125, "completions/mean_terminated_length": 458.95001220703125, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "epoch": 0.9775019394879751, "grad_norm": 5.925112724304199, "kl": 20.640625, "learning_rate": 5.061812644259583e-07, "loss": 0.3133, "num_tokens": 458551083.0, "reward": 0.3537500500679016, "reward_std": 0.165181964635849, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.2750000059604645, "rewards/unicoder_reward_fn/std": 0.32666322588920593, "step": 2520 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3059.0, "completions/max_terminated_length": 3059.0, "completions/mean_length": 484.57501220703125, "completions/mean_terminated_length": 484.57501220703125, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.9778898370830101, "grad_norm": 17.236177444458008, "kl": 23.1875, "learning_rate": 5.059735977343546e-07, "loss": 0.508, "num_tokens": 458640465.0, "reward": 0.24625001847743988, "reward_std": 0.20605607330799103, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.16999998688697815, "rewards/unicoder_reward_fn/std": 0.24046368896961212, "step": 2521 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2847.0, "completions/max_terminated_length": 2847.0, "completions/mean_length": 560.4375, "completions/mean_terminated_length": 560.4375, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.978277734678045, "grad_norm": 9.479538917541504, "kl": 29.53125, "learning_rate": 5.057694746667416e-07, "loss": 0.5462, "num_tokens": 458736020.0, "reward": 0.2775000035762787, "reward_std": 0.23966236412525177, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.21250000596046448, "rewards/unicoder_reward_fn/std": 0.3090921640396118, "step": 2522 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3734.0, "completions/max_terminated_length": 3734.0, "completions/mean_length": 408.8374938964844, "completions/mean_terminated_length": 408.8374938964844, "completions/min_length": 164.0, "completions/min_terminated_length": 164.0, "epoch": 0.9786656322730799, "grad_norm": 11.54926586151123, "kl": 16.265625, "learning_rate": 5.05568895545457e-07, "loss": 0.341, "num_tokens": 458812647.0, "reward": 0.26875001192092896, "reward_std": 0.1654243767261505, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.21723231673240662, "step": 2523 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2883.0, "completions/mean_length": 613.7875366210938, "completions/mean_terminated_length": 569.7088623046875, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.9790535298681148, "grad_norm": 7.658037185668945, "kl": 23.03125, "learning_rate": 5.053718606872433e-07, "loss": 0.4593, "num_tokens": 458911990.0, "reward": 0.18125000596046448, "reward_std": 0.12370200455188751, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.10250000655651093, "rewards/unicoder_reward_fn/std": 0.15259361267089844, "step": 2524 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2219.0, "completions/max_terminated_length": 2219.0, "completions/mean_length": 477.9375, "completions/mean_terminated_length": 477.9375, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.9794414274631498, "grad_norm": 11.99586009979248, "kl": 23.09375, "learning_rate": 5.051783704032452e-07, "loss": 0.5251, "num_tokens": 458997145.0, "reward": 0.35499998927116394, "reward_std": 0.1731906682252884, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.27000001072883606, "rewards/unicoder_reward_fn/std": 0.2252424955368042, "step": 2525 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1651.0, "completions/max_terminated_length": 1651.0, "completions/mean_length": 460.20001220703125, "completions/mean_terminated_length": 460.20001220703125, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "epoch": 0.9798293250581847, "grad_norm": 7.458778381347656, "kl": 21.9375, "learning_rate": 5.049884249990101e-07, "loss": 0.3882, "num_tokens": 459074425.0, "reward": 0.3075000047683716, "reward_std": 0.2062445878982544, "rewards/format_reward/mean": 0.8500000238418579, "rewards/format_reward/std": 0.35932427644729614, "rewards/unicoder_reward_fn/mean": 0.2225000113248825, "rewards/unicoder_reward_fn/std": 0.31821295619010925, "step": 2526 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3119.0, "completions/max_terminated_length": 3119.0, "completions/mean_length": 413.9750061035156, "completions/mean_terminated_length": 413.9750061035156, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.9802172226532195, "grad_norm": 16.078638076782227, "kl": 20.609375, "learning_rate": 5.048020247744881e-07, "loss": 0.4144, "num_tokens": 459150727.0, "reward": 0.36625000834465027, "reward_std": 0.277203768491745, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.2874999940395355, "rewards/unicoder_reward_fn/std": 0.3115396797657013, "step": 2527 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2482.0, "completions/max_terminated_length": 2482.0, "completions/mean_length": 490.0874938964844, "completions/mean_terminated_length": 490.0874938964844, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "epoch": 0.9806051202482544, "grad_norm": 12.806673049926758, "kl": 29.75, "learning_rate": 5.046191700240301e-07, "loss": 0.6116, "num_tokens": 459234222.0, "reward": 0.29625001549720764, "reward_std": 0.23228268325328827, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.22499999403953552, "rewards/unicoder_reward_fn/std": 0.29230642318725586, "step": 2528 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3451.0, "completions/max_terminated_length": 3451.0, "completions/mean_length": 509.4624938964844, "completions/mean_terminated_length": 509.4624938964844, "completions/min_length": 93.0, "completions/min_terminated_length": 93.0, "epoch": 0.9809930178432894, "grad_norm": 6.136139392852783, "kl": 30.8125, "learning_rate": 5.04439861036389e-07, "loss": 0.5927, "num_tokens": 459314731.0, "reward": 0.27625003457069397, "reward_std": 0.2420826405286789, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.20250001549720764, "rewards/unicoder_reward_fn/std": 0.26718172430992126, "step": 2529 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2524.0, "completions/max_terminated_length": 2524.0, "completions/mean_length": 364.32501220703125, "completions/mean_terminated_length": 364.32501220703125, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.9813809154383243, "grad_norm": 6.452450275421143, "kl": 14.828125, "learning_rate": 5.042640980947183e-07, "loss": 0.2381, "num_tokens": 459393221.0, "reward": 0.24625001847743988, "reward_std": 0.10400787740945816, "rewards/format_reward/mean": 0.887499988079071, "rewards/format_reward/std": 0.3179742097854614, "rewards/unicoder_reward_fn/mean": 0.1574999988079071, "rewards/unicoder_reward_fn/std": 0.1734059453010559, "step": 2530 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3461.0, "completions/max_terminated_length": 3461.0, "completions/mean_length": 462.6875, "completions/mean_terminated_length": 462.6875, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "epoch": 0.9817688130333592, "grad_norm": 14.77818775177002, "kl": 23.84375, "learning_rate": 5.040918814765714e-07, "loss": 0.5156, "num_tokens": 459471388.0, "reward": 0.23250000178813934, "reward_std": 0.13231994211673737, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.15000000596046448, "rewards/unicoder_reward_fn/std": 0.18960317969322205, "step": 2531 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3058.0, "completions/mean_length": 481.63751220703125, "completions/mean_terminated_length": 435.8860778808594, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "epoch": 0.9821567106283942, "grad_norm": 9.826220512390137, "kl": 24.953125, "learning_rate": 5.039232114539019e-07, "loss": 0.4871, "num_tokens": 459558159.0, "reward": 0.24625001847743988, "reward_std": 0.16036182641983032, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.19985756278038025, "step": 2532 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2940.0, "completions/max_terminated_length": 2940.0, "completions/mean_length": 507.5625, "completions/mean_terminated_length": 507.5625, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "epoch": 0.982544608223429, "grad_norm": 4.147197723388672, "kl": 29.125, "learning_rate": 5.037580882930628e-07, "loss": 0.4899, "num_tokens": 459650260.0, "reward": 0.3062500059604645, "reward_std": 0.20166073739528656, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.22999998927116394, "rewards/unicoder_reward_fn/std": 0.2986891269683838, "step": 2533 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 2976.0, "completions/mean_length": 655.8875122070312, "completions/mean_terminated_length": 567.6795043945312, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "epoch": 0.9829325058184639, "grad_norm": 8.381826400756836, "kl": 45.875, "learning_rate": 5.035965122548063e-07, "loss": 0.6854, "num_tokens": 459747435.0, "reward": 0.2787500321865082, "reward_std": 0.18522246181964874, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.20999999344348907, "rewards/unicoder_reward_fn/std": 0.31163108348846436, "step": 2534 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3680.0, "completions/max_terminated_length": 3680.0, "completions/mean_length": 603.9125366210938, "completions/mean_terminated_length": 603.9125366210938, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.9833204034134988, "grad_norm": 5.599853515625, "kl": 40.625, "learning_rate": 5.034384835942826e-07, "loss": 0.6847, "num_tokens": 459840964.0, "reward": 0.1900000125169754, "reward_std": 0.15691879391670227, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.20048990845680237, "step": 2535 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1955.0, "completions/max_terminated_length": 1955.0, "completions/mean_length": 442.875, "completions/mean_terminated_length": 442.875, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.9837083010085338, "grad_norm": 5.476648330688477, "kl": 36.34375, "learning_rate": 5.032840025610412e-07, "loss": 0.6287, "num_tokens": 459921178.0, "reward": 0.33250001072883606, "reward_std": 0.276477575302124, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.2550000250339508, "rewards/unicoder_reward_fn/std": 0.35432231426239014, "step": 2536 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3505.0, "completions/max_terminated_length": 3505.0, "completions/mean_length": 619.8375244140625, "completions/mean_terminated_length": 619.8375244140625, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "epoch": 0.9840961986035687, "grad_norm": 5.789279460906982, "kl": 47.5625, "learning_rate": 5.03133069399028e-07, "loss": 0.835, "num_tokens": 460018077.0, "reward": 0.20375001430511475, "reward_std": 0.1770145744085312, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.13250000774860382, "rewards/unicoder_reward_fn/std": 0.2484254240989685, "step": 2537 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3780.0, "completions/mean_length": 702.9375, "completions/mean_terminated_length": 659.9873657226562, "completions/min_length": 36.0, "completions/min_terminated_length": 36.0, "epoch": 0.9844840961986036, "grad_norm": 15.064844131469727, "kl": 49.6875, "learning_rate": 5.029856843465875e-07, "loss": 0.7004, "num_tokens": 460128656.0, "reward": 0.17875002324581146, "reward_std": 0.1898636817932129, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.11750000715255737, "rewards/unicoder_reward_fn/std": 0.2768608033657074, "step": 2538 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3005.0, "completions/mean_length": 721.4249877929688, "completions/mean_terminated_length": 678.7088623046875, "completions/min_length": 65.0, "completions/min_terminated_length": 65.0, "epoch": 0.9848719937936384, "grad_norm": 18.48716163635254, "kl": 53.625, "learning_rate": 5.028418476364608e-07, "loss": 0.8191, "num_tokens": 460229642.0, "reward": 0.23250000178813934, "reward_std": 0.18306808173656464, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.17249999940395355, "rewards/unicoder_reward_fn/std": 0.30810773372650146, "step": 2539 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 3988.0, "completions/mean_length": 892.4249877929688, "completions/mean_terminated_length": 810.2820434570312, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.9852598913886734, "grad_norm": 16.490951538085938, "kl": 62.375, "learning_rate": 5.027015594957852e-07, "loss": 1.0508, "num_tokens": 460339340.0, "reward": 0.16625002026557922, "reward_std": 0.16247081756591797, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.10750000178813934, "rewards/unicoder_reward_fn/std": 0.19858679175376892, "step": 2540 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 4075.0, "completions/max_terminated_length": 4075.0, "completions/mean_length": 651.8125, "completions/mean_terminated_length": 651.8125, "completions/min_length": 152.0, "completions/min_terminated_length": 152.0, "epoch": 0.9856477889837083, "grad_norm": 17.630943298339844, "kl": 51.0625, "learning_rate": 5.025648201460959e-07, "loss": 0.7674, "num_tokens": 460435981.0, "reward": 0.14624999463558197, "reward_std": 0.12322944402694702, "rewards/format_reward/mean": 0.612500011920929, "rewards/format_reward/std": 0.4902530908584595, "rewards/unicoder_reward_fn/mean": 0.08500000089406967, "rewards/unicoder_reward_fn/std": 0.17653486132621765, "step": 2541 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2883.0, "completions/max_terminated_length": 2883.0, "completions/mean_length": 664.3624877929688, "completions/mean_terminated_length": 664.3624877929688, "completions/min_length": 136.0, "completions/min_terminated_length": 136.0, "epoch": 0.9860356865787432, "grad_norm": 18.530258178710938, "kl": 55.125, "learning_rate": 5.024316298033218e-07, "loss": 0.841, "num_tokens": 460537314.0, "reward": 0.17625001072883606, "reward_std": 0.18132583796977997, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.4992082417011261, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.2263636738061905, "step": 2542 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3499.0, "completions/mean_length": 727.375, "completions/mean_terminated_length": 684.7341918945312, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.9864235841737781, "grad_norm": 18.85887336730957, "kl": 57.5625, "learning_rate": 5.023019886777894e-07, "loss": 0.8842, "num_tokens": 460642816.0, "reward": 0.18375001847743988, "reward_std": 0.2076731026172638, "rewards/format_reward/mean": 0.48750001192092896, "rewards/format_reward/std": 0.5029973387718201, "rewards/unicoder_reward_fn/mean": 0.13499999046325684, "rewards/unicoder_reward_fn/std": 0.2695988714694977, "step": 2543 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3382.0, "completions/max_terminated_length": 3382.0, "completions/mean_length": 599.6625366210938, "completions/mean_terminated_length": 599.6625366210938, "completions/min_length": 59.0, "completions/min_terminated_length": 59.0, "epoch": 0.9868114817688131, "grad_norm": 3.3124475479125977, "kl": 37.375, "learning_rate": 5.021758969742196e-07, "loss": 0.7026, "num_tokens": 460740493.0, "reward": 0.2825000286102295, "reward_std": 0.20860052108764648, "rewards/format_reward/mean": 0.7749999761581421, "rewards/format_reward/std": 0.42021694779396057, "rewards/unicoder_reward_fn/mean": 0.20500001311302185, "rewards/unicoder_reward_fn/std": 0.26809218525886536, "step": 2544 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3192.0, "completions/max_terminated_length": 3192.0, "completions/mean_length": 749.3500366210938, "completions/mean_terminated_length": 749.3500366210938, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.9871993793638479, "grad_norm": 9.387686729431152, "kl": 49.5, "learning_rate": 5.020533548917279e-07, "loss": 0.8682, "num_tokens": 460842681.0, "reward": 0.21375000476837158, "reward_std": 0.2281012088060379, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.14249999821186066, "rewards/unicoder_reward_fn/std": 0.25694161653518677, "step": 2545 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2285.0, "completions/max_terminated_length": 2285.0, "completions/mean_length": 511.51251220703125, "completions/mean_terminated_length": 511.51251220703125, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.9875872769588828, "grad_norm": 3.5814096927642822, "kl": 29.3125, "learning_rate": 5.019343626238256e-07, "loss": 0.4547, "num_tokens": 460931434.0, "reward": 0.21125002205371857, "reward_std": 0.22233128547668457, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.2466997504234314, "step": 2546 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3824.0, "completions/mean_length": 801.8875122070312, "completions/mean_terminated_length": 760.1898803710938, "completions/min_length": 169.0, "completions/min_terminated_length": 169.0, "epoch": 0.9879751745539178, "grad_norm": 8.709131240844727, "kl": 47.1875, "learning_rate": 5.018189203584177e-07, "loss": 0.8419, "num_tokens": 461043961.0, "reward": 0.14249999821186066, "reward_std": 0.1479910910129547, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.07250000536441803, "rewards/unicoder_reward_fn/std": 0.17787423729896545, "step": 2547 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1955.0, "completions/max_terminated_length": 1955.0, "completions/mean_length": 496.3999938964844, "completions/mean_terminated_length": 496.3999938964844, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.9883630721489527, "grad_norm": 8.439908981323242, "kl": 28.75, "learning_rate": 5.01707028277803e-07, "loss": 0.5433, "num_tokens": 461133649.0, "reward": 0.20375001430511475, "reward_std": 0.1717177778482437, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.2167072594165802, "step": 2548 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1683.0, "completions/mean_length": 500.57501220703125, "completions/mean_terminated_length": 455.06329345703125, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.9887509697439876, "grad_norm": 4.86032772064209, "kl": 29.46875, "learning_rate": 5.015986865586743e-07, "loss": 0.4937, "num_tokens": 461215927.0, "reward": 0.22874999046325684, "reward_std": 0.1562034636735916, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.1525000035762787, "rewards/unicoder_reward_fn/std": 0.2814856469631195, "step": 2549 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2380.0, "completions/mean_length": 561.4750366210938, "completions/mean_terminated_length": 516.7341918945312, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "epoch": 0.9891388673390225, "grad_norm": 2.712918281555176, "kl": 24.125, "learning_rate": 5.014938953721183e-07, "loss": 0.3799, "num_tokens": 461312917.0, "reward": 0.1612500101327896, "reward_std": 0.11889948695898056, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.13684973120689392, "step": 2550 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3459.0, "completions/max_terminated_length": 3459.0, "completions/mean_length": 652.6000366210938, "completions/mean_terminated_length": 652.6000366210938, "completions/min_length": 31.0, "completions/min_terminated_length": 31.0, "epoch": 0.9895267649340574, "grad_norm": 3.928154945373535, "kl": 34.25, "learning_rate": 5.013926548836144e-07, "loss": 0.5885, "num_tokens": 461407701.0, "reward": 0.19500000774860382, "reward_std": 0.24301829934120178, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997891902923584, "rewards/unicoder_reward_fn/mean": 0.1300000101327896, "rewards/unicoder_reward_fn/std": 0.24256019294261932, "step": 2551 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3233.0, "completions/max_terminated_length": 3233.0, "completions/mean_length": 539.6000366210938, "completions/mean_terminated_length": 539.6000366210938, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.9899146625290923, "grad_norm": 10.697370529174805, "kl": 24.8125, "learning_rate": 5.01294965253035e-07, "loss": 0.556, "num_tokens": 461496117.0, "reward": 0.2212500125169754, "reward_std": 0.18391677737236023, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.23296384513378143, "step": 2552 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1974.0, "completions/max_terminated_length": 1974.0, "completions/mean_length": 499.1000061035156, "completions/mean_terminated_length": 499.1000061035156, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "epoch": 0.9903025601241272, "grad_norm": 6.139554023742676, "kl": 31.15625, "learning_rate": 5.012008266346457e-07, "loss": 0.5792, "num_tokens": 461584197.0, "reward": 0.2550000250339508, "reward_std": 0.23907189071178436, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.18250000476837158, "rewards/unicoder_reward_fn/std": 0.2988480031490326, "step": 2553 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2814.0, "completions/max_terminated_length": 2814.0, "completions/mean_length": 675.4125366210938, "completions/mean_terminated_length": 675.4125366210938, "completions/min_length": 110.0, "completions/min_terminated_length": 110.0, "epoch": 0.9906904577191621, "grad_norm": 6.20737886428833, "kl": 37.46875, "learning_rate": 5.011102391771039e-07, "loss": 0.5764, "num_tokens": 461680046.0, "reward": 0.19875000417232513, "reward_std": 0.2305753231048584, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.14000001549720764, "rewards/unicoder_reward_fn/std": 0.23898519575595856, "step": 2554 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.025000000000000022, "completions/max_length": 4096.0, "completions/max_terminated_length": 1971.0, "completions/mean_length": 650.6500244140625, "completions/mean_terminated_length": 562.3076782226562, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.9910783553141971, "grad_norm": 12.735006332397461, "kl": 40.375, "learning_rate": 5.0102320302346e-07, "loss": 0.6228, "num_tokens": 461778946.0, "reward": 0.1862500011920929, "reward_std": 0.18514445424079895, "rewards/format_reward/mean": 0.5874999761581421, "rewards/format_reward/std": 0.4953901469707489, "rewards/unicoder_reward_fn/mean": 0.1274999976158142, "rewards/unicoder_reward_fn/std": 0.2164003551006317, "step": 2555 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2275.0, "completions/max_terminated_length": 2275.0, "completions/mean_length": 523.7999877929688, "completions/mean_terminated_length": 523.7999877929688, "completions/min_length": 119.0, "completions/min_terminated_length": 119.0, "epoch": 0.991466252909232, "grad_norm": 6.696477890014648, "kl": 31.21875, "learning_rate": 5.009397183111557e-07, "loss": 0.4883, "num_tokens": 461867306.0, "reward": 0.14750000834465027, "reward_std": 0.13803832232952118, "rewards/format_reward/mean": 0.699999988079071, "rewards/format_reward/std": 0.46114879846572876, "rewards/unicoder_reward_fn/mean": 0.07750000059604645, "rewards/unicoder_reward_fn/std": 0.17282097041606903, "step": 2556 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2671.0, "completions/mean_length": 784.2999877929688, "completions/mean_terminated_length": 742.3797607421875, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.9918541505042668, "grad_norm": 15.954662322998047, "kl": 37.0625, "learning_rate": 5.008597851720245e-07, "loss": 0.5381, "num_tokens": 461983218.0, "reward": 0.19625000655651093, "reward_std": 0.17091555893421173, "rewards/format_reward/mean": 0.737500011920929, "rewards/format_reward/std": 0.4427689015865326, "rewards/unicoder_reward_fn/mean": 0.12250000238418579, "rewards/unicoder_reward_fn/std": 0.28727221488952637, "step": 2557 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2021.0, "completions/max_terminated_length": 2021.0, "completions/mean_length": 515.8875122070312, "completions/mean_terminated_length": 515.8875122070312, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.9922420480993018, "grad_norm": 9.627727508544922, "kl": 26.28125, "learning_rate": 5.007834037322922e-07, "loss": 0.4799, "num_tokens": 462065145.0, "reward": 0.21250000596046448, "reward_std": 0.19659917056560516, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.13750000298023224, "rewards/unicoder_reward_fn/std": 0.25376906991004944, "step": 2558 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3140.0, "completions/max_terminated_length": 3140.0, "completions/mean_length": 522.3375244140625, "completions/mean_terminated_length": 522.3375244140625, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.9926299456943367, "grad_norm": 10.838343620300293, "kl": 31.15625, "learning_rate": 5.007105741125752e-07, "loss": 0.4446, "num_tokens": 462153076.0, "reward": 0.1887500137090683, "reward_std": 0.16074678301811218, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.46643680334091187, "rewards/unicoder_reward_fn/mean": 0.12000000476837158, "rewards/unicoder_reward_fn/std": 0.28921544551849365, "step": 2559 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2840.0, "completions/max_terminated_length": 2840.0, "completions/mean_length": 583.5, "completions/mean_terminated_length": 583.5, "completions/min_length": 77.0, "completions/min_terminated_length": 77.0, "epoch": 0.9930178432893716, "grad_norm": 8.413047790527344, "kl": 31.125, "learning_rate": 5.006412964278816e-07, "loss": 0.4926, "num_tokens": 462241020.0, "reward": 0.1875000149011612, "reward_std": 0.15134096145629883, "rewards/format_reward/mean": 0.6000000238418579, "rewards/format_reward/std": 0.49298879504203796, "rewards/unicoder_reward_fn/mean": 0.1274999976158142, "rewards/unicoder_reward_fn/std": 0.27786481380462646, "step": 2560 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 1490.0, "completions/mean_length": 491.9125061035156, "completions/mean_terminated_length": 446.2911376953125, "completions/min_length": 167.0, "completions/min_terminated_length": 167.0, "epoch": 0.9934057408844065, "grad_norm": 13.16689395904541, "kl": 23.96875, "learning_rate": 5.0057557078761e-07, "loss": 0.5133, "num_tokens": 462326509.0, "reward": 0.1912500113248825, "reward_std": 0.16437070071697235, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.20812605321407318, "step": 2561 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2128.0, "completions/max_terminated_length": 2128.0, "completions/mean_length": 452.7124938964844, "completions/mean_terminated_length": 452.7124938964844, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.9937936384794415, "grad_norm": 20.42721939086914, "kl": 22.40625, "learning_rate": 5.005133972955506e-07, "loss": 0.5624, "num_tokens": 462409078.0, "reward": 0.19750002026557922, "reward_std": 0.164617657661438, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.23117750883102417, "step": 2562 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2784.0, "completions/max_terminated_length": 2784.0, "completions/mean_length": 405.7875061035156, "completions/mean_terminated_length": 405.7875061035156, "completions/min_length": 155.0, "completions/min_terminated_length": 155.0, "epoch": 0.9941815360744763, "grad_norm": 8.62731647491455, "kl": 19.75, "learning_rate": 5.004547760498833e-07, "loss": 0.4138, "num_tokens": 462491861.0, "reward": 0.20125000178813934, "reward_std": 0.15225912630558014, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.11749999225139618, "rewards/unicoder_reward_fn/std": 0.23318108916282654, "step": 2563 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2082.0, "completions/max_terminated_length": 2082.0, "completions/mean_length": 523.8624877929688, "completions/mean_terminated_length": 523.8624877929688, "completions/min_length": 154.0, "completions/min_terminated_length": 154.0, "epoch": 0.9945694336695112, "grad_norm": 10.151226043701172, "kl": 27.40625, "learning_rate": 5.003997071431799e-07, "loss": 0.5682, "num_tokens": 462574314.0, "reward": 0.24375000596046448, "reward_std": 0.21220770478248596, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360076904297, "rewards/unicoder_reward_fn/mean": 0.16749998927116394, "rewards/unicoder_reward_fn/std": 0.29927128553390503, "step": 2564 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1808.0, "completions/max_terminated_length": 1808.0, "completions/mean_length": 445.2875061035156, "completions/mean_terminated_length": 445.2875061035156, "completions/min_length": 135.0, "completions/min_terminated_length": 135.0, "epoch": 0.9949573312645461, "grad_norm": 21.570499420166016, "kl": 20.3125, "learning_rate": 5.003481906624008e-07, "loss": 0.4778, "num_tokens": 462658497.0, "reward": 0.4362500309944153, "reward_std": 0.3089696168899536, "rewards/format_reward/mean": 0.8374999761581421, "rewards/format_reward/std": 0.3712363839149475, "rewards/unicoder_reward_fn/mean": 0.3525000214576721, "rewards/unicoder_reward_fn/std": 0.36870455741882324, "step": 2565 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3029.0, "completions/max_terminated_length": 3029.0, "completions/mean_length": 586.1375122070312, "completions/mean_terminated_length": 586.1375122070312, "completions/min_length": 96.0, "completions/min_terminated_length": 96.0, "epoch": 0.9953452288595811, "grad_norm": 6.735319137573242, "kl": 33.9375, "learning_rate": 5.003002266888981e-07, "loss": 0.5102, "num_tokens": 462751356.0, "reward": 0.1575000137090683, "reward_std": 0.19291971623897552, "rewards/format_reward/mean": 0.550000011920929, "rewards/format_reward/std": 0.5006325244903564, "rewards/unicoder_reward_fn/mean": 0.10249999910593033, "rewards/unicoder_reward_fn/std": 0.18824642896652222, "step": 2566 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3247.0, "completions/max_terminated_length": 3247.0, "completions/mean_length": 430.0874938964844, "completions/mean_terminated_length": 430.0874938964844, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.995733126454616, "grad_norm": 21.857494354248047, "kl": 18.625, "learning_rate": 5.002558152984132e-07, "loss": 0.4618, "num_tokens": 462827211.0, "reward": 0.19749999046325684, "reward_std": 0.1396467536687851, "rewards/format_reward/mean": 0.824999988079071, "rewards/format_reward/std": 0.3823643922805786, "rewards/unicoder_reward_fn/mean": 0.11500000953674316, "rewards/unicoder_reward_fn/std": 0.14849455654621124, "step": 2567 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1712.0, "completions/max_terminated_length": 1712.0, "completions/mean_length": 484.2375183105469, "completions/mean_terminated_length": 484.2375183105469, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.9961210240496509, "grad_norm": 12.068586349487305, "kl": 24.34375, "learning_rate": 5.00214956561078e-07, "loss": 0.5375, "num_tokens": 462915422.0, "reward": 0.20375001430511475, "reward_std": 0.16991297900676727, "rewards/format_reward/mean": 0.762499988079071, "rewards/format_reward/std": 0.4282360374927521, "rewards/unicoder_reward_fn/mean": 0.1275000125169754, "rewards/unicoder_reward_fn/std": 0.23220191895961761, "step": 2568 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3191.0, "completions/mean_length": 699.25, "completions/mean_terminated_length": 656.253173828125, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.9965089216446859, "grad_norm": 4.572378158569336, "kl": 27.46875, "learning_rate": 5.001776505414138e-07, "loss": 0.4167, "num_tokens": 463016306.0, "reward": 0.18500001728534698, "reward_std": 0.18362991511821747, "rewards/format_reward/mean": 0.7250000238418579, "rewards/format_reward/std": 0.4493314325809479, "rewards/unicoder_reward_fn/mean": 0.11249999701976776, "rewards/unicoder_reward_fn/std": 0.20337344706058502, "step": 2569 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2016.0, "completions/max_terminated_length": 2016.0, "completions/mean_length": 413.7875061035156, "completions/mean_terminated_length": 413.7875061035156, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.9968968192397207, "grad_norm": 19.432527542114258, "kl": 19.515625, "learning_rate": 5.001438972983321e-07, "loss": 0.4128, "num_tokens": 463094649.0, "reward": 0.45375004410743713, "reward_std": 0.33608385920524597, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.39277493953704834, "rewards/unicoder_reward_fn/mean": 0.3725000023841858, "rewards/unicoder_reward_fn/std": 0.3610725700855255, "step": 2570 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 3738.0, "completions/mean_length": 585.2750244140625, "completions/mean_terminated_length": 540.83544921875, "completions/min_length": 25.0, "completions/min_terminated_length": 25.0, "epoch": 0.9972847168347556, "grad_norm": 8.126594543457031, "kl": 30.5, "learning_rate": 5.001136968851336e-07, "loss": 0.5549, "num_tokens": 463192575.0, "reward": 0.3212500214576721, "reward_std": 0.28005924820899963, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.2500000298023224, "rewards/unicoder_reward_fn/std": 0.3419971168041229, "step": 2571 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 2780.0, "completions/mean_length": 556.5875244140625, "completions/mean_terminated_length": 511.7848205566406, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.9976726144297905, "grad_norm": 4.399582862854004, "kl": 31.5625, "learning_rate": 5.00087049349509e-07, "loss": 0.5688, "num_tokens": 463281046.0, "reward": 0.2512499988079071, "reward_std": 0.1985575258731842, "rewards/format_reward/mean": 0.7124999761581421, "rewards/format_reward/std": 0.45545223355293274, "rewards/unicoder_reward_fn/mean": 0.17999999225139618, "rewards/unicoder_reward_fn/std": 0.2730048596858978, "step": 2572 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3227.0, "completions/max_terminated_length": 3227.0, "completions/mean_length": 516.1000366210938, "completions/mean_terminated_length": 516.1000366210938, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "epoch": 0.9980605120248255, "grad_norm": 4.99653434753418, "kl": 25.78125, "learning_rate": 5.000639547335384e-07, "loss": 0.476, "num_tokens": 463369606.0, "reward": 0.14750000834465027, "reward_std": 0.11983654648065567, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.06750000268220901, "rewards/unicoder_reward_fn/std": 0.14563262462615967, "step": 2573 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1897.0, "completions/max_terminated_length": 1897.0, "completions/mean_length": 427.3125, "completions/mean_terminated_length": 427.3125, "completions/min_length": 134.0, "completions/min_terminated_length": 134.0, "epoch": 0.9984484096198604, "grad_norm": 3.197005033493042, "kl": 22.53125, "learning_rate": 5.000444130736916e-07, "loss": 0.3952, "num_tokens": 463447471.0, "reward": 0.20500002801418304, "reward_std": 0.11563372611999512, "rewards/format_reward/mean": 0.800000011920929, "rewards/format_reward/std": 0.4025236964225769, "rewards/unicoder_reward_fn/mean": 0.125, "rewards/unicoder_reward_fn/std": 0.21435804665088654, "step": 2574 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2040.0, "completions/max_terminated_length": 2040.0, "completions/mean_length": 548.9125366210938, "completions/mean_terminated_length": 548.9125366210938, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.9988363072148952, "grad_norm": 8.313701629638672, "kl": 37.09375, "learning_rate": 5.000284244008272e-07, "loss": 0.6068, "num_tokens": 463539976.0, "reward": 0.16500000655651093, "reward_std": 0.14277949929237366, "rewards/format_reward/mean": 0.6499999761581421, "rewards/format_reward/std": 0.47997888922691345, "rewards/unicoder_reward_fn/mean": 0.10000000149011612, "rewards/unicoder_reward_fn/std": 0.19093374907970428, "step": 2575 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.012499999999999956, "completions/max_length": 4096.0, "completions/max_terminated_length": 4089.0, "completions/mean_length": 534.2875366210938, "completions/mean_terminated_length": 489.2025451660156, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "epoch": 0.9992242048099301, "grad_norm": 7.911056041717529, "kl": 31.703125, "learning_rate": 5.000159887401937e-07, "loss": 0.6086, "num_tokens": 463623287.0, "reward": 0.2550000250339508, "reward_std": 0.1937754899263382, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.43574467301368713, "rewards/unicoder_reward_fn/mean": 0.18000000715255737, "rewards/unicoder_reward_fn/std": 0.21955075860023499, "step": 2576 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1649.0, "completions/max_terminated_length": 1649.0, "completions/mean_length": 529.4210815429688, "completions/mean_terminated_length": 529.4210815429688, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "epoch": 0.9996121024049651, "grad_norm": 9.203536033630371, "kl": 30.78125, "learning_rate": 5.000071061114284e-07, "loss": 0.4906, "num_tokens": 463713115.0, "reward": 0.19375000894069672, "reward_std": 0.14919963479042053, "rewards/format_reward/mean": 0.7875000238418579, "rewards/format_reward/std": 0.4116576611995697, "rewards/unicoder_reward_fn/mean": 0.11499999463558197, "rewards/unicoder_reward_fn/std": 0.2006950080394745, "step": 2577 }, { "epoch": 0.9996121024049651, "step": 2577, "total_flos": 0.0, "train_loss": 0.351450573751997, "train_runtime": 154865.0862, "train_samples_per_second": 0.166, "train_steps_per_second": 0.017 } ], "logging_steps": 1, "max_steps": 2578, "num_input_tokens_seen": 463713115, "num_train_epochs": 1, "save_steps": 10, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }