{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 882, "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": 1433.0, "completions/max_terminated_length": 1433.0, "completions/mean_length": 787.265625, "completions/mean_terminated_length": 787.265625, "completions/min_length": 415.0, "completions/min_terminated_length": 415.0, "entropy": 0.7912661358714104, "epoch": 0.0005668934240362812, "frac_reward_zero_std": 0.75, "grad_norm": 0.3359375, "learning_rate": 1e-06, "loss": -0.0145, "num_tokens": 58849.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 1, "step_time": 53.208744254894555 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1396.0, "completions/max_terminated_length": 1396.0, "completions/mean_length": 720.953125, "completions/mean_terminated_length": 720.953125, "completions/min_length": 275.0, "completions/min_terminated_length": 275.0, "entropy": 0.8267659060657024, "epoch": 0.0011337868480725624, "frac_reward_zero_std": 0.75, "grad_norm": 0.314453125, "learning_rate": 9.994331065759637e-07, "loss": 0.0152, "num_tokens": 116702.0, "reward": 0.0703125, "reward_std": 0.35486623644828796, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 2, "step_time": 52.46114222612232 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 722.0625, "completions/mean_terminated_length": 722.0625, "completions/min_length": 338.0, "completions/min_terminated_length": 338.0, "entropy": 0.9556892551481724, "epoch": 0.0017006802721088435, "frac_reward_zero_std": 0.625, "grad_norm": 0.466796875, "learning_rate": 9.988662131519275e-07, "loss": 0.0057, "num_tokens": 172418.0, "reward": 0.0703125, "reward_std": 0.35486623644828796, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 3, "step_time": 51.01601474173367 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1369.0, "completions/max_terminated_length": 1369.0, "completions/mean_length": 726.9375, "completions/mean_terminated_length": 726.9375, "completions/min_length": 33.0, "completions/min_terminated_length": 33.0, "entropy": 0.8995401579886675, "epoch": 0.0022675736961451248, "frac_reward_zero_std": 0.875, "grad_norm": 0.232421875, "learning_rate": 9.982993197278912e-07, "loss": 0.0002, "num_tokens": 229150.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 4, "step_time": 50.94503266736865 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2224.0, "completions/max_terminated_length": 2224.0, "completions/mean_length": 798.359375, "completions/mean_terminated_length": 798.359375, "completions/min_length": 384.0, "completions/min_terminated_length": 384.0, "entropy": 0.870071342214942, "epoch": 0.002834467120181406, "frac_reward_zero_std": 0.875, "grad_norm": 0.24609375, "learning_rate": 9.977324263038547e-07, "loss": 0.0055, "num_tokens": 289981.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 5, "step_time": 84.87245248258114 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 777.734375, "completions/mean_terminated_length": 777.734375, "completions/min_length": 328.0, "completions/min_terminated_length": 328.0, "entropy": 0.5162014495581388, "epoch": 0.003401360544217687, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 9.971655328798185e-07, "loss": 0.0, "num_tokens": 350580.0, "reward": 0.0, "reward_std": 0.0, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 6, "step_time": 57.13563923817128 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 745.953125, "completions/mean_terminated_length": 745.953125, "completions/min_length": 350.0, "completions/min_terminated_length": 350.0, "entropy": 0.9106043875217438, "epoch": 0.003968253968253968, "frac_reward_zero_std": 0.625, "grad_norm": 0.44140625, "learning_rate": 9.965986394557822e-07, "loss": 0.027, "num_tokens": 411089.0, "reward": 0.1015625, "reward_std": 0.42891231179237366, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 7, "step_time": 56.46439360547811 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 663.9375, "completions/mean_terminated_length": 663.9375, "completions/min_length": 210.0, "completions/min_terminated_length": 210.0, "entropy": 0.7852413542568684, "epoch": 0.0045351473922902496, "frac_reward_zero_std": 0.875, "grad_norm": 0.2275390625, "learning_rate": 9.96031746031746e-07, "loss": 0.0029, "num_tokens": 465445.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 8, "step_time": 44.557856001891196 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1411.0, "completions/max_terminated_length": 1411.0, "completions/mean_length": 754.640625, "completions/mean_terminated_length": 754.640625, "completions/min_length": 415.0, "completions/min_terminated_length": 415.0, "entropy": 0.6974279470741749, "epoch": 0.00510204081632653, "frac_reward_zero_std": 0.875, "grad_norm": 0.2021484375, "learning_rate": 9.954648526077097e-07, "loss": -0.0029, "num_tokens": 525142.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 9, "step_time": 53.639586839824915 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1870.0, "completions/max_terminated_length": 1870.0, "completions/mean_length": 809.296875, "completions/mean_terminated_length": 809.296875, "completions/min_length": 454.0, "completions/min_terminated_length": 454.0, "entropy": 0.8046941738575697, "epoch": 0.005668934240362812, "frac_reward_zero_std": 0.875, "grad_norm": 0.271484375, "learning_rate": 9.948979591836735e-07, "loss": 0.0001, "num_tokens": 586313.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 10, "step_time": 70.25523345544934 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1355.0, "completions/max_terminated_length": 1355.0, "completions/mean_length": 843.921875, "completions/mean_terminated_length": 843.921875, "completions/min_length": 379.0, "completions/min_terminated_length": 379.0, "entropy": 0.7391964737325907, "epoch": 0.006235827664399093, "frac_reward_zero_std": 0.75, "grad_norm": 0.337890625, "learning_rate": 9.94331065759637e-07, "loss": 0.0022, "num_tokens": 650068.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 11, "step_time": 50.96021834015846 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1932.0, "completions/mean_length": 940.75, "completions/mean_terminated_length": 890.666748046875, "completions/min_length": 354.0, "completions/min_terminated_length": 354.0, "entropy": 0.6950870044529438, "epoch": 0.006802721088435374, "frac_reward_zero_std": 0.875, "grad_norm": 0.2255859375, "learning_rate": 9.937641723356008e-07, "loss": 0.0054, "num_tokens": 720268.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 12, "step_time": 175.690794384107 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1704.0, "completions/max_terminated_length": 1704.0, "completions/mean_length": 896.03125, "completions/mean_terminated_length": 896.03125, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "entropy": 0.7128410097211599, "epoch": 0.007369614512471655, "frac_reward_zero_std": 0.75, "grad_norm": 0.314453125, "learning_rate": 9.931972789115645e-07, "loss": 0.0013, "num_tokens": 788798.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 13, "step_time": 65.53362340852618 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1462.0, "completions/max_terminated_length": 1462.0, "completions/mean_length": 774.09375, "completions/mean_terminated_length": 774.09375, "completions/min_length": 406.0, "completions/min_terminated_length": 406.0, "entropy": 0.8867389839142561, "epoch": 0.007936507936507936, "frac_reward_zero_std": 0.75, "grad_norm": 0.37890625, "learning_rate": 9.926303854875283e-07, "loss": -0.0287, "num_tokens": 849668.0, "reward": 0.046875, "reward_std": 0.2630521357059479, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 14, "step_time": 56.14234718307853 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1392.0, "completions/max_terminated_length": 1392.0, "completions/mean_length": 806.171875, "completions/mean_terminated_length": 806.171875, "completions/min_length": 346.0, "completions/min_terminated_length": 346.0, "entropy": 0.680386021733284, "epoch": 0.008503401360544218, "frac_reward_zero_std": 0.625, "grad_norm": 0.373046875, "learning_rate": 9.92063492063492e-07, "loss": -0.007, "num_tokens": 912015.0, "reward": 0.15625, "reward_std": 0.5409793853759766, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 15, "step_time": 52.486787100322545 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1232.0, "completions/max_terminated_length": 1232.0, "completions/mean_length": 737.03125, "completions/mean_terminated_length": 737.03125, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "entropy": 0.897417239844799, "epoch": 0.009070294784580499, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 9.914965986394558e-07, "loss": 0.0, "num_tokens": 969129.0, "reward": 0.0, "reward_std": 0.0, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 16, "step_time": 46.14659155998379 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1403.0, "completions/max_terminated_length": 1403.0, "completions/mean_length": 765.453125, "completions/mean_terminated_length": 765.453125, "completions/min_length": 403.0, "completions/min_terminated_length": 403.0, "entropy": 0.9181688856333494, "epoch": 0.00963718820861678, "frac_reward_zero_std": 0.625, "grad_norm": 0.451171875, "learning_rate": 9.909297052154193e-07, "loss": -0.0073, "num_tokens": 1029094.0, "reward": 0.1328125, "reward_std": 0.489915668964386, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 17, "step_time": 52.30169053096324 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1291.0, "completions/max_terminated_length": 1291.0, "completions/mean_length": 795.53125, "completions/mean_terminated_length": 795.53125, "completions/min_length": 375.0, "completions/min_terminated_length": 375.0, "entropy": 0.7358187921345234, "epoch": 0.01020408163265306, "frac_reward_zero_std": 0.875, "grad_norm": 0.234375, "learning_rate": 9.90362811791383e-07, "loss": -0.0018, "num_tokens": 1089720.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 18, "step_time": 48.606017048470676 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1462.0, "completions/max_terminated_length": 1462.0, "completions/mean_length": 711.921875, "completions/mean_terminated_length": 711.921875, "completions/min_length": 405.0, "completions/min_terminated_length": 405.0, "entropy": 0.9461786858737469, "epoch": 0.010770975056689343, "frac_reward_zero_std": 0.875, "grad_norm": 0.28515625, "learning_rate": 9.897959183673468e-07, "loss": -0.0098, "num_tokens": 1144827.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 19, "step_time": 54.73584616929293 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 717.609375, "completions/mean_terminated_length": 717.609375, "completions/min_length": 362.0, "completions/min_terminated_length": 362.0, "entropy": 0.7700510062277317, "epoch": 0.011337868480725623, "frac_reward_zero_std": 0.625, "grad_norm": 0.44921875, "learning_rate": 9.892290249433106e-07, "loss": -0.0043, "num_tokens": 1200890.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 20, "step_time": 74.84192624688148 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1629.0, "completions/max_terminated_length": 1629.0, "completions/mean_length": 862.453125, "completions/mean_terminated_length": 862.453125, "completions/min_length": 399.0, "completions/min_terminated_length": 399.0, "entropy": 0.7410835660994053, "epoch": 0.011904761904761904, "frac_reward_zero_std": 0.875, "grad_norm": 0.205078125, "learning_rate": 9.886621315192743e-07, "loss": -0.0016, "num_tokens": 1269087.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 21, "step_time": 62.02429956756532 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 655.640625, "completions/mean_terminated_length": 655.640625, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "entropy": 1.062008362263441, "epoch": 0.012471655328798186, "frac_reward_zero_std": 0.875, "grad_norm": 0.330078125, "learning_rate": 9.88095238095238e-07, "loss": 0.0065, "num_tokens": 1329856.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 22, "step_time": 56.07742853090167 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1237.0, "completions/max_terminated_length": 1237.0, "completions/mean_length": 741.921875, "completions/mean_terminated_length": 741.921875, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "entropy": 0.8006270676851273, "epoch": 0.013038548752834467, "frac_reward_zero_std": 0.75, "grad_norm": 0.486328125, "learning_rate": 9.875283446712018e-07, "loss": -0.0079, "num_tokens": 1388403.0, "reward": 0.0625, "reward_std": 0.2745848298072815, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 23, "step_time": 46.51893087476492 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1368.0, "completions/max_terminated_length": 1368.0, "completions/mean_length": 685.40625, "completions/mean_terminated_length": 685.40625, "completions/min_length": 289.0, "completions/min_terminated_length": 289.0, "entropy": 0.9855949841439724, "epoch": 0.013605442176870748, "frac_reward_zero_std": 0.75, "grad_norm": 0.41796875, "learning_rate": 9.869614512471656e-07, "loss": 0.0031, "num_tokens": 1442821.0, "reward": 0.0390625, "reward_std": 0.2567298710346222, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 24, "step_time": 51.394197618588805 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1163.0, "completions/max_terminated_length": 1163.0, "completions/mean_length": 753.390625, "completions/mean_terminated_length": 753.390625, "completions/min_length": 31.0, "completions/min_terminated_length": 31.0, "entropy": 0.8061298858374357, "epoch": 0.01417233560090703, "frac_reward_zero_std": 0.75, "grad_norm": 0.3515625, "learning_rate": 9.86394557823129e-07, "loss": 0.0055, "num_tokens": 1501582.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 25, "step_time": 44.023181112483144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 733.625, "completions/mean_terminated_length": 733.625, "completions/min_length": 20.0, "completions/min_terminated_length": 20.0, "entropy": 0.8445104025304317, "epoch": 0.01473922902494331, "frac_reward_zero_std": 0.75, "grad_norm": 0.365234375, "learning_rate": 9.858276643990928e-07, "loss": -0.0117, "num_tokens": 1557806.0, "reward": 0.0703125, "reward_std": 0.39708447456359863, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 26, "step_time": 53.340097087435424 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1720.0, "completions/max_terminated_length": 1720.0, "completions/mean_length": 724.765625, "completions/mean_terminated_length": 724.765625, "completions/min_length": 353.0, "completions/min_terminated_length": 353.0, "entropy": 1.036143023520708, "epoch": 0.015306122448979591, "frac_reward_zero_std": 0.75, "grad_norm": 0.345703125, "learning_rate": 9.852607709750566e-07, "loss": 0.0017, "num_tokens": 1614767.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 27, "step_time": 64.8979874541983 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1388.0, "completions/max_terminated_length": 1388.0, "completions/mean_length": 730.71875, "completions/mean_terminated_length": 730.71875, "completions/min_length": 177.0, "completions/min_terminated_length": 177.0, "entropy": 0.7145545389503241, "epoch": 0.015873015873015872, "frac_reward_zero_std": 0.625, "grad_norm": 0.392578125, "learning_rate": 9.846938775510203e-07, "loss": 0.0183, "num_tokens": 1671429.0, "reward": 0.1015625, "reward_std": 0.42891231179237366, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 28, "step_time": 52.964277925901115 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 806.96875, "completions/mean_terminated_length": 806.96875, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "entropy": 0.9442988224327564, "epoch": 0.016439909297052153, "frac_reward_zero_std": 0.75, "grad_norm": 0.359375, "learning_rate": 9.84126984126984e-07, "loss": 0.0258, "num_tokens": 1736139.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 29, "step_time": 50.33796348888427 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1497.0, "completions/max_terminated_length": 1497.0, "completions/mean_length": 751.125, "completions/mean_terminated_length": 751.125, "completions/min_length": 304.0, "completions/min_terminated_length": 304.0, "entropy": 0.9300839081406593, "epoch": 0.017006802721088437, "frac_reward_zero_std": 0.5, "grad_norm": 0.5625, "learning_rate": 9.835600907029478e-07, "loss": -0.0215, "num_tokens": 1795699.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 30, "step_time": 58.92226161248982 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1455.0, "completions/max_terminated_length": 1455.0, "completions/mean_length": 748.625, "completions/mean_terminated_length": 748.625, "completions/min_length": 364.0, "completions/min_terminated_length": 364.0, "entropy": 0.9305003769695759, "epoch": 0.017573696145124718, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 9.829931972789116e-07, "loss": 0.0, "num_tokens": 1853307.0, "reward": 0.0, "reward_std": 0.0, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 31, "step_time": 55.984347719699144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1257.0, "completions/max_terminated_length": 1257.0, "completions/mean_length": 756.6875, "completions/mean_terminated_length": 756.6875, "completions/min_length": 330.0, "completions/min_terminated_length": 330.0, "entropy": 1.0353158302605152, "epoch": 0.018140589569160998, "frac_reward_zero_std": 0.75, "grad_norm": 0.39453125, "learning_rate": 9.824263038548753e-07, "loss": -0.028, "num_tokens": 1914359.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 32, "step_time": 48.36243443284184 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 752.609375, "completions/mean_terminated_length": 752.609375, "completions/min_length": 268.0, "completions/min_terminated_length": 268.0, "entropy": 0.7963434420526028, "epoch": 0.01870748299319728, "frac_reward_zero_std": 0.625, "grad_norm": 0.486328125, "learning_rate": 9.818594104308389e-07, "loss": -0.0104, "num_tokens": 1971198.0, "reward": 0.1875, "reward_std": 0.5875696539878845, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 33, "step_time": 47.883975078351796 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2255.0, "completions/max_terminated_length": 2255.0, "completions/mean_length": 762.859375, "completions/mean_terminated_length": 762.859375, "completions/min_length": 358.0, "completions/min_terminated_length": 358.0, "entropy": 0.7279575727880001, "epoch": 0.01927437641723356, "frac_reward_zero_std": 0.875, "grad_norm": 0.2431640625, "learning_rate": 9.812925170068026e-07, "loss": 0.0052, "num_tokens": 2030189.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 34, "step_time": 87.02857953030616 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1432.0, "completions/max_terminated_length": 1432.0, "completions/mean_length": 823.03125, "completions/mean_terminated_length": 823.03125, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "entropy": 0.8386581428349018, "epoch": 0.01984126984126984, "frac_reward_zero_std": 0.5, "grad_norm": 0.486328125, "learning_rate": 9.807256235827664e-07, "loss": 0.002, "num_tokens": 2094439.0, "reward": 0.1953125, "reward_std": 0.5883605480194092, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 35, "step_time": 54.508007938973606 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1454.0, "completions/max_terminated_length": 1454.0, "completions/mean_length": 729.65625, "completions/mean_terminated_length": 729.65625, "completions/min_length": 312.0, "completions/min_terminated_length": 312.0, "entropy": 0.9026074595749378, "epoch": 0.02040816326530612, "frac_reward_zero_std": 0.875, "grad_norm": 0.220703125, "learning_rate": 9.801587301587301e-07, "loss": 0.0048, "num_tokens": 2152001.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 36, "step_time": 55.1510893618688 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 730.28125, "completions/mean_terminated_length": 730.28125, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "entropy": 0.7190846279263496, "epoch": 0.020975056689342405, "frac_reward_zero_std": 0.625, "grad_norm": 0.380859375, "learning_rate": 9.795918367346939e-07, "loss": 0.0085, "num_tokens": 2210139.0, "reward": 0.1640625, "reward_std": 0.5422958731651306, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 37, "step_time": 45.36145204398781 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1166.0, "completions/max_terminated_length": 1166.0, "completions/mean_length": 751.625, "completions/mean_terminated_length": 751.625, "completions/min_length": 69.0, "completions/min_terminated_length": 69.0, "entropy": 0.7631574179977179, "epoch": 0.021541950113378686, "frac_reward_zero_std": 0.5, "grad_norm": 0.5078125, "learning_rate": 9.790249433106576e-07, "loss": 0.0038, "num_tokens": 2267779.0, "reward": 0.25, "reward_std": 0.6666666865348816, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 38, "step_time": 43.88726707175374 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1659.0, "completions/max_terminated_length": 1659.0, "completions/mean_length": 777.15625, "completions/mean_terminated_length": 777.15625, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "entropy": 1.039644118398428, "epoch": 0.022108843537414966, "frac_reward_zero_std": 0.875, "grad_norm": 0.25, "learning_rate": 9.784580498866214e-07, "loss": -0.0216, "num_tokens": 2328453.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 39, "step_time": 63.13692790642381 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1534.0, "completions/max_terminated_length": 1534.0, "completions/mean_length": 730.8125, "completions/mean_terminated_length": 730.8125, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "entropy": 1.0110173597931862, "epoch": 0.022675736961451247, "frac_reward_zero_std": 0.625, "grad_norm": 0.45703125, "learning_rate": 9.778911564625851e-07, "loss": 0.0069, "num_tokens": 2389921.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 40, "step_time": 60.27846311684698 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1052.0, "completions/mean_length": 690.328125, "completions/mean_terminated_length": 636.2698974609375, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "entropy": 1.0578433759510517, "epoch": 0.023242630385487528, "frac_reward_zero_std": 0.75, "grad_norm": 0.439453125, "learning_rate": 9.773242630385487e-07, "loss": -0.0046, "num_tokens": 2444566.0, "reward": 0.046875, "reward_std": 0.2630521357059479, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 41, "step_time": 177.39915514178574 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1472.0, "completions/max_terminated_length": 1472.0, "completions/mean_length": 838.0, "completions/mean_terminated_length": 838.0, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "entropy": 0.8202644884586334, "epoch": 0.023809523809523808, "frac_reward_zero_std": 0.75, "grad_norm": 0.375, "learning_rate": 9.767573696145124e-07, "loss": -0.0029, "num_tokens": 2509566.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 42, "step_time": 56.90364085417241 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1452.0, "completions/max_terminated_length": 1452.0, "completions/mean_length": 734.171875, "completions/mean_terminated_length": 734.171875, "completions/min_length": 320.0, "completions/min_terminated_length": 320.0, "entropy": 0.7877535559237003, "epoch": 0.024376417233560092, "frac_reward_zero_std": 0.625, "grad_norm": 0.396484375, "learning_rate": 9.761904761904762e-07, "loss": 0.008, "num_tokens": 2566745.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 43, "step_time": 55.869060398079455 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 814.359375, "completions/mean_terminated_length": 814.359375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "entropy": 0.7830752562731504, "epoch": 0.024943310657596373, "frac_reward_zero_std": 0.75, "grad_norm": 0.33203125, "learning_rate": 9.7562358276644e-07, "loss": -0.0146, "num_tokens": 2628928.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 44, "step_time": 55.32074639480561 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 732.40625, "completions/mean_terminated_length": 732.40625, "completions/min_length": 148.0, "completions/min_terminated_length": 148.0, "entropy": 0.9837802685797215, "epoch": 0.025510204081632654, "frac_reward_zero_std": 0.75, "grad_norm": 0.3984375, "learning_rate": 9.750566893424036e-07, "loss": 0.0002, "num_tokens": 2687210.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 45, "step_time": 48.296295342966914 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1630.0, "completions/max_terminated_length": 1630.0, "completions/mean_length": 826.90625, "completions/mean_terminated_length": 826.90625, "completions/min_length": 285.0, "completions/min_terminated_length": 285.0, "entropy": 0.8597951680421829, "epoch": 0.026077097505668934, "frac_reward_zero_std": 0.75, "grad_norm": 0.5, "learning_rate": 9.744897959183674e-07, "loss": -0.0273, "num_tokens": 2750828.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 46, "step_time": 62.49716377258301 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1509.0, "completions/max_terminated_length": 1509.0, "completions/mean_length": 766.0625, "completions/mean_terminated_length": 766.0625, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "entropy": 0.8254429344087839, "epoch": 0.026643990929705215, "frac_reward_zero_std": 0.625, "grad_norm": 0.4296875, "learning_rate": 9.739229024943311e-07, "loss": 0.0219, "num_tokens": 2813176.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 47, "step_time": 58.94816019106656 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1348.0, "completions/max_terminated_length": 1348.0, "completions/mean_length": 796.171875, "completions/mean_terminated_length": 796.171875, "completions/min_length": 338.0, "completions/min_terminated_length": 338.0, "entropy": 0.70322042144835, "epoch": 0.027210884353741496, "frac_reward_zero_std": 0.625, "grad_norm": 0.447265625, "learning_rate": 9.733560090702947e-07, "loss": -0.0156, "num_tokens": 2873451.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 48, "step_time": 51.40357466414571 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 642.03125, "completions/mean_terminated_length": 642.03125, "completions/min_length": 28.0, "completions/min_terminated_length": 28.0, "entropy": 1.0623097755014896, "epoch": 0.027777777777777776, "frac_reward_zero_std": 0.625, "grad_norm": 0.546875, "learning_rate": 9.727891156462584e-07, "loss": 0.0177, "num_tokens": 2923909.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 49, "step_time": 39.97872541937977 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1270.0, "completions/max_terminated_length": 1270.0, "completions/mean_length": 765.9375, "completions/mean_terminated_length": 765.9375, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "entropy": 0.8148674815893173, "epoch": 0.02834467120181406, "frac_reward_zero_std": 0.75, "grad_norm": 0.38671875, "learning_rate": 9.722222222222222e-07, "loss": -0.0207, "num_tokens": 2983305.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 50, "step_time": 48.24801494460553 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1388.0, "completions/mean_length": 821.859375, "completions/mean_terminated_length": 769.888916015625, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "entropy": 0.7434265501797199, "epoch": 0.02891156462585034, "frac_reward_zero_std": 0.75, "grad_norm": 0.33203125, "learning_rate": 9.71655328798186e-07, "loss": 0.0041, "num_tokens": 3049592.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 51, "step_time": 188.78284354228526 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1639.0, "completions/max_terminated_length": 1639.0, "completions/mean_length": 812.09375, "completions/mean_terminated_length": 812.09375, "completions/min_length": 421.0, "completions/min_terminated_length": 421.0, "entropy": 0.9160140417516232, "epoch": 0.02947845804988662, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 9.710884353741497e-07, "loss": 0.0, "num_tokens": 3110910.0, "reward": 0.0, "reward_std": 0.0, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 52, "step_time": 61.9008500315249 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 694.109375, "completions/mean_terminated_length": 694.109375, "completions/min_length": 349.0, "completions/min_terminated_length": 349.0, "entropy": 1.017433162778616, "epoch": 0.030045351473922902, "frac_reward_zero_std": 0.875, "grad_norm": 0.31640625, "learning_rate": 9.705215419501134e-07, "loss": -0.004, "num_tokens": 3165373.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 53, "step_time": 56.43010967876762 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1607.0, "completions/max_terminated_length": 1607.0, "completions/mean_length": 828.953125, "completions/mean_terminated_length": 828.953125, "completions/min_length": 375.0, "completions/min_terminated_length": 375.0, "entropy": 0.7901779320091009, "epoch": 0.030612244897959183, "frac_reward_zero_std": 0.875, "grad_norm": 0.1689453125, "learning_rate": 9.69954648526077e-07, "loss": 0.0009, "num_tokens": 3229466.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 54, "step_time": 61.095109866932034 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1296.0, "completions/max_terminated_length": 1296.0, "completions/mean_length": 707.84375, "completions/mean_terminated_length": 707.84375, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "entropy": 0.9682547487318516, "epoch": 0.031179138321995464, "frac_reward_zero_std": 0.625, "grad_norm": 0.455078125, "learning_rate": 9.693877551020407e-07, "loss": 0.0097, "num_tokens": 3286064.0, "reward": 0.140625, "reward_std": 0.4917473793029785, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 55, "step_time": 49.53000052366406 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1407.0, "completions/max_terminated_length": 1407.0, "completions/mean_length": 747.03125, "completions/mean_terminated_length": 747.03125, "completions/min_length": 376.0, "completions/min_terminated_length": 376.0, "entropy": 0.8757524751126766, "epoch": 0.031746031746031744, "frac_reward_zero_std": 0.875, "grad_norm": 0.294921875, "learning_rate": 9.688208616780045e-07, "loss": 0.0033, "num_tokens": 3345554.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 56, "step_time": 54.49297194555402 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1591.0, "completions/max_terminated_length": 1591.0, "completions/mean_length": 735.1875, "completions/mean_terminated_length": 735.1875, "completions/min_length": 327.0, "completions/min_terminated_length": 327.0, "entropy": 0.7640372887253761, "epoch": 0.03231292517006803, "frac_reward_zero_std": 0.625, "grad_norm": 0.63671875, "learning_rate": 9.682539682539682e-07, "loss": 0.0149, "num_tokens": 3405638.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 57, "step_time": 63.35214420687407 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 664.578125, "completions/mean_terminated_length": 664.578125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "entropy": 0.7871280889958143, "epoch": 0.032879818594104306, "frac_reward_zero_std": 0.375, "grad_norm": 0.6796875, "learning_rate": 9.67687074829932e-07, "loss": 0.016, "num_tokens": 3457443.0, "reward": 0.171875, "reward_std": 0.5434950590133667, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 58, "step_time": 53.59946721140295 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1491.0, "completions/mean_length": 756.1875, "completions/mean_terminated_length": 703.1746215820312, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "entropy": 0.9718666542321444, "epoch": 0.03344671201814059, "frac_reward_zero_std": 0.75, "grad_norm": 0.412109375, "learning_rate": 9.671201814058957e-07, "loss": -0.0036, "num_tokens": 3516543.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 59, "step_time": 176.96932354010642 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1990.0, "completions/max_terminated_length": 1990.0, "completions/mean_length": 785.265625, "completions/mean_terminated_length": 785.265625, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "entropy": 0.9386273510754108, "epoch": 0.034013605442176874, "frac_reward_zero_std": 0.875, "grad_norm": 0.224609375, "learning_rate": 9.665532879818595e-07, "loss": 0.0006, "num_tokens": 3577824.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 60, "step_time": 76.28884133696556 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2057.0, "completions/max_terminated_length": 2057.0, "completions/mean_length": 800.65625, "completions/mean_terminated_length": 800.65625, "completions/min_length": 371.0, "completions/min_terminated_length": 371.0, "entropy": 0.9068919755518436, "epoch": 0.03458049886621315, "frac_reward_zero_std": 0.5, "grad_norm": 0.498046875, "learning_rate": 9.65986394557823e-07, "loss": 0.0151, "num_tokens": 3640170.0, "reward": 0.15625, "reward_std": 0.5409793853759766, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 61, "step_time": 79.0729680787772 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1312.0, "completions/max_terminated_length": 1312.0, "completions/mean_length": 678.296875, "completions/mean_terminated_length": 678.296875, "completions/min_length": 339.0, "completions/min_terminated_length": 339.0, "entropy": 0.9080477468669415, "epoch": 0.035147392290249435, "frac_reward_zero_std": 0.75, "grad_norm": 0.4375, "learning_rate": 9.654195011337868e-07, "loss": 0.0082, "num_tokens": 3692229.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 62, "step_time": 49.03559826966375 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1295.0, "completions/max_terminated_length": 1295.0, "completions/mean_length": 733.890625, "completions/mean_terminated_length": 733.890625, "completions/min_length": 276.0, "completions/min_terminated_length": 276.0, "entropy": 1.0118759498000145, "epoch": 0.03571428571428571, "frac_reward_zero_std": 0.875, "grad_norm": 0.25, "learning_rate": 9.648526077097505e-07, "loss": 0.0029, "num_tokens": 3749190.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 63, "step_time": 48.98392139002681 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1745.0, "completions/max_terminated_length": 1745.0, "completions/mean_length": 793.625, "completions/mean_terminated_length": 793.625, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "entropy": 0.8415250703692436, "epoch": 0.036281179138321996, "frac_reward_zero_std": 0.875, "grad_norm": 0.1875, "learning_rate": 9.642857142857142e-07, "loss": -0.0139, "num_tokens": 3812422.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 64, "step_time": 66.7228564498946 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1802.0, "completions/mean_length": 813.0, "completions/mean_terminated_length": 760.888916015625, "completions/min_length": 253.0, "completions/min_terminated_length": 253.0, "entropy": 0.9139939080923796, "epoch": 0.036848072562358274, "frac_reward_zero_std": 0.625, "grad_norm": 0.41015625, "learning_rate": 9.63718820861678e-07, "loss": -0.0032, "num_tokens": 3876014.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 65, "step_time": 178.41644801665097 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1971.0, "completions/max_terminated_length": 1971.0, "completions/mean_length": 887.59375, "completions/mean_terminated_length": 887.59375, "completions/min_length": 386.0, "completions/min_terminated_length": 386.0, "entropy": 0.7073164060711861, "epoch": 0.03741496598639456, "frac_reward_zero_std": 0.75, "grad_norm": 0.3046875, "learning_rate": 9.631519274376417e-07, "loss": -0.0054, "num_tokens": 3945324.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 66, "step_time": 75.84505661018193 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1218.0, "completions/max_terminated_length": 1218.0, "completions/mean_length": 760.796875, "completions/mean_terminated_length": 760.796875, "completions/min_length": 327.0, "completions/min_terminated_length": 327.0, "entropy": 0.7621981650590897, "epoch": 0.03798185941043084, "frac_reward_zero_std": 0.625, "grad_norm": 0.52734375, "learning_rate": 9.625850340136053e-07, "loss": -0.0338, "num_tokens": 4004503.0, "reward": 0.1015625, "reward_std": 0.42891231179237366, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 67, "step_time": 46.655542940832675 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1647.0, "completions/max_terminated_length": 1647.0, "completions/mean_length": 870.921875, "completions/mean_terminated_length": 870.921875, "completions/min_length": 335.0, "completions/min_terminated_length": 335.0, "entropy": 0.6609465070068836, "epoch": 0.03854875283446712, "frac_reward_zero_std": 0.625, "grad_norm": 0.359375, "learning_rate": 9.62018140589569e-07, "loss": 0.0241, "num_tokens": 4070570.0, "reward": 0.15625, "reward_std": 0.5409793853759766, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 68, "step_time": 64.57767817284912 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1423.0, "completions/max_terminated_length": 1423.0, "completions/mean_length": 810.359375, "completions/mean_terminated_length": 810.359375, "completions/min_length": 299.0, "completions/min_terminated_length": 299.0, "entropy": 0.7866592034697533, "epoch": 0.0391156462585034, "frac_reward_zero_std": 0.875, "grad_norm": 0.19921875, "learning_rate": 9.614512471655328e-07, "loss": -0.0032, "num_tokens": 4134009.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 69, "step_time": 54.660805217921734 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1522.0, "completions/max_terminated_length": 1522.0, "completions/mean_length": 793.765625, "completions/mean_terminated_length": 793.765625, "completions/min_length": 355.0, "completions/min_terminated_length": 355.0, "entropy": 0.8732692003250122, "epoch": 0.03968253968253968, "frac_reward_zero_std": 0.75, "grad_norm": 0.3359375, "learning_rate": 9.608843537414965e-07, "loss": 0.0036, "num_tokens": 4195066.0, "reward": 0.0703125, "reward_std": 0.35486623644828796, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 70, "step_time": 58.68067215476185 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 666.203125, "completions/mean_terminated_length": 666.203125, "completions/min_length": 273.0, "completions/min_terminated_length": 273.0, "entropy": 0.7925896048545837, "epoch": 0.040249433106575964, "frac_reward_zero_std": 0.875, "grad_norm": 0.30078125, "learning_rate": 9.603174603174603e-07, "loss": 0.0067, "num_tokens": 4247335.0, "reward": 0.015625, "reward_std": 0.08768405020236969, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 71, "step_time": 45.4902665251866 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1203.0, "completions/max_terminated_length": 1203.0, "completions/mean_length": 583.375, "completions/mean_terminated_length": 583.375, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "entropy": 0.8753242213279009, "epoch": 0.04081632653061224, "frac_reward_zero_std": 0.625, "grad_norm": 0.5625, "learning_rate": 9.59750566893424e-07, "loss": 0.0005, "num_tokens": 4292759.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 72, "step_time": 44.927397458814085 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 742.28125, "completions/mean_terminated_length": 742.28125, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "entropy": 0.8473658263683319, "epoch": 0.041383219954648526, "frac_reward_zero_std": 0.875, "grad_norm": 0.2734375, "learning_rate": 9.591836734693876e-07, "loss": 0.0034, "num_tokens": 4358281.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 73, "step_time": 81.2505574543029 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 655.671875, "completions/mean_terminated_length": 655.671875, "completions/min_length": 259.0, "completions/min_terminated_length": 259.0, "entropy": 1.105690035969019, "epoch": 0.04195011337868481, "frac_reward_zero_std": 0.625, "grad_norm": 0.546875, "learning_rate": 9.586167800453513e-07, "loss": -0.0107, "num_tokens": 4409628.0, "reward": 0.1875, "reward_std": 0.5875696539878845, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 74, "step_time": 43.225347391329706 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1425.0, "completions/max_terminated_length": 1425.0, "completions/mean_length": 799.859375, "completions/mean_terminated_length": 799.859375, "completions/min_length": 254.0, "completions/min_terminated_length": 254.0, "entropy": 0.7495364528149366, "epoch": 0.04251700680272109, "frac_reward_zero_std": 0.75, "grad_norm": 0.33203125, "learning_rate": 9.58049886621315e-07, "loss": -0.0215, "num_tokens": 4471059.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 75, "step_time": 54.34133177995682 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 701.703125, "completions/mean_terminated_length": 701.703125, "completions/min_length": 315.0, "completions/min_terminated_length": 315.0, "entropy": 0.7809962090104818, "epoch": 0.04308390022675737, "frac_reward_zero_std": 0.75, "grad_norm": 0.349609375, "learning_rate": 9.574829931972788e-07, "loss": 0.0035, "num_tokens": 4524760.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 76, "step_time": 51.31494827289134 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1443.0, "completions/max_terminated_length": 1443.0, "completions/mean_length": 782.890625, "completions/mean_terminated_length": 782.890625, "completions/min_length": 416.0, "completions/min_terminated_length": 416.0, "entropy": 0.9791093207895756, "epoch": 0.04365079365079365, "frac_reward_zero_std": 0.625, "grad_norm": 0.52734375, "learning_rate": 9.569160997732426e-07, "loss": -0.0071, "num_tokens": 4584777.0, "reward": 0.09375, "reward_std": 0.36596253514289856, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 77, "step_time": 54.10253896191716 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1675.0, "completions/max_terminated_length": 1675.0, "completions/mean_length": 823.84375, "completions/mean_terminated_length": 823.84375, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "entropy": 0.8013289663940668, "epoch": 0.04421768707482993, "frac_reward_zero_std": 0.75, "grad_norm": 0.34765625, "learning_rate": 9.563492063492063e-07, "loss": 0.0228, "num_tokens": 4649015.0, "reward": 0.1015625, "reward_std": 0.42891231179237366, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 78, "step_time": 63.31329367030412 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 789.515625, "completions/mean_terminated_length": 789.515625, "completions/min_length": 349.0, "completions/min_terminated_length": 349.0, "entropy": 0.7154162228107452, "epoch": 0.04478458049886622, "frac_reward_zero_std": 0.5, "grad_norm": 0.48046875, "learning_rate": 9.5578231292517e-07, "loss": 0.0182, "num_tokens": 4710208.0, "reward": 0.1484375, "reward_std": 0.49344661831855774, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 79, "step_time": 54.76761769596487 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1659.0, "completions/max_terminated_length": 1659.0, "completions/mean_length": 806.765625, "completions/mean_terminated_length": 806.765625, "completions/min_length": 8.0, "completions/min_terminated_length": 8.0, "entropy": 0.7990427520126104, "epoch": 0.045351473922902494, "frac_reward_zero_std": 0.75, "grad_norm": 0.34765625, "learning_rate": 9.552154195011338e-07, "loss": -0.0069, "num_tokens": 4772081.0, "reward": 0.109375, "reward_std": 0.4315783977508545, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 80, "step_time": 62.50960550829768 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1240.0, "completions/max_terminated_length": 1240.0, "completions/mean_length": 713.703125, "completions/mean_terminated_length": 713.703125, "completions/min_length": 337.0, "completions/min_terminated_length": 337.0, "entropy": 0.8844226524233818, "epoch": 0.04591836734693878, "frac_reward_zero_std": 0.875, "grad_norm": 0.291015625, "learning_rate": 9.546485260770974e-07, "loss": 0.0076, "num_tokens": 4827382.0, "reward": 0.0078125, "reward_std": 0.0625, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 81, "step_time": 46.83812944497913 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1190.0, "completions/max_terminated_length": 1190.0, "completions/mean_length": 640.75, "completions/mean_terminated_length": 640.75, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "entropy": 0.9863478187471628, "epoch": 0.046485260770975055, "frac_reward_zero_std": 0.625, "grad_norm": 0.6875, "learning_rate": 9.54081632653061e-07, "loss": 0.0056, "num_tokens": 4882950.0, "reward": 0.109375, "reward_std": 0.4315783977508545, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 82, "step_time": 47.364373445510864 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1294.0, "completions/mean_length": 836.703125, "completions/mean_terminated_length": 784.9683227539062, "completions/min_length": 384.0, "completions/min_terminated_length": 384.0, "entropy": 0.7088115140795708, "epoch": 0.04705215419501134, "frac_reward_zero_std": 0.625, "grad_norm": 0.337890625, "learning_rate": 9.53514739229025e-07, "loss": 0.0159, "num_tokens": 4945459.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 83, "step_time": 177.49385809525847 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1561.0, "completions/max_terminated_length": 1561.0, "completions/mean_length": 784.484375, "completions/mean_terminated_length": 784.484375, "completions/min_length": 450.0, "completions/min_terminated_length": 450.0, "entropy": 0.9638878479599953, "epoch": 0.047619047619047616, "frac_reward_zero_std": 0.75, "grad_norm": 0.33984375, "learning_rate": 9.529478458049887e-07, "loss": 0.0057, "num_tokens": 5006394.0, "reward": 0.1875, "reward_std": 0.5875696539878845, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 84, "step_time": 58.921785313636065 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 925.96875, "completions/mean_terminated_length": 925.96875, "completions/min_length": 464.0, "completions/min_terminated_length": 464.0, "entropy": 0.7016284838318825, "epoch": 0.0481859410430839, "frac_reward_zero_std": 0.875, "grad_norm": 0.2373046875, "learning_rate": 9.523809523809522e-07, "loss": 0.0028, "num_tokens": 5077376.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 85, "step_time": 50.29745448753238 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 808.703125, "completions/mean_terminated_length": 808.703125, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "entropy": 0.8223045617341995, "epoch": 0.048752834467120185, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 9.51814058956916e-07, "loss": 0.0, "num_tokens": 5140029.0, "reward": 0.0, "reward_std": 0.0, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 86, "step_time": 50.399625142104924 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 835.5625, "completions/mean_terminated_length": 835.5625, "completions/min_length": 342.0, "completions/min_terminated_length": 342.0, "entropy": 0.9934357516467571, "epoch": 0.04931972789115646, "frac_reward_zero_std": 0.75, "grad_norm": 0.330078125, "learning_rate": 9.512471655328797e-07, "loss": 0.0201, "num_tokens": 5206257.0, "reward": 0.0703125, "reward_std": 0.35486623644828796, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 87, "step_time": 58.83743468020111 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 807.65625, "completions/mean_terminated_length": 807.65625, "completions/min_length": 425.0, "completions/min_terminated_length": 425.0, "entropy": 0.9231362417340279, "epoch": 0.049886621315192746, "frac_reward_zero_std": 0.875, "grad_norm": 0.306640625, "learning_rate": 9.506802721088435e-07, "loss": 0.0085, "num_tokens": 5277547.0, "reward": 0.015625, "reward_std": 0.08768405020236969, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 88, "step_time": 64.1513889292255 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1456.0, "completions/max_terminated_length": 1456.0, "completions/mean_length": 725.40625, "completions/mean_terminated_length": 725.40625, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "entropy": 0.8447514325380325, "epoch": 0.05045351473922902, "frac_reward_zero_std": 0.75, "grad_norm": 0.349609375, "learning_rate": 9.501133786848072e-07, "loss": 0.0131, "num_tokens": 5334709.0, "reward": 0.0546875, "reward_std": 0.26899558305740356, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 89, "step_time": 55.60451160091907 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1420.0, "completions/max_terminated_length": 1420.0, "completions/mean_length": 801.5625, "completions/mean_terminated_length": 801.5625, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "entropy": 0.9044834561645985, "epoch": 0.05102040816326531, "frac_reward_zero_std": 0.875, "grad_norm": 0.2216796875, "learning_rate": 9.49546485260771e-07, "loss": 0.0025, "num_tokens": 5397441.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 90, "step_time": 53.2394202882424 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 795.109375, "completions/mean_terminated_length": 795.109375, "completions/min_length": 200.0, "completions/min_terminated_length": 200.0, "entropy": 0.9016128778457642, "epoch": 0.051587301587301584, "frac_reward_zero_std": 0.75, "grad_norm": 0.36328125, "learning_rate": 9.489795918367347e-07, "loss": -0.0106, "num_tokens": 5460848.0, "reward": 0.15625, "reward_std": 0.5409793853759766, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 91, "step_time": 57.886049394495785 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1684.0, "completions/max_terminated_length": 1684.0, "completions/mean_length": 761.984375, "completions/mean_terminated_length": 761.984375, "completions/min_length": 339.0, "completions/min_terminated_length": 339.0, "entropy": 0.7569958660751581, "epoch": 0.05215419501133787, "frac_reward_zero_std": 0.75, "grad_norm": 0.35546875, "learning_rate": 9.484126984126984e-07, "loss": 0.0051, "num_tokens": 5519727.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 92, "step_time": 63.76254240516573 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2312.0, "completions/max_terminated_length": 2312.0, "completions/mean_length": 824.375, "completions/mean_terminated_length": 824.375, "completions/min_length": 262.0, "completions/min_terminated_length": 262.0, "entropy": 0.8328082952648401, "epoch": 0.05272108843537415, "frac_reward_zero_std": 0.875, "grad_norm": 0.30078125, "learning_rate": 9.47845804988662e-07, "loss": -0.0043, "num_tokens": 5584447.0, "reward": 0.0078125, "reward_std": 0.0625, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 93, "step_time": 89.38122331723571 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 682.65625, "completions/mean_terminated_length": 682.65625, "completions/min_length": 368.0, "completions/min_terminated_length": 368.0, "entropy": 1.166295189410448, "epoch": 0.05328798185941043, "frac_reward_zero_std": 0.75, "grad_norm": 0.404296875, "learning_rate": 9.472789115646258e-07, "loss": 0.0014, "num_tokens": 5639625.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 94, "step_time": 40.6492495816201 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 710.375, "completions/mean_terminated_length": 710.375, "completions/min_length": 313.0, "completions/min_terminated_length": 313.0, "entropy": 0.9604226835072041, "epoch": 0.053854875283446714, "frac_reward_zero_std": 0.875, "grad_norm": 0.279296875, "learning_rate": 9.467120181405895e-07, "loss": 0.0113, "num_tokens": 5695033.0, "reward": 0.0078125, "reward_std": 0.0625, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 95, "step_time": 53.77150951512158 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1531.0, "completions/max_terminated_length": 1531.0, "completions/mean_length": 730.59375, "completions/mean_terminated_length": 730.59375, "completions/min_length": 320.0, "completions/min_terminated_length": 320.0, "entropy": 0.7991134524345398, "epoch": 0.05442176870748299, "frac_reward_zero_std": 0.625, "grad_norm": 0.4140625, "learning_rate": 9.461451247165533e-07, "loss": 0.0271, "num_tokens": 5752639.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 96, "step_time": 59.56408981606364 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 776.53125, "completions/mean_terminated_length": 776.53125, "completions/min_length": 402.0, "completions/min_terminated_length": 402.0, "entropy": 0.9329401925206184, "epoch": 0.054988662131519275, "frac_reward_zero_std": 0.75, "grad_norm": 0.40625, "learning_rate": 9.45578231292517e-07, "loss": -0.0129, "num_tokens": 5813473.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 97, "step_time": 47.72492705844343 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1445.0, "completions/max_terminated_length": 1445.0, "completions/mean_length": 850.984375, "completions/mean_terminated_length": 850.984375, "completions/min_length": 438.0, "completions/min_terminated_length": 438.0, "entropy": 0.7094423845410347, "epoch": 0.05555555555555555, "frac_reward_zero_std": 0.75, "grad_norm": 0.3046875, "learning_rate": 9.450113378684807e-07, "loss": 0.0181, "num_tokens": 5878904.0, "reward": 0.046875, "reward_std": 0.2630521357059479, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 98, "step_time": 56.325513327494264 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1597.0, "completions/max_terminated_length": 1597.0, "completions/mean_length": 879.5, "completions/mean_terminated_length": 879.5, "completions/min_length": 358.0, "completions/min_terminated_length": 358.0, "entropy": 0.7695587165653706, "epoch": 0.05612244897959184, "frac_reward_zero_std": 0.875, "grad_norm": 0.2265625, "learning_rate": 9.444444444444444e-07, "loss": -0.0, "num_tokens": 5948408.0, "reward": 0.0078125, "reward_std": 0.0625, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 99, "step_time": 63.71368716843426 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1517.0, "completions/max_terminated_length": 1517.0, "completions/mean_length": 794.984375, "completions/mean_terminated_length": 794.984375, "completions/min_length": 344.0, "completions/min_terminated_length": 344.0, "entropy": 0.7346187010407448, "epoch": 0.05668934240362812, "frac_reward_zero_std": 0.5, "grad_norm": 0.53125, "learning_rate": 9.438775510204082e-07, "loss": 0.0258, "num_tokens": 6010095.0, "reward": 0.203125, "reward_std": 0.5890451669692993, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 100, "step_time": 57.80585550609976 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1724.0, "completions/max_terminated_length": 1724.0, "completions/mean_length": 755.90625, "completions/mean_terminated_length": 755.90625, "completions/min_length": 311.0, "completions/min_terminated_length": 311.0, "entropy": 0.7995297256857157, "epoch": 0.0572562358276644, "frac_reward_zero_std": 0.75, "grad_norm": 0.33984375, "learning_rate": 9.433106575963718e-07, "loss": 0.0019, "num_tokens": 6068169.0, "reward": 0.125, "reward_std": 0.48795002698898315, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 101, "step_time": 65.5606642998755 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1146.0, "completions/max_terminated_length": 1146.0, "completions/mean_length": 799.046875, "completions/mean_terminated_length": 799.046875, "completions/min_length": 403.0, "completions/min_terminated_length": 403.0, "entropy": 0.7355476096272469, "epoch": 0.05782312925170068, "frac_reward_zero_std": 0.75, "grad_norm": 0.353515625, "learning_rate": 9.427437641723356e-07, "loss": 0.0018, "num_tokens": 6129364.0, "reward": 0.0703125, "reward_std": 0.35486623644828796, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 102, "step_time": 43.53606962505728 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1734.0, "completions/max_terminated_length": 1734.0, "completions/mean_length": 830.765625, "completions/mean_terminated_length": 830.765625, "completions/min_length": 460.0, "completions/min_terminated_length": 460.0, "entropy": 0.8551389183849096, "epoch": 0.05839002267573696, "frac_reward_zero_std": 0.625, "grad_norm": 0.388671875, "learning_rate": 9.421768707482993e-07, "loss": 0.0269, "num_tokens": 6192565.0, "reward": 0.1953125, "reward_std": 0.5883605480194092, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 103, "step_time": 66.09313382022083 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 813.265625, "completions/mean_terminated_length": 813.265625, "completions/min_length": 11.0, "completions/min_terminated_length": 11.0, "entropy": 0.7450036723166704, "epoch": 0.05895691609977324, "frac_reward_zero_std": 0.625, "grad_norm": 0.353515625, "learning_rate": 9.41609977324263e-07, "loss": 0.0198, "num_tokens": 6257174.0, "reward": 0.0859375, "reward_std": 0.36247265338897705, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 104, "step_time": 57.193663199432194 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1614.0, "completions/max_terminated_length": 1614.0, "completions/mean_length": 689.921875, "completions/mean_terminated_length": 689.921875, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "entropy": 0.7431341577321291, "epoch": 0.05952380952380952, "frac_reward_zero_std": 0.625, "grad_norm": 0.416015625, "learning_rate": 9.410430839002267e-07, "loss": -0.0001, "num_tokens": 6310881.0, "reward": 0.1875, "reward_std": 0.5875696539878845, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 105, "step_time": 61.51257424522191 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2372.0, "completions/max_terminated_length": 2372.0, "completions/mean_length": 755.953125, "completions/mean_terminated_length": 755.953125, "completions/min_length": 282.0, "completions/min_terminated_length": 282.0, "entropy": 0.7027264926582575, "epoch": 0.060090702947845805, "frac_reward_zero_std": 0.75, "grad_norm": 0.34375, "learning_rate": 9.404761904761904e-07, "loss": -0.0087, "num_tokens": 6368918.0, "reward": 0.21875, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 106, "step_time": 92.04900511726737 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 800.71875, "completions/mean_terminated_length": 800.71875, "completions/min_length": 453.0, "completions/min_terminated_length": 453.0, "entropy": 0.7530606593936682, "epoch": 0.06065759637188209, "frac_reward_zero_std": 0.75, "grad_norm": 0.302734375, "learning_rate": 9.399092970521542e-07, "loss": 0.0055, "num_tokens": 6432204.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 107, "step_time": 49.60997568629682 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1581.0, "completions/max_terminated_length": 1581.0, "completions/mean_length": 834.015625, "completions/mean_terminated_length": 834.015625, "completions/min_length": 472.0, "completions/min_terminated_length": 472.0, "entropy": 0.6710873208940029, "epoch": 0.061224489795918366, "frac_reward_zero_std": 0.75, "grad_norm": 0.328125, "learning_rate": 9.393424036281179e-07, "loss": -0.0135, "num_tokens": 6494893.0, "reward": 0.0625, "reward_std": 0.2745848298072815, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 108, "step_time": 60.02176415827125 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 681.40625, "completions/mean_terminated_length": 681.40625, "completions/min_length": 245.0, "completions/min_terminated_length": 245.0, "entropy": 1.0327333752065897, "epoch": 0.06179138321995465, "frac_reward_zero_std": 0.625, "grad_norm": 0.50390625, "learning_rate": 9.387755102040816e-07, "loss": 0.0027, "num_tokens": 6547679.0, "reward": 0.1015625, "reward_std": 0.42891231179237366, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 109, "step_time": 54.22360837832093 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1361.0, "completions/max_terminated_length": 1361.0, "completions/mean_length": 715.171875, "completions/mean_terminated_length": 715.171875, "completions/min_length": 353.0, "completions/min_terminated_length": 353.0, "entropy": 0.7466475442051888, "epoch": 0.06235827664399093, "frac_reward_zero_std": 0.75, "grad_norm": 0.359375, "learning_rate": 9.382086167800452e-07, "loss": -0.0057, "num_tokens": 6603386.0, "reward": 0.25, "reward_std": 0.7237468957901001, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 110, "step_time": 51.49598786607385 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1622.0, "completions/max_terminated_length": 1622.0, "completions/mean_length": 757.90625, "completions/mean_terminated_length": 757.90625, "completions/min_length": 323.0, "completions/min_terminated_length": 323.0, "entropy": 0.8256755955517292, "epoch": 0.06292517006802721, "frac_reward_zero_std": 0.875, "grad_norm": 0.287109375, "learning_rate": 9.37641723356009e-07, "loss": -0.0077, "num_tokens": 6663844.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 111, "step_time": 62.764745325781405 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1264.0, "completions/max_terminated_length": 1264.0, "completions/mean_length": 717.5625, "completions/mean_terminated_length": 717.5625, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "entropy": 0.8293892573565245, "epoch": 0.06349206349206349, "frac_reward_zero_std": 0.75, "grad_norm": 0.421875, "learning_rate": 9.370748299319727e-07, "loss": -0.0144, "num_tokens": 6720208.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 112, "step_time": 48.27981117833406 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1483.0, "completions/max_terminated_length": 1483.0, "completions/mean_length": 743.015625, "completions/mean_terminated_length": 743.015625, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "entropy": 0.8873293474316597, "epoch": 0.06405895691609978, "frac_reward_zero_std": 0.875, "grad_norm": 0.205078125, "learning_rate": 9.365079365079365e-07, "loss": 0.0021, "num_tokens": 6779377.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 113, "step_time": 58.38554639369249 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1565.0, "completions/max_terminated_length": 1565.0, "completions/mean_length": 751.15625, "completions/mean_terminated_length": 751.15625, "completions/min_length": 395.0, "completions/min_terminated_length": 395.0, "entropy": 0.9458110071718693, "epoch": 0.06462585034013606, "frac_reward_zero_std": 0.625, "grad_norm": 0.4453125, "learning_rate": 9.359410430839002e-07, "loss": -0.0463, "num_tokens": 6839931.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 114, "step_time": 61.09925704449415 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 601.734375, "completions/mean_terminated_length": 601.734375, "completions/min_length": 233.0, "completions/min_terminated_length": 233.0, "entropy": 0.9978123791515827, "epoch": 0.06519274376417233, "frac_reward_zero_std": 0.75, "grad_norm": 0.478515625, "learning_rate": 9.35374149659864e-07, "loss": -0.0051, "num_tokens": 6888666.0, "reward": 0.1875, "reward_std": 0.5455447435379028, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 115, "step_time": 49.341413654387 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1531.0, "completions/max_terminated_length": 1531.0, "completions/mean_length": 724.390625, "completions/mean_terminated_length": 724.390625, "completions/min_length": 302.0, "completions/min_terminated_length": 302.0, "entropy": 0.6574038378894329, "epoch": 0.06575963718820861, "frac_reward_zero_std": 0.875, "grad_norm": 0.2412109375, "learning_rate": 9.348072562358276e-07, "loss": -0.0035, "num_tokens": 6951011.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 116, "step_time": 61.11546309199184 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1904.0, "completions/max_terminated_length": 1904.0, "completions/mean_length": 814.578125, "completions/mean_terminated_length": 814.578125, "completions/min_length": 298.0, "completions/min_terminated_length": 298.0, "entropy": 0.8695726897567511, "epoch": 0.0663265306122449, "frac_reward_zero_std": 0.5, "grad_norm": 0.494140625, "learning_rate": 9.342403628117913e-07, "loss": -0.0349, "num_tokens": 7016536.0, "reward": 0.1015625, "reward_std": 0.42891231179237366, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 117, "step_time": 74.26174595579505 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1279.0, "completions/max_terminated_length": 1279.0, "completions/mean_length": 706.703125, "completions/mean_terminated_length": 706.703125, "completions/min_length": 279.0, "completions/min_terminated_length": 279.0, "entropy": 0.785661468282342, "epoch": 0.06689342403628118, "frac_reward_zero_std": 0.625, "grad_norm": 0.443359375, "learning_rate": 9.33673469387755e-07, "loss": -0.0256, "num_tokens": 7073613.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 118, "step_time": 48.82230156194419 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1785.0, "completions/max_terminated_length": 1785.0, "completions/mean_length": 718.65625, "completions/mean_terminated_length": 718.65625, "completions/min_length": 286.0, "completions/min_terminated_length": 286.0, "entropy": 0.8869774080812931, "epoch": 0.06746031746031746, "frac_reward_zero_std": 0.75, "grad_norm": 0.345703125, "learning_rate": 9.331065759637188e-07, "loss": -0.0002, "num_tokens": 7128311.0, "reward": 0.15625, "reward_std": 0.5409793853759766, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 119, "step_time": 66.68152226600796 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 865.1875, "completions/mean_terminated_length": 865.1875, "completions/min_length": 466.0, "completions/min_terminated_length": 466.0, "entropy": 0.7705853059887886, "epoch": 0.06802721088435375, "frac_reward_zero_std": 0.5, "grad_norm": 0.451171875, "learning_rate": 9.325396825396825e-07, "loss": -0.0049, "num_tokens": 7194195.0, "reward": 0.2265625, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 120, "step_time": 62.07781745027751 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1929.0, "completions/max_terminated_length": 1929.0, "completions/mean_length": 821.390625, "completions/mean_terminated_length": 821.390625, "completions/min_length": 356.0, "completions/min_terminated_length": 356.0, "entropy": 0.6111497618257999, "epoch": 0.06859410430839002, "frac_reward_zero_std": 0.75, "grad_norm": 0.310546875, "learning_rate": 9.319727891156463e-07, "loss": -0.0137, "num_tokens": 7258436.0, "reward": 0.046875, "reward_std": 0.2630521357059479, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 121, "step_time": 74.65006504673511 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1487.0, "completions/mean_length": 881.765625, "completions/mean_terminated_length": 830.74609375, "completions/min_length": 424.0, "completions/min_terminated_length": 424.0, "entropy": 0.6640894301235676, "epoch": 0.0691609977324263, "frac_reward_zero_std": 0.625, "grad_norm": 0.3984375, "learning_rate": 9.3140589569161e-07, "loss": 0.0248, "num_tokens": 7326005.0, "reward": 0.15625, "reward_std": 0.5409793853759766, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 122, "step_time": 178.2570448340848 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1314.0, "completions/max_terminated_length": 1314.0, "completions/mean_length": 593.46875, "completions/mean_terminated_length": 593.46875, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "entropy": 0.9700460098683834, "epoch": 0.06972789115646258, "frac_reward_zero_std": 0.75, "grad_norm": 0.400390625, "learning_rate": 9.308390022675737e-07, "loss": 0.0146, "num_tokens": 7377635.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 123, "step_time": 51.13704787194729 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 579.09375, "completions/mean_terminated_length": 579.09375, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "entropy": 1.065275426954031, "epoch": 0.07029478458049887, "frac_reward_zero_std": 0.875, "grad_norm": 0.318359375, "learning_rate": 9.302721088435374e-07, "loss": -0.0053, "num_tokens": 7424409.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 124, "step_time": 37.35036712232977 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 777.375, "completions/mean_terminated_length": 777.375, "completions/min_length": 296.0, "completions/min_terminated_length": 296.0, "entropy": 0.9211017340421677, "epoch": 0.07086167800453515, "frac_reward_zero_std": 0.75, "grad_norm": 0.30078125, "learning_rate": 9.29705215419501e-07, "loss": 0.0199, "num_tokens": 7486521.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 125, "step_time": 57.95168686658144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1432.0, "completions/max_terminated_length": 1432.0, "completions/mean_length": 770.90625, "completions/mean_terminated_length": 770.90625, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "entropy": 0.8487519323825836, "epoch": 0.07142857142857142, "frac_reward_zero_std": 0.875, "grad_norm": 0.28125, "learning_rate": 9.291383219954648e-07, "loss": -0.0032, "num_tokens": 7546539.0, "reward": 0.1015625, "reward_std": 0.40999117493629456, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 126, "step_time": 54.54803218692541 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 780.203125, "completions/mean_terminated_length": 780.203125, "completions/min_length": 9.0, "completions/min_terminated_length": 9.0, "entropy": 0.8813410773873329, "epoch": 0.07199546485260772, "frac_reward_zero_std": 0.75, "grad_norm": 0.380859375, "learning_rate": 9.285714285714285e-07, "loss": -0.0022, "num_tokens": 7607624.0, "reward": 0.046875, "reward_std": 0.2630521357059479, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 127, "step_time": 51.01048209145665 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2171.0, "completions/max_terminated_length": 2171.0, "completions/mean_length": 727.984375, "completions/mean_terminated_length": 727.984375, "completions/min_length": 296.0, "completions/min_terminated_length": 296.0, "entropy": 0.8846711702644825, "epoch": 0.07256235827664399, "frac_reward_zero_std": 0.875, "grad_norm": 0.306640625, "learning_rate": 9.280045351473923e-07, "loss": 0.001, "num_tokens": 7664759.0, "reward": 0.0625, "reward_std": 0.35073620080947876, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 128, "step_time": 82.85143149085343 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 785.359375, "completions/mean_terminated_length": 785.359375, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "entropy": 0.6674748081713915, "epoch": 0.07312925170068027, "frac_reward_zero_std": 0.625, "grad_norm": 0.46875, "learning_rate": 9.274376417233559e-07, "loss": -0.0248, "num_tokens": 7724710.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 129, "step_time": 46.273179126903415 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1516.0, "completions/max_terminated_length": 1516.0, "completions/mean_length": 840.625, "completions/mean_terminated_length": 840.625, "completions/min_length": 305.0, "completions/min_terminated_length": 305.0, "entropy": 0.777132011950016, "epoch": 0.07369614512471655, "frac_reward_zero_std": 0.625, "grad_norm": 0.390625, "learning_rate": 9.268707482993197e-07, "loss": 0.0132, "num_tokens": 7788878.0, "reward": 0.0859375, "reward_std": 0.36247265338897705, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 130, "step_time": 57.40823499020189 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1504.0, "completions/max_terminated_length": 1504.0, "completions/mean_length": 703.9375, "completions/mean_terminated_length": 703.9375, "completions/min_length": 327.0, "completions/min_terminated_length": 327.0, "entropy": 0.9815900027751923, "epoch": 0.07426303854875284, "frac_reward_zero_std": 0.5, "grad_norm": 0.6171875, "learning_rate": 9.263038548752834e-07, "loss": -0.0003, "num_tokens": 7844554.0, "reward": 0.1328125, "reward_std": 0.489915668964386, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 131, "step_time": 57.04882840067148 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1490.0, "completions/max_terminated_length": 1490.0, "completions/mean_length": 837.015625, "completions/mean_terminated_length": 837.015625, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "entropy": 0.7387430816888809, "epoch": 0.07482993197278912, "frac_reward_zero_std": 0.625, "grad_norm": 0.35546875, "learning_rate": 9.257369614512472e-07, "loss": 0.0338, "num_tokens": 7908603.0, "reward": 0.1953125, "reward_std": 0.6147472858428955, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 132, "step_time": 56.945903515443206 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 692.265625, "completions/mean_terminated_length": 692.265625, "completions/min_length": 333.0, "completions/min_terminated_length": 333.0, "entropy": 0.7826277390122414, "epoch": 0.07539682539682539, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 9.251700680272108e-07, "loss": 0.0395, "num_tokens": 7963396.0, "reward": 0.15625, "reward_std": 0.49501484632492065, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 133, "step_time": 48.388916741125286 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1395.0, "completions/max_terminated_length": 1395.0, "completions/mean_length": 789.28125, "completions/mean_terminated_length": 789.28125, "completions/min_length": 326.0, "completions/min_terminated_length": 326.0, "entropy": 0.7802317626774311, "epoch": 0.07596371882086168, "frac_reward_zero_std": 0.625, "grad_norm": 0.40234375, "learning_rate": 9.246031746031746e-07, "loss": -0.0041, "num_tokens": 8025022.0, "reward": 0.1015625, "reward_std": 0.3692515790462494, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 134, "step_time": 53.4332234794274 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1510.0, "completions/max_terminated_length": 1510.0, "completions/mean_length": 618.0, "completions/mean_terminated_length": 618.0, "completions/min_length": 64.0, "completions/min_terminated_length": 64.0, "entropy": 0.8391514271497726, "epoch": 0.07653061224489796, "frac_reward_zero_std": 0.5, "grad_norm": 0.5625, "learning_rate": 9.240362811791382e-07, "loss": 0.0519, "num_tokens": 8073766.0, "reward": 0.328125, "reward_std": 0.7622999548912048, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 135, "step_time": 56.804534021764994 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 829.875, "completions/mean_terminated_length": 829.875, "completions/min_length": 275.0, "completions/min_terminated_length": 275.0, "entropy": 0.7645440567284822, "epoch": 0.07709750566893424, "frac_reward_zero_std": 0.625, "grad_norm": 0.43359375, "learning_rate": 9.23469387755102e-07, "loss": 0.0181, "num_tokens": 8136686.0, "reward": 0.09375, "reward_std": 0.36596253514289856, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 136, "step_time": 75.18006173055619 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1155.0, "completions/max_terminated_length": 1155.0, "completions/mean_length": 691.046875, "completions/mean_terminated_length": 691.046875, "completions/min_length": 335.0, "completions/min_terminated_length": 335.0, "entropy": 1.0325789265334606, "epoch": 0.07766439909297052, "frac_reward_zero_std": 0.75, "grad_norm": 0.37890625, "learning_rate": 9.229024943310657e-07, "loss": 0.021, "num_tokens": 8191625.0, "reward": 0.109375, "reward_std": 0.46691107749938965, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 137, "step_time": 44.112369927577674 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 788.390625, "completions/mean_terminated_length": 788.390625, "completions/min_length": 345.0, "completions/min_terminated_length": 345.0, "entropy": 0.8837102428078651, "epoch": 0.0782312925170068, "frac_reward_zero_std": 0.5, "grad_norm": 0.54296875, "learning_rate": 9.223356009070295e-07, "loss": 0.0164, "num_tokens": 8255322.0, "reward": 0.1875, "reward_std": 0.5455447435379028, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 138, "step_time": 50.744255722500384 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1479.0, "completions/max_terminated_length": 1479.0, "completions/mean_length": 714.609375, "completions/mean_terminated_length": 714.609375, "completions/min_length": 403.0, "completions/min_terminated_length": 403.0, "entropy": 0.803769901394844, "epoch": 0.07879818594104308, "frac_reward_zero_std": 0.625, "grad_norm": 0.462890625, "learning_rate": 9.217687074829932e-07, "loss": 0.0093, "num_tokens": 8312857.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 139, "step_time": 56.574633402749896 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1645.0, "completions/mean_length": 887.4375, "completions/mean_terminated_length": 836.5079956054688, "completions/min_length": 397.0, "completions/min_terminated_length": 397.0, "entropy": 0.6532765198498964, "epoch": 0.07936507936507936, "frac_reward_zero_std": 0.5, "grad_norm": 0.47265625, "learning_rate": 9.21201814058957e-07, "loss": -0.0169, "num_tokens": 8378405.0, "reward": 0.1171875, "reward_std": 0.43408530950546265, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 140, "step_time": 176.8985596196726 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 686.421875, "completions/mean_terminated_length": 686.421875, "completions/min_length": 268.0, "completions/min_terminated_length": 268.0, "entropy": 0.8225036654621363, "epoch": 0.07993197278911565, "frac_reward_zero_std": 0.625, "grad_norm": 0.4765625, "learning_rate": 9.206349206349205e-07, "loss": -0.0061, "num_tokens": 8431888.0, "reward": 0.15625, "reward_std": 0.5409793853759766, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 141, "step_time": 49.527410535141826 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1527.0, "completions/max_terminated_length": 1527.0, "completions/mean_length": 781.078125, "completions/mean_terminated_length": 781.078125, "completions/min_length": 436.0, "completions/min_terminated_length": 436.0, "entropy": 0.8449603244662285, "epoch": 0.08049886621315193, "frac_reward_zero_std": 0.875, "grad_norm": 0.3046875, "learning_rate": 9.200680272108843e-07, "loss": -0.0182, "num_tokens": 8491925.0, "reward": 0.046875, "reward_std": 0.3177144229412079, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 142, "step_time": 58.418819515034556 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 795.625, "completions/mean_terminated_length": 795.625, "completions/min_length": 431.0, "completions/min_terminated_length": 431.0, "entropy": 0.8637427650392056, "epoch": 0.0810657596371882, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 9.19501133786848e-07, "loss": 0.0067, "num_tokens": 8553165.0, "reward": 0.1796875, "reward_std": 0.5445777773857117, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 143, "step_time": 52.9595802705735 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 659.859375, "completions/mean_terminated_length": 659.859375, "completions/min_length": 183.0, "completions/min_terminated_length": 183.0, "entropy": 0.8931311331689358, "epoch": 0.08163265306122448, "frac_reward_zero_std": 0.75, "grad_norm": 0.3828125, "learning_rate": 9.189342403628118e-07, "loss": -0.0258, "num_tokens": 8605972.0, "reward": 0.046875, "reward_std": 0.2630521357059479, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 144, "step_time": 51.218512947671115 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1310.0, "completions/max_terminated_length": 1310.0, "completions/mean_length": 780.859375, "completions/mean_terminated_length": 780.859375, "completions/min_length": 380.0, "completions/min_terminated_length": 380.0, "entropy": 0.8530476838350296, "epoch": 0.08219954648526077, "frac_reward_zero_std": 0.75, "grad_norm": 0.388671875, "learning_rate": 9.183673469387755e-07, "loss": 0.0051, "num_tokens": 8667731.0, "reward": 0.0546875, "reward_std": 0.26899558305740356, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 145, "step_time": 49.99847659096122 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 727.046875, "completions/mean_terminated_length": 727.046875, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "entropy": 0.7862106617540121, "epoch": 0.08276643990929705, "frac_reward_zero_std": 0.375, "grad_norm": 0.56640625, "learning_rate": 9.178004535147393e-07, "loss": 0.02, "num_tokens": 8727598.0, "reward": 0.203125, "reward_std": 0.5890451669692993, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 146, "step_time": 63.81167842168361 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 816.78125, "completions/mean_terminated_length": 816.78125, "completions/min_length": 388.0, "completions/min_terminated_length": 388.0, "entropy": 0.8034506104886532, "epoch": 0.08333333333333333, "frac_reward_zero_std": 0.5, "grad_norm": 0.49609375, "learning_rate": 9.172335600907029e-07, "loss": 0.0094, "num_tokens": 8790232.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 147, "step_time": 52.338583000935614 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 726.625, "completions/mean_terminated_length": 726.625, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "entropy": 0.8533828165382147, "epoch": 0.08390022675736962, "frac_reward_zero_std": 0.75, "grad_norm": 0.3671875, "learning_rate": 9.166666666666665e-07, "loss": 0.0042, "num_tokens": 8846584.0, "reward": 0.109375, "reward_std": 0.4315783977508545, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 148, "step_time": 41.05783475842327 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1764.0, "completions/max_terminated_length": 1764.0, "completions/mean_length": 781.515625, "completions/mean_terminated_length": 781.515625, "completions/min_length": 328.0, "completions/min_terminated_length": 328.0, "entropy": 0.7435462810099125, "epoch": 0.0844671201814059, "frac_reward_zero_std": 0.5, "grad_norm": 0.5078125, "learning_rate": 9.160997732426303e-07, "loss": 0.0042, "num_tokens": 8907089.0, "reward": 0.1875, "reward_std": 0.5455447435379028, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 149, "step_time": 67.09477553144097 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1922.0, "completions/max_terminated_length": 1922.0, "completions/mean_length": 794.375, "completions/mean_terminated_length": 794.375, "completions/min_length": 301.0, "completions/min_terminated_length": 301.0, "entropy": 0.5769841112196445, "epoch": 0.08503401360544217, "frac_reward_zero_std": 0.625, "grad_norm": 0.419921875, "learning_rate": 9.15532879818594e-07, "loss": 0.0269, "num_tokens": 8968057.0, "reward": 0.140625, "reward_std": 0.4917473793029785, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 150, "step_time": 73.29869169741869 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1612.0, "completions/max_terminated_length": 1612.0, "completions/mean_length": 775.984375, "completions/mean_terminated_length": 775.984375, "completions/min_length": 395.0, "completions/min_terminated_length": 395.0, "entropy": 0.9547706730663776, "epoch": 0.08560090702947845, "frac_reward_zero_std": 0.75, "grad_norm": 0.330078125, "learning_rate": 9.149659863945578e-07, "loss": 0.0039, "num_tokens": 9028016.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 151, "step_time": 61.82334432005882 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 736.53125, "completions/mean_terminated_length": 736.53125, "completions/min_length": 286.0, "completions/min_terminated_length": 286.0, "entropy": 0.8605308867990971, "epoch": 0.08616780045351474, "frac_reward_zero_std": 0.5, "grad_norm": 0.65234375, "learning_rate": 9.143990929705215e-07, "loss": -0.0206, "num_tokens": 9086130.0, "reward": 0.1171875, "reward_std": 0.43408530950546265, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 152, "step_time": 52.93216622900218 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1617.0, "completions/max_terminated_length": 1617.0, "completions/mean_length": 731.5, "completions/mean_terminated_length": 731.5, "completions/min_length": 334.0, "completions/min_terminated_length": 334.0, "entropy": 0.8619019202888012, "epoch": 0.08673469387755102, "frac_reward_zero_std": 0.875, "grad_norm": 0.267578125, "learning_rate": 9.138321995464853e-07, "loss": -0.0132, "num_tokens": 9143674.0, "reward": 0.0234375, "reward_std": 0.10652101039886475, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 153, "step_time": 61.16917031444609 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 795.3125, "completions/mean_terminated_length": 795.3125, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 0.8963804133236408, "epoch": 0.0873015873015873, "frac_reward_zero_std": 0.75, "grad_norm": 0.33984375, "learning_rate": 9.132653061224489e-07, "loss": 0.0011, "num_tokens": 9204638.0, "reward": 0.0546875, "reward_std": 0.26899558305740356, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 154, "step_time": 48.77144674025476 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1480.0, "completions/max_terminated_length": 1480.0, "completions/mean_length": 802.421875, "completions/mean_terminated_length": 802.421875, "completions/min_length": 360.0, "completions/min_terminated_length": 360.0, "entropy": 0.8584241755306721, "epoch": 0.08786848072562359, "frac_reward_zero_std": 0.625, "grad_norm": 0.439453125, "learning_rate": 9.126984126984127e-07, "loss": 0.0057, "num_tokens": 9270737.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 155, "step_time": 58.38633061386645 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 786.234375, "completions/mean_terminated_length": 786.234375, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "entropy": 0.7147355619817972, "epoch": 0.08843537414965986, "frac_reward_zero_std": 0.625, "grad_norm": 0.408203125, "learning_rate": 9.121315192743763e-07, "loss": -0.019, "num_tokens": 9332240.0, "reward": 0.1328125, "reward_std": 0.489915668964386, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 156, "step_time": 47.72322576586157 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1425.0, "completions/max_terminated_length": 1425.0, "completions/mean_length": 698.53125, "completions/mean_terminated_length": 698.53125, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "entropy": 0.9679968729615211, "epoch": 0.08900226757369614, "frac_reward_zero_std": 0.5, "grad_norm": 0.4765625, "learning_rate": 9.115646258503401e-07, "loss": -0.0082, "num_tokens": 9387138.0, "reward": 0.21875, "reward_std": 0.5764904618263245, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 157, "step_time": 54.63682753685862 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1328.0, "completions/max_terminated_length": 1328.0, "completions/mean_length": 686.03125, "completions/mean_terminated_length": 686.03125, "completions/min_length": 9.0, "completions/min_terminated_length": 9.0, "entropy": 0.8306704685091972, "epoch": 0.08956916099773243, "frac_reward_zero_std": 0.625, "grad_norm": 0.5625, "learning_rate": 9.109977324263038e-07, "loss": -0.0268, "num_tokens": 9439876.0, "reward": 0.1015625, "reward_std": 0.42891231179237366, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 158, "step_time": 50.167413502931595 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1665.0, "completions/max_terminated_length": 1665.0, "completions/mean_length": 804.484375, "completions/mean_terminated_length": 804.484375, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "entropy": 0.7235718593001366, "epoch": 0.09013605442176871, "frac_reward_zero_std": 0.5, "grad_norm": 0.4296875, "learning_rate": 9.104308390022676e-07, "loss": 0.0075, "num_tokens": 9502635.0, "reward": 0.1015625, "reward_std": 0.42891231179237366, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 159, "step_time": 63.26405255496502 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1866.0, "completions/max_terminated_length": 1866.0, "completions/mean_length": 707.109375, "completions/mean_terminated_length": 707.109375, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "entropy": 0.8616630770266056, "epoch": 0.09070294784580499, "frac_reward_zero_std": 0.75, "grad_norm": 0.373046875, "learning_rate": 9.098639455782312e-07, "loss": 0.0051, "num_tokens": 9557034.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 160, "step_time": 70.56565687526017 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 768.59375, "completions/mean_terminated_length": 768.59375, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "entropy": 0.9379465486854315, "epoch": 0.09126984126984126, "frac_reward_zero_std": 0.25, "grad_norm": 0.62890625, "learning_rate": 9.09297052154195e-07, "loss": 0.0187, "num_tokens": 9617640.0, "reward": 0.25, "reward_std": 0.6546536684036255, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 161, "step_time": 50.923309312202036 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1423.0, "completions/max_terminated_length": 1423.0, "completions/mean_length": 716.359375, "completions/mean_terminated_length": 716.359375, "completions/min_length": 321.0, "completions/min_terminated_length": 321.0, "entropy": 0.589280379936099, "epoch": 0.09183673469387756, "frac_reward_zero_std": 0.375, "grad_norm": 0.51953125, "learning_rate": 9.087301587301587e-07, "loss": 0.0246, "num_tokens": 9672767.0, "reward": 0.234375, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 162, "step_time": 53.908720002509654 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1277.0, "completions/max_terminated_length": 1277.0, "completions/mean_length": 778.640625, "completions/mean_terminated_length": 778.640625, "completions/min_length": 104.0, "completions/min_terminated_length": 104.0, "entropy": 0.9540521949529648, "epoch": 0.09240362811791383, "frac_reward_zero_std": 0.75, "grad_norm": 0.359375, "learning_rate": 9.081632653061225e-07, "loss": 0.009, "num_tokens": 9733720.0, "reward": 0.1171875, "reward_std": 0.43408530950546265, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 163, "step_time": 48.228524887003005 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 798.796875, "completions/mean_terminated_length": 798.796875, "completions/min_length": 430.0, "completions/min_terminated_length": 430.0, "entropy": 0.9369715973734856, "epoch": 0.09297052154195011, "frac_reward_zero_std": 0.625, "grad_norm": 0.458984375, "learning_rate": 9.075963718820861e-07, "loss": -0.0221, "num_tokens": 9796915.0, "reward": 0.125, "reward_std": 0.4364357888698578, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 164, "step_time": 61.795095819979906 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1539.0, "completions/max_terminated_length": 1539.0, "completions/mean_length": 717.359375, "completions/mean_terminated_length": 717.359375, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "entropy": 0.8061159290373325, "epoch": 0.0935374149659864, "frac_reward_zero_std": 0.75, "grad_norm": 0.337890625, "learning_rate": 9.070294784580499e-07, "loss": 0.0034, "num_tokens": 9851794.0, "reward": 0.109375, "reward_std": 0.41277948021888733, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 165, "step_time": 58.08895343821496 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 686.625, "completions/mean_terminated_length": 686.625, "completions/min_length": 311.0, "completions/min_terminated_length": 311.0, "entropy": 0.9296346679329872, "epoch": 0.09410430839002268, "frac_reward_zero_std": 0.875, "grad_norm": 0.3125, "learning_rate": 9.064625850340135e-07, "loss": 0.0125, "num_tokens": 9906666.0, "reward": 0.03125, "reward_std": 0.12198750674724579, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 166, "step_time": 52.012140300124884 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1470.0, "completions/max_terminated_length": 1470.0, "completions/mean_length": 749.0, "completions/mean_terminated_length": 749.0, "completions/min_length": 316.0, "completions/min_terminated_length": 316.0, "entropy": 0.6455147042870522, "epoch": 0.09467120181405896, "frac_reward_zero_std": 0.5, "grad_norm": 0.435546875, "learning_rate": 9.058956916099772e-07, "loss": 0.0119, "num_tokens": 9964658.0, "reward": 0.3828125, "reward_std": 0.8344857692718506, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 167, "step_time": 56.627189056016505 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1420.0, "completions/max_terminated_length": 1420.0, "completions/mean_length": 728.671875, "completions/mean_terminated_length": 728.671875, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "entropy": 0.8471074942499399, "epoch": 0.09523809523809523, "frac_reward_zero_std": 0.5, "grad_norm": 0.49609375, "learning_rate": 9.05328798185941e-07, "loss": -0.0208, "num_tokens": 10022517.0, "reward": 0.1875, "reward_std": 0.5455447435379028, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 168, "step_time": 55.34592215809971 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1452.0, "completions/max_terminated_length": 1452.0, "completions/mean_length": 789.671875, "completions/mean_terminated_length": 789.671875, "completions/min_length": 349.0, "completions/min_terminated_length": 349.0, "entropy": 0.9063250534236431, "epoch": 0.09580498866213152, "frac_reward_zero_std": 0.875, "grad_norm": 0.28515625, "learning_rate": 9.047619047619047e-07, "loss": 0.0261, "num_tokens": 10084376.0, "reward": 0.0390625, "reward_std": 0.13524484634399414, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 169, "step_time": 56.44608405418694 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 740.6875, "completions/mean_terminated_length": 740.6875, "completions/min_length": 319.0, "completions/min_terminated_length": 319.0, "entropy": 0.9391395784914494, "epoch": 0.0963718820861678, "frac_reward_zero_std": 0.625, "grad_norm": 0.435546875, "learning_rate": 9.041950113378685e-07, "loss": -0.0212, "num_tokens": 10145532.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 170, "step_time": 45.221647115424275 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 722.953125, "completions/mean_terminated_length": 722.953125, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "entropy": 0.8834142237901688, "epoch": 0.09693877551020408, "frac_reward_zero_std": 0.75, "grad_norm": 0.345703125, "learning_rate": 9.036281179138322e-07, "loss": 0.009, "num_tokens": 10200745.0, "reward": 0.2109375, "reward_std": 0.5896238088607788, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 171, "step_time": 66.33940491080284 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1845.0, "completions/max_terminated_length": 1845.0, "completions/mean_length": 715.828125, "completions/mean_terminated_length": 715.828125, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "entropy": 0.9365953505039215, "epoch": 0.09750566893424037, "frac_reward_zero_std": 0.875, "grad_norm": 0.275390625, "learning_rate": 9.030612244897958e-07, "loss": 0.0068, "num_tokens": 10256590.0, "reward": 0.0390625, "reward_std": 0.13524484634399414, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 172, "step_time": 72.47474573273212 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1415.0, "completions/max_terminated_length": 1415.0, "completions/mean_length": 774.4375, "completions/mean_terminated_length": 774.4375, "completions/min_length": 278.0, "completions/min_terminated_length": 278.0, "entropy": 1.0347714982926846, "epoch": 0.09807256235827665, "frac_reward_zero_std": 0.625, "grad_norm": 0.447265625, "learning_rate": 9.024943310657595e-07, "loss": -0.0001, "num_tokens": 10316010.0, "reward": 0.203125, "reward_std": 0.5754139423370361, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 173, "step_time": 54.19198610819876 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1635.0, "completions/max_terminated_length": 1635.0, "completions/mean_length": 823.203125, "completions/mean_terminated_length": 823.203125, "completions/min_length": 434.0, "completions/min_terminated_length": 434.0, "entropy": 0.9763973355293274, "epoch": 0.09863945578231292, "frac_reward_zero_std": 0.875, "grad_norm": 0.255859375, "learning_rate": 9.019274376417233e-07, "loss": 0.0109, "num_tokens": 10379279.0, "reward": 0.015625, "reward_std": 0.08768405020236969, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 174, "step_time": 63.09931514598429 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1525.0, "completions/max_terminated_length": 1525.0, "completions/mean_length": 809.859375, "completions/mean_terminated_length": 809.859375, "completions/min_length": 410.0, "completions/min_terminated_length": 410.0, "entropy": 0.8472088351845741, "epoch": 0.0992063492063492, "frac_reward_zero_std": 0.75, "grad_norm": 0.30078125, "learning_rate": 9.01360544217687e-07, "loss": -0.018, "num_tokens": 10444046.0, "reward": 0.0546875, "reward_std": 0.26899558305740356, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 175, "step_time": 59.40772425476462 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 773.359375, "completions/mean_terminated_length": 773.359375, "completions/min_length": 403.0, "completions/min_terminated_length": 403.0, "entropy": 0.889940544962883, "epoch": 0.09977324263038549, "frac_reward_zero_std": 0.75, "grad_norm": 0.40234375, "learning_rate": 9.007936507936508e-07, "loss": 0.0082, "num_tokens": 10503645.0, "reward": 0.15625, "reward_std": 0.5261042714118958, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 176, "step_time": 45.98833618219942 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 721.875, "completions/mean_terminated_length": 721.875, "completions/min_length": 284.0, "completions/min_terminated_length": 284.0, "entropy": 0.5890161376446486, "epoch": 0.10034013605442177, "frac_reward_zero_std": 0.625, "grad_norm": 0.41015625, "learning_rate": 9.002267573696145e-07, "loss": -0.0192, "num_tokens": 10559589.0, "reward": 0.1796875, "reward_std": 0.5298036932945251, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 177, "step_time": 48.5907125454396 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1195.0, "completions/max_terminated_length": 1195.0, "completions/mean_length": 707.140625, "completions/mean_terminated_length": 707.140625, "completions/min_length": 21.0, "completions/min_terminated_length": 21.0, "entropy": 1.0144722759723663, "epoch": 0.10090702947845805, "frac_reward_zero_std": 0.75, "grad_norm": 0.375, "learning_rate": 8.996598639455783e-07, "loss": 0.0149, "num_tokens": 10614134.0, "reward": 0.0625, "reward_std": 0.2745848298072815, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 178, "step_time": 45.42839828785509 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1314.0, "completions/max_terminated_length": 1314.0, "completions/mean_length": 801.234375, "completions/mean_terminated_length": 801.234375, "completions/min_length": 304.0, "completions/min_terminated_length": 304.0, "entropy": 0.7205090411007404, "epoch": 0.10147392290249434, "frac_reward_zero_std": 0.875, "grad_norm": 0.25, "learning_rate": 8.990929705215419e-07, "loss": -0.0033, "num_tokens": 10677061.0, "reward": 0.046875, "reward_std": 0.14689241349697113, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 179, "step_time": 50.42318244557828 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1579.0, "completions/max_terminated_length": 1579.0, "completions/mean_length": 756.328125, "completions/mean_terminated_length": 756.328125, "completions/min_length": 11.0, "completions/min_terminated_length": 11.0, "entropy": 0.7697293981909752, "epoch": 0.10204081632653061, "frac_reward_zero_std": 0.625, "grad_norm": 0.474609375, "learning_rate": 8.985260770975056e-07, "loss": 0.0173, "num_tokens": 10738306.0, "reward": 0.1015625, "reward_std": 0.3692515790462494, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 180, "step_time": 60.82005000580102 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2502.0, "completions/max_terminated_length": 2502.0, "completions/mean_length": 806.46875, "completions/mean_terminated_length": 806.46875, "completions/min_length": 337.0, "completions/min_terminated_length": 337.0, "entropy": 0.8854840584099293, "epoch": 0.10260770975056689, "frac_reward_zero_std": 0.75, "grad_norm": 0.373046875, "learning_rate": 8.979591836734693e-07, "loss": -0.0048, "num_tokens": 10801224.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 181, "step_time": 99.66593520808965 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1445.0, "completions/max_terminated_length": 1445.0, "completions/mean_length": 681.46875, "completions/mean_terminated_length": 681.46875, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "entropy": 0.833324022591114, "epoch": 0.10317460317460317, "frac_reward_zero_std": 0.625, "grad_norm": 0.50390625, "learning_rate": 8.973922902494331e-07, "loss": -0.0122, "num_tokens": 10854622.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 182, "step_time": 56.18346180021763 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1328.0, "completions/mean_length": 756.9375, "completions/mean_terminated_length": 703.9365234375, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "entropy": 0.8603761028498411, "epoch": 0.10374149659863946, "frac_reward_zero_std": 0.625, "grad_norm": 0.4765625, "learning_rate": 8.968253968253968e-07, "loss": -0.0064, "num_tokens": 10912946.0, "reward": 0.09375, "reward_std": 0.36596253514289856, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 183, "step_time": 176.22284319531173 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1420.0, "completions/max_terminated_length": 1420.0, "completions/mean_length": 771.984375, "completions/mean_terminated_length": 771.984375, "completions/min_length": 351.0, "completions/min_terminated_length": 351.0, "entropy": 0.679333757609129, "epoch": 0.10430839002267574, "frac_reward_zero_std": 0.375, "grad_norm": 0.5625, "learning_rate": 8.962585034013606e-07, "loss": -0.0128, "num_tokens": 10972257.0, "reward": 0.21875, "reward_std": 0.5900968313217163, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 184, "step_time": 54.31486776098609 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2259.0, "completions/max_terminated_length": 2259.0, "completions/mean_length": 856.28125, "completions/mean_terminated_length": 856.28125, "completions/min_length": 364.0, "completions/min_terminated_length": 364.0, "entropy": 0.8160925414413214, "epoch": 0.10487528344671201, "frac_reward_zero_std": 0.375, "grad_norm": 0.59375, "learning_rate": 8.956916099773242e-07, "loss": 0.0073, "num_tokens": 11036563.0, "reward": 0.171875, "reward_std": 0.5434950590133667, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 185, "step_time": 88.11578060686588 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1472.0, "completions/max_terminated_length": 1472.0, "completions/mean_length": 739.234375, "completions/mean_terminated_length": 739.234375, "completions/min_length": 296.0, "completions/min_terminated_length": 296.0, "entropy": 0.7380147892981768, "epoch": 0.1054421768707483, "frac_reward_zero_std": 0.875, "grad_norm": 0.236328125, "learning_rate": 8.95124716553288e-07, "loss": 0.0029, "num_tokens": 11095386.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 186, "step_time": 56.094907987862825 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1792.0, "completions/max_terminated_length": 1792.0, "completions/mean_length": 750.96875, "completions/mean_terminated_length": 750.96875, "completions/min_length": 350.0, "completions/min_terminated_length": 350.0, "entropy": 0.642386669293046, "epoch": 0.10600907029478458, "frac_reward_zero_std": 0.625, "grad_norm": 0.375, "learning_rate": 8.945578231292517e-07, "loss": 0.0299, "num_tokens": 11153080.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 187, "step_time": 67.72712143044919 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1212.0, "completions/max_terminated_length": 1212.0, "completions/mean_length": 745.625, "completions/mean_terminated_length": 745.625, "completions/min_length": 307.0, "completions/min_terminated_length": 307.0, "entropy": 0.6487344261258841, "epoch": 0.10657596371882086, "frac_reward_zero_std": 0.375, "grad_norm": 0.54296875, "learning_rate": 8.939909297052153e-07, "loss": -0.0064, "num_tokens": 11210912.0, "reward": 0.265625, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 188, "step_time": 45.404655010439456 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1832.0, "completions/max_terminated_length": 1832.0, "completions/mean_length": 784.28125, "completions/mean_terminated_length": 784.28125, "completions/min_length": 369.0, "completions/min_terminated_length": 369.0, "entropy": 0.9743715263903141, "epoch": 0.10714285714285714, "frac_reward_zero_std": 0.5, "grad_norm": 0.578125, "learning_rate": 8.934240362811791e-07, "loss": -0.0049, "num_tokens": 11272802.0, "reward": 0.1171875, "reward_std": 0.43408530950546265, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 189, "step_time": 69.66925997380167 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1579.0, "completions/max_terminated_length": 1579.0, "completions/mean_length": 730.0625, "completions/mean_terminated_length": 730.0625, "completions/min_length": 264.0, "completions/min_terminated_length": 264.0, "entropy": 0.8052809778600931, "epoch": 0.10770975056689343, "frac_reward_zero_std": 0.375, "grad_norm": 0.62890625, "learning_rate": 8.928571428571428e-07, "loss": -0.0014, "num_tokens": 11329454.0, "reward": 0.1328125, "reward_std": 0.489915668964386, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 190, "step_time": 59.718092283234 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1164.0, "completions/max_terminated_length": 1164.0, "completions/mean_length": 689.609375, "completions/mean_terminated_length": 689.609375, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "entropy": 0.8131397124379873, "epoch": 0.1082766439909297, "frac_reward_zero_std": 0.375, "grad_norm": 0.65234375, "learning_rate": 8.922902494331065e-07, "loss": 0.0137, "num_tokens": 11386005.0, "reward": 0.375, "reward_std": 0.745356023311615, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 191, "step_time": 44.201239502988756 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 770.609375, "completions/mean_terminated_length": 770.609375, "completions/min_length": 369.0, "completions/min_terminated_length": 369.0, "entropy": 0.8108047619462013, "epoch": 0.10884353741496598, "frac_reward_zero_std": 0.625, "grad_norm": 0.43359375, "learning_rate": 8.917233560090702e-07, "loss": 0.0036, "num_tokens": 11445404.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 192, "step_time": 52.274463404901326 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1482.0, "completions/max_terminated_length": 1482.0, "completions/mean_length": 731.984375, "completions/mean_terminated_length": 731.984375, "completions/min_length": 320.0, "completions/min_terminated_length": 320.0, "entropy": 0.7214385736733675, "epoch": 0.10941043083900227, "frac_reward_zero_std": 0.875, "grad_norm": 0.294921875, "learning_rate": 8.91156462585034e-07, "loss": 0.0048, "num_tokens": 11501987.0, "reward": 0.03125, "reward_std": 0.25, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 193, "step_time": 56.17669538129121 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1884.0, "completions/max_terminated_length": 1884.0, "completions/mean_length": 858.609375, "completions/mean_terminated_length": 858.609375, "completions/min_length": 422.0, "completions/min_terminated_length": 422.0, "entropy": 0.5905744098126888, "epoch": 0.10997732426303855, "frac_reward_zero_std": 0.625, "grad_norm": 0.37890625, "learning_rate": 8.905895691609977e-07, "loss": -0.0082, "num_tokens": 11565898.0, "reward": 0.09375, "reward_std": 0.36596253514289856, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 194, "step_time": 71.28613732848316 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 710.1875, "completions/mean_terminated_length": 710.1875, "completions/min_length": 217.0, "completions/min_terminated_length": 217.0, "entropy": 0.8393162861466408, "epoch": 0.11054421768707483, "frac_reward_zero_std": 0.625, "grad_norm": 0.490234375, "learning_rate": 8.900226757369615e-07, "loss": -0.012, "num_tokens": 11620614.0, "reward": 0.1875, "reward_std": 0.5739033818244934, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 195, "step_time": 48.9538035010919 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1361.0, "completions/max_terminated_length": 1361.0, "completions/mean_length": 799.625, "completions/mean_terminated_length": 799.625, "completions/min_length": 291.0, "completions/min_terminated_length": 291.0, "entropy": 0.7309198118746281, "epoch": 0.1111111111111111, "frac_reward_zero_std": 0.375, "grad_norm": 0.60546875, "learning_rate": 8.894557823129251e-07, "loss": -0.0092, "num_tokens": 11686094.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 196, "step_time": 54.0608956348151 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 708.078125, "completions/mean_terminated_length": 708.078125, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "entropy": 0.8973634522408247, "epoch": 0.1116780045351474, "frac_reward_zero_std": 0.625, "grad_norm": 0.44921875, "learning_rate": 8.888888888888888e-07, "loss": 0.0047, "num_tokens": 11743739.0, "reward": 0.21875, "reward_std": 0.5900968313217163, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 197, "step_time": 47.21101175248623 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 664.328125, "completions/mean_terminated_length": 664.328125, "completions/min_length": 283.0, "completions/min_terminated_length": 283.0, "entropy": 0.8888151589781046, "epoch": 0.11224489795918367, "frac_reward_zero_std": 0.75, "grad_norm": 0.3359375, "learning_rate": 8.883219954648525e-07, "loss": -0.0159, "num_tokens": 11797512.0, "reward": 0.15625, "reward_std": 0.4787135720252991, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 198, "step_time": 47.433251860551536 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 736.75, "completions/mean_terminated_length": 736.75, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "entropy": 0.617466252297163, "epoch": 0.11281179138321995, "frac_reward_zero_std": 0.375, "grad_norm": 0.5625, "learning_rate": 8.877551020408163e-07, "loss": 0.0137, "num_tokens": 11853832.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 199, "step_time": 67.18875078298151 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1274.0, "completions/max_terminated_length": 1274.0, "completions/mean_length": 754.609375, "completions/mean_terminated_length": 754.609375, "completions/min_length": 461.0, "completions/min_terminated_length": 461.0, "entropy": 0.993266124278307, "epoch": 0.11337868480725624, "frac_reward_zero_std": 0.5, "grad_norm": 0.53125, "learning_rate": 8.8718820861678e-07, "loss": -0.0036, "num_tokens": 11912055.0, "reward": 0.15625, "reward_std": 0.49501484632492065, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 200, "step_time": 48.555252633057535 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1785.0, "completions/max_terminated_length": 1785.0, "completions/mean_length": 734.9375, "completions/mean_terminated_length": 734.9375, "completions/min_length": 393.0, "completions/min_terminated_length": 393.0, "entropy": 0.8032393399626017, "epoch": 0.11394557823129252, "frac_reward_zero_std": 0.5, "grad_norm": 0.46875, "learning_rate": 8.866213151927438e-07, "loss": 0.0151, "num_tokens": 11969403.0, "reward": 0.3125, "reward_std": 0.6986381411552429, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 201, "step_time": 68.53788980189711 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1584.0, "completions/max_terminated_length": 1584.0, "completions/mean_length": 687.6875, "completions/mean_terminated_length": 687.6875, "completions/min_length": 217.0, "completions/min_terminated_length": 217.0, "entropy": 0.8932653181254864, "epoch": 0.1145124716553288, "frac_reward_zero_std": 0.625, "grad_norm": 0.486328125, "learning_rate": 8.860544217687075e-07, "loss": 0.0114, "num_tokens": 12025543.0, "reward": 0.15625, "reward_std": 0.49501484632492065, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 202, "step_time": 61.52193641010672 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1410.0, "completions/max_terminated_length": 1410.0, "completions/mean_length": 721.234375, "completions/mean_terminated_length": 721.234375, "completions/min_length": 364.0, "completions/min_terminated_length": 364.0, "entropy": 1.0084965340793133, "epoch": 0.11507936507936507, "frac_reward_zero_std": 0.875, "grad_norm": 0.3203125, "learning_rate": 8.854875283446712e-07, "loss": 0.0078, "num_tokens": 12081302.0, "reward": 0.0234375, "reward_std": 0.10652101039886475, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 203, "step_time": 53.666412461549044 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 768.046875, "completions/mean_terminated_length": 768.046875, "completions/min_length": 344.0, "completions/min_terminated_length": 344.0, "entropy": 0.9337150938808918, "epoch": 0.11564625850340136, "frac_reward_zero_std": 0.75, "grad_norm": 0.349609375, "learning_rate": 8.849206349206348e-07, "loss": 0.0497, "num_tokens": 12140425.0, "reward": 0.0859375, "reward_std": 0.36247265338897705, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 204, "step_time": 52.04016512725502 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1596.0, "completions/max_terminated_length": 1596.0, "completions/mean_length": 843.734375, "completions/mean_terminated_length": 843.734375, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "entropy": 0.617034338414669, "epoch": 0.11621315192743764, "frac_reward_zero_std": 0.875, "grad_norm": 0.1806640625, "learning_rate": 8.843537414965985e-07, "loss": 0.0015, "num_tokens": 12203368.0, "reward": 0.140625, "reward_std": 0.5379911661148071, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 205, "step_time": 60.39713575411588 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2372.0, "completions/max_terminated_length": 2372.0, "completions/mean_length": 866.125, "completions/mean_terminated_length": 866.125, "completions/min_length": 381.0, "completions/min_terminated_length": 381.0, "entropy": 0.6286310963332653, "epoch": 0.11678004535147392, "frac_reward_zero_std": 0.625, "grad_norm": 0.3515625, "learning_rate": 8.837868480725623e-07, "loss": 0.021, "num_tokens": 12270552.0, "reward": 0.2265625, "reward_std": 0.6662480235099792, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 206, "step_time": 92.44137762580067 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 693.765625, "completions/mean_terminated_length": 693.765625, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "entropy": 0.7869409117847681, "epoch": 0.11734693877551021, "frac_reward_zero_std": 0.625, "grad_norm": 0.439453125, "learning_rate": 8.83219954648526e-07, "loss": 0.0016, "num_tokens": 12326353.0, "reward": 0.1796875, "reward_std": 0.5445777773857117, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 207, "step_time": 53.71699774544686 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 710.984375, "completions/mean_terminated_length": 710.984375, "completions/min_length": 356.0, "completions/min_terminated_length": 356.0, "entropy": 0.8867442682385445, "epoch": 0.11791383219954649, "frac_reward_zero_std": 0.625, "grad_norm": 0.49609375, "learning_rate": 8.826530612244898e-07, "loss": 0.0062, "num_tokens": 12381736.0, "reward": 0.09375, "reward_std": 0.36596253514289856, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 208, "step_time": 54.03470201138407 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 801.8125, "completions/mean_terminated_length": 801.8125, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 0.6637982688844204, "epoch": 0.11848072562358276, "frac_reward_zero_std": 0.75, "grad_norm": 0.400390625, "learning_rate": 8.820861678004535e-07, "loss": 0.0169, "num_tokens": 12442460.0, "reward": 0.046875, "reward_std": 0.14689241349697113, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 209, "step_time": 55.399893024004996 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1073.0, "completions/max_terminated_length": 1073.0, "completions/mean_length": 677.4375, "completions/mean_terminated_length": 677.4375, "completions/min_length": 335.0, "completions/min_terminated_length": 335.0, "entropy": 0.8804828450083733, "epoch": 0.11904761904761904, "frac_reward_zero_std": 0.75, "grad_norm": 0.4609375, "learning_rate": 8.815192743764172e-07, "loss": 0.0067, "num_tokens": 12495944.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 210, "step_time": 40.4301584828645 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1643.0, "completions/max_terminated_length": 1643.0, "completions/mean_length": 783.734375, "completions/mean_terminated_length": 783.734375, "completions/min_length": 327.0, "completions/min_terminated_length": 327.0, "entropy": 0.8268555216491222, "epoch": 0.11961451247165533, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 8.809523809523809e-07, "loss": 0.03, "num_tokens": 12555895.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 211, "step_time": 61.941131071187556 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1534.0, "completions/max_terminated_length": 1534.0, "completions/mean_length": 808.421875, "completions/mean_terminated_length": 808.421875, "completions/min_length": 241.0, "completions/min_terminated_length": 241.0, "entropy": 0.6941155325621367, "epoch": 0.12018140589569161, "frac_reward_zero_std": 0.625, "grad_norm": 0.408203125, "learning_rate": 8.803854875283446e-07, "loss": -0.0081, "num_tokens": 12617538.0, "reward": 0.0859375, "reward_std": 0.36247265338897705, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 212, "step_time": 58.329736151732504 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1460.0, "completions/max_terminated_length": 1460.0, "completions/mean_length": 784.46875, "completions/mean_terminated_length": 784.46875, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "entropy": 0.7919981423765421, "epoch": 0.12074829931972789, "frac_reward_zero_std": 0.875, "grad_norm": 0.19140625, "learning_rate": 8.798185941043083e-07, "loss": 0.0028, "num_tokens": 12679336.0, "reward": 0.0859375, "reward_std": 0.3398725986480713, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 213, "step_time": 55.956489546224475 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 754.890625, "completions/mean_terminated_length": 754.890625, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "entropy": 0.7367960270494223, "epoch": 0.12131519274376418, "frac_reward_zero_std": 0.625, "grad_norm": 0.419921875, "learning_rate": 8.792517006802721e-07, "loss": -0.017, "num_tokens": 12738401.0, "reward": 0.1171875, "reward_std": 0.43408530950546265, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 214, "step_time": 55.02792660146952 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 758.40625, "completions/mean_terminated_length": 758.40625, "completions/min_length": 418.0, "completions/min_terminated_length": 418.0, "entropy": 0.6321735177189112, "epoch": 0.12188208616780045, "frac_reward_zero_std": 0.625, "grad_norm": 0.435546875, "learning_rate": 8.786848072562358e-07, "loss": 0.0405, "num_tokens": 12797347.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 215, "step_time": 58.47775779198855 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1197.0, "completions/max_terminated_length": 1197.0, "completions/mean_length": 781.21875, "completions/mean_terminated_length": 781.21875, "completions/min_length": 352.0, "completions/min_terminated_length": 352.0, "entropy": 0.970495417714119, "epoch": 0.12244897959183673, "frac_reward_zero_std": 0.625, "grad_norm": 0.46484375, "learning_rate": 8.781179138321995e-07, "loss": 0.0257, "num_tokens": 12861937.0, "reward": 0.1640625, "reward_std": 0.5422958731651306, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0078125, "rewards/format_reward_func/std": 0.0625, "step": 216, "step_time": 48.29055262543261 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 863.328125, "completions/mean_terminated_length": 863.328125, "completions/min_length": 445.0, "completions/min_terminated_length": 445.0, "entropy": 0.7418898530304432, "epoch": 0.12301587301587301, "frac_reward_zero_std": 0.5, "grad_norm": 0.52734375, "learning_rate": 8.775510204081632e-07, "loss": -0.0217, "num_tokens": 12927302.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 217, "step_time": 57.822010475210845 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1969.0, "completions/max_terminated_length": 1969.0, "completions/mean_length": 851.828125, "completions/mean_terminated_length": 851.828125, "completions/min_length": 345.0, "completions/min_terminated_length": 345.0, "entropy": 0.876748263835907, "epoch": 0.1235827664399093, "frac_reward_zero_std": 0.625, "grad_norm": 0.412109375, "learning_rate": 8.76984126984127e-07, "loss": 0.0152, "num_tokens": 12992571.0, "reward": 0.1796875, "reward_std": 0.5445777773857117, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 218, "step_time": 76.29621310438961 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2382.0, "completions/max_terminated_length": 2382.0, "completions/mean_length": 736.34375, "completions/mean_terminated_length": 736.34375, "completions/min_length": 170.0, "completions/min_terminated_length": 170.0, "entropy": 0.7815339379012585, "epoch": 0.12414965986394558, "frac_reward_zero_std": 0.625, "grad_norm": 0.419921875, "learning_rate": 8.764172335600907e-07, "loss": -0.0102, "num_tokens": 13049665.0, "reward": 0.328125, "reward_std": 0.6970388293266296, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 219, "step_time": 90.68243482988328 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 751.296875, "completions/mean_terminated_length": 751.296875, "completions/min_length": 294.0, "completions/min_terminated_length": 294.0, "entropy": 0.8732363618910313, "epoch": 0.12471655328798185, "frac_reward_zero_std": 0.875, "grad_norm": 0.2216796875, "learning_rate": 8.758503401360544e-07, "loss": -0.0087, "num_tokens": 13108468.0, "reward": 0.0546875, "reward_std": 0.15728822350502014, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 220, "step_time": 54.02226519025862 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1302.0, "completions/max_terminated_length": 1302.0, "completions/mean_length": 800.765625, "completions/mean_terminated_length": 800.765625, "completions/min_length": 450.0, "completions/min_terminated_length": 450.0, "entropy": 0.7857203856110573, "epoch": 0.12528344671201813, "frac_reward_zero_std": 0.625, "grad_norm": 0.376953125, "learning_rate": 8.752834467120181e-07, "loss": 0.0214, "num_tokens": 13170357.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 221, "step_time": 49.31268990319222 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1215.0, "completions/max_terminated_length": 1215.0, "completions/mean_length": 746.875, "completions/mean_terminated_length": 746.875, "completions/min_length": 366.0, "completions/min_terminated_length": 366.0, "entropy": 0.9006501343101263, "epoch": 0.12585034013605442, "frac_reward_zero_std": 0.625, "grad_norm": 0.451171875, "learning_rate": 8.747165532879818e-07, "loss": 0.0168, "num_tokens": 13230621.0, "reward": 0.2265625, "reward_std": 0.6167611479759216, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 222, "step_time": 47.496775266714394 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1563.0, "completions/max_terminated_length": 1563.0, "completions/mean_length": 678.0625, "completions/mean_terminated_length": 678.0625, "completions/min_length": 326.0, "completions/min_terminated_length": 326.0, "entropy": 0.9045848734676838, "epoch": 0.1264172335600907, "frac_reward_zero_std": 0.625, "grad_norm": 0.54296875, "learning_rate": 8.741496598639455e-07, "loss": -0.014, "num_tokens": 13283545.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 223, "step_time": 59.5722237136215 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1134.0, "completions/max_terminated_length": 1134.0, "completions/mean_length": 696.53125, "completions/mean_terminated_length": 696.53125, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "entropy": 0.6746862567961216, "epoch": 0.12698412698412698, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 8.735827664399093e-07, "loss": 0.029, "num_tokens": 13337867.0, "reward": 0.2265625, "reward_std": 0.5904644727706909, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 224, "step_time": 42.94442406669259 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1520.0, "completions/max_terminated_length": 1520.0, "completions/mean_length": 682.421875, "completions/mean_terminated_length": 682.421875, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "entropy": 0.9444940574467182, "epoch": 0.12755102040816327, "frac_reward_zero_std": 0.75, "grad_norm": 0.376953125, "learning_rate": 8.73015873015873e-07, "loss": -0.0031, "num_tokens": 13397838.0, "reward": 0.15625, "reward_std": 0.5107961297035217, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 225, "step_time": 61.212527259252965 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 4083.0, "completions/mean_length": 863.296875, "completions/mean_terminated_length": 811.9841918945312, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "entropy": 0.5170101718977094, "epoch": 0.12811791383219956, "frac_reward_zero_std": 0.625, "grad_norm": 0.369140625, "learning_rate": 8.724489795918368e-07, "loss": 0.0063, "num_tokens": 13462673.0, "reward": 0.2578125, "reward_std": 0.6784212589263916, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 226, "step_time": 175.77780503034592 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 737.75, "completions/mean_terminated_length": 737.75, "completions/min_length": 28.0, "completions/min_terminated_length": 28.0, "entropy": 0.8340022526681423, "epoch": 0.12868480725623582, "frac_reward_zero_std": 0.75, "grad_norm": 0.3828125, "learning_rate": 8.718820861678005e-07, "loss": -0.0183, "num_tokens": 13520017.0, "reward": 0.09375, "reward_std": 0.36596253514289856, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 227, "step_time": 51.04294812493026 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1305.0, "completions/max_terminated_length": 1305.0, "completions/mean_length": 688.578125, "completions/mean_terminated_length": 688.578125, "completions/min_length": 308.0, "completions/min_terminated_length": 308.0, "entropy": 0.8905489556491375, "epoch": 0.1292517006802721, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 8.71315192743764e-07, "loss": 0.002, "num_tokens": 13573430.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 228, "step_time": 49.98010636493564 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 823.40625, "completions/mean_terminated_length": 823.40625, "completions/min_length": 394.0, "completions/min_terminated_length": 394.0, "entropy": 0.7021019477397203, "epoch": 0.12981859410430838, "frac_reward_zero_std": 0.625, "grad_norm": 0.388671875, "learning_rate": 8.707482993197278e-07, "loss": 0.0176, "num_tokens": 13636056.0, "reward": 0.2109375, "reward_std": 0.6029338836669922, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 229, "step_time": 56.998927718028426 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 683.25, "completions/mean_terminated_length": 683.25, "completions/min_length": 273.0, "completions/min_terminated_length": 273.0, "entropy": 0.8584277145564556, "epoch": 0.13038548752834467, "frac_reward_zero_std": 0.75, "grad_norm": 0.404296875, "learning_rate": 8.701814058956915e-07, "loss": 0.0128, "num_tokens": 13690056.0, "reward": 0.0859375, "reward_std": 0.36247265338897705, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 230, "step_time": 54.34538010787219 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1381.0, "completions/max_terminated_length": 1381.0, "completions/mean_length": 786.390625, "completions/mean_terminated_length": 786.390625, "completions/min_length": 239.0, "completions/min_terminated_length": 239.0, "entropy": 0.7083199769258499, "epoch": 0.13095238095238096, "frac_reward_zero_std": 0.375, "grad_norm": 0.546875, "learning_rate": 8.696145124716553e-07, "loss": 0.0133, "num_tokens": 13749385.0, "reward": 0.265625, "reward_std": 0.6664806604385376, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 231, "step_time": 52.82335822377354 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 731.09375, "completions/mean_terminated_length": 731.09375, "completions/min_length": 427.0, "completions/min_terminated_length": 427.0, "entropy": 0.6988869290798903, "epoch": 0.13151927437641722, "frac_reward_zero_std": 0.625, "grad_norm": 0.47265625, "learning_rate": 8.69047619047619e-07, "loss": -0.0041, "num_tokens": 13807071.0, "reward": 0.09375, "reward_std": 0.36596253514289856, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 232, "step_time": 49.35344780609012 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 773.171875, "completions/mean_terminated_length": 773.171875, "completions/min_length": 297.0, "completions/min_terminated_length": 297.0, "entropy": 0.7230974994599819, "epoch": 0.1320861678004535, "frac_reward_zero_std": 0.75, "grad_norm": 0.39453125, "learning_rate": 8.684807256235828e-07, "loss": -0.0015, "num_tokens": 13866754.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 233, "step_time": 67.96496311109513 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1691.0, "completions/max_terminated_length": 1691.0, "completions/mean_length": 722.4375, "completions/mean_terminated_length": 722.4375, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "entropy": 0.9679892286658287, "epoch": 0.1326530612244898, "frac_reward_zero_std": 0.75, "grad_norm": 0.423828125, "learning_rate": 8.679138321995464e-07, "loss": 0.0132, "num_tokens": 13924182.0, "reward": 0.0859375, "reward_std": 0.36247265338897705, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 234, "step_time": 64.32111782580614 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1209.0, "completions/max_terminated_length": 1209.0, "completions/mean_length": 697.359375, "completions/mean_terminated_length": 697.359375, "completions/min_length": 313.0, "completions/min_terminated_length": 313.0, "entropy": 0.8101966269314289, "epoch": 0.13321995464852607, "frac_reward_zero_std": 0.625, "grad_norm": 0.515625, "learning_rate": 8.673469387755102e-07, "loss": 0.036, "num_tokens": 13983613.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 235, "step_time": 46.66053644847125 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1462.0, "completions/max_terminated_length": 1462.0, "completions/mean_length": 848.953125, "completions/mean_terminated_length": 848.953125, "completions/min_length": 352.0, "completions/min_terminated_length": 352.0, "entropy": 0.745408920571208, "epoch": 0.13378684807256236, "frac_reward_zero_std": 0.875, "grad_norm": 0.197265625, "learning_rate": 8.667800453514738e-07, "loss": -0.0025, "num_tokens": 14052034.0, "reward": 0.0234375, "reward_std": 0.10652101039886475, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 236, "step_time": 55.998151410371065 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 670.515625, "completions/mean_terminated_length": 670.515625, "completions/min_length": 297.0, "completions/min_terminated_length": 297.0, "entropy": 0.8511352762579918, "epoch": 0.13435374149659865, "frac_reward_zero_std": 0.5, "grad_norm": 0.55078125, "learning_rate": 8.662131519274376e-07, "loss": 0.0455, "num_tokens": 14105475.0, "reward": 0.1484375, "reward_std": 0.49344661831855774, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 237, "step_time": 54.208372516557574 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2160.0, "completions/max_terminated_length": 2160.0, "completions/mean_length": 886.609375, "completions/mean_terminated_length": 886.609375, "completions/min_length": 306.0, "completions/min_terminated_length": 306.0, "entropy": 0.6885136272758245, "epoch": 0.1349206349206349, "frac_reward_zero_std": 0.5, "grad_norm": 0.45703125, "learning_rate": 8.656462585034013e-07, "loss": -0.0186, "num_tokens": 14172330.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 238, "step_time": 82.4482323108241 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1512.0, "completions/max_terminated_length": 1512.0, "completions/mean_length": 776.484375, "completions/mean_terminated_length": 776.484375, "completions/min_length": 336.0, "completions/min_terminated_length": 336.0, "entropy": 0.8836219422519207, "epoch": 0.1354875283446712, "frac_reward_zero_std": 0.5, "grad_norm": 0.498046875, "learning_rate": 8.650793650793651e-07, "loss": -0.0082, "num_tokens": 14231937.0, "reward": 0.15625, "reward_std": 0.49501484632492065, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 239, "step_time": 56.863407094962895 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 772.421875, "completions/mean_terminated_length": 772.421875, "completions/min_length": 275.0, "completions/min_terminated_length": 275.0, "entropy": 0.8125546742230654, "epoch": 0.1360544217687075, "frac_reward_zero_std": 0.75, "grad_norm": 0.341796875, "learning_rate": 8.645124716553288e-07, "loss": 0.0253, "num_tokens": 14291156.0, "reward": 0.09375, "reward_std": 0.426084041595459, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0, "rewards/format_reward_func/std": 0.0, "step": 240, "step_time": 54.15421236958355 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1214.0, "completions/max_terminated_length": 1214.0, "completions/mean_length": 717.5, "completions/mean_terminated_length": 717.5, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "entropy": 0.7670545168220997, "epoch": 0.13662131519274376, "frac_reward_zero_std": 0.5, "grad_norm": 0.55078125, "learning_rate": 8.639455782312925e-07, "loss": 0.0004, "num_tokens": 14347132.0, "reward": 0.2265625, "reward_std": 0.5904644727706909, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 241, "step_time": 45.79305484518409 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1725.0, "completions/max_terminated_length": 1725.0, "completions/mean_length": 817.71875, "completions/mean_terminated_length": 817.71875, "completions/min_length": 408.0, "completions/min_terminated_length": 408.0, "entropy": 0.7369464822113514, "epoch": 0.13718820861678005, "frac_reward_zero_std": 0.75, "grad_norm": 0.359375, "learning_rate": 8.633786848072562e-07, "loss": -0.0089, "num_tokens": 14409370.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 242, "step_time": 64.96719512436539 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1240.0, "completions/max_terminated_length": 1240.0, "completions/mean_length": 643.890625, "completions/mean_terminated_length": 643.890625, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "entropy": 0.760059030726552, "epoch": 0.1377551020408163, "frac_reward_zero_std": 0.75, "grad_norm": 0.431640625, "learning_rate": 8.6281179138322e-07, "loss": 0.0022, "num_tokens": 14461675.0, "reward": 0.15625, "reward_std": 0.49501484632492065, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 243, "step_time": 47.31172960717231 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 779.359375, "completions/mean_terminated_length": 779.359375, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "entropy": 0.711810328066349, "epoch": 0.1383219954648526, "frac_reward_zero_std": 0.625, "grad_norm": 0.46484375, "learning_rate": 8.622448979591836e-07, "loss": -0.0117, "num_tokens": 14521482.0, "reward": 0.15625, "reward_std": 0.5261042714118958, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 244, "step_time": 49.666923959739506 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1311.0, "completions/max_terminated_length": 1311.0, "completions/mean_length": 749.359375, "completions/mean_terminated_length": 749.359375, "completions/min_length": 363.0, "completions/min_terminated_length": 363.0, "entropy": 0.8499616701155901, "epoch": 0.1388888888888889, "frac_reward_zero_std": 0.375, "grad_norm": 0.66796875, "learning_rate": 8.616780045351474e-07, "loss": 0.0022, "num_tokens": 14582617.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 245, "step_time": 52.20430337637663 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1238.0, "completions/max_terminated_length": 1238.0, "completions/mean_length": 779.84375, "completions/mean_terminated_length": 779.84375, "completions/min_length": 388.0, "completions/min_terminated_length": 388.0, "entropy": 0.8445747178047895, "epoch": 0.13945578231292516, "frac_reward_zero_std": 0.75, "grad_norm": 0.361328125, "learning_rate": 8.611111111111111e-07, "loss": 0.0172, "num_tokens": 14644599.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 246, "step_time": 47.973739959299564 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1796.0, "completions/max_terminated_length": 1796.0, "completions/mean_length": 858.078125, "completions/mean_terminated_length": 858.078125, "completions/min_length": 353.0, "completions/min_terminated_length": 353.0, "entropy": 0.7805829998105764, "epoch": 0.14002267573696145, "frac_reward_zero_std": 0.75, "grad_norm": 0.341796875, "learning_rate": 8.605442176870747e-07, "loss": -0.0442, "num_tokens": 14714260.0, "reward": 0.09375, "reward_std": 0.40703144669532776, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 247, "step_time": 71.51869424339384 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1315.0, "completions/max_terminated_length": 1315.0, "completions/mean_length": 689.59375, "completions/mean_terminated_length": 689.59375, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "entropy": 0.6935113929212093, "epoch": 0.14058956916099774, "frac_reward_zero_std": 0.625, "grad_norm": 0.44921875, "learning_rate": 8.599773242630385e-07, "loss": -0.0056, "num_tokens": 14767354.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 248, "step_time": 50.06727380864322 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1367.0, "completions/max_terminated_length": 1367.0, "completions/mean_length": 817.921875, "completions/mean_terminated_length": 817.921875, "completions/min_length": 359.0, "completions/min_terminated_length": 359.0, "entropy": 0.8837875053286552, "epoch": 0.141156462585034, "frac_reward_zero_std": 0.5, "grad_norm": 0.51953125, "learning_rate": 8.594104308390022e-07, "loss": 0.028, "num_tokens": 14831589.0, "reward": 0.2890625, "reward_std": 0.6655030846595764, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 249, "step_time": 52.77266918309033 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1598.0, "completions/max_terminated_length": 1598.0, "completions/mean_length": 743.78125, "completions/mean_terminated_length": 743.78125, "completions/min_length": 227.0, "completions/min_terminated_length": 227.0, "entropy": 0.8470449112355709, "epoch": 0.1417233560090703, "frac_reward_zero_std": 0.5, "grad_norm": 0.54296875, "learning_rate": 8.58843537414966e-07, "loss": 0.0148, "num_tokens": 14889359.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 250, "step_time": 60.69317118264735 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 639.578125, "completions/mean_terminated_length": 639.578125, "completions/min_length": 336.0, "completions/min_terminated_length": 336.0, "entropy": 1.1404040716588497, "epoch": 0.14229024943310659, "frac_reward_zero_std": 0.625, "grad_norm": 0.5703125, "learning_rate": 8.582766439909297e-07, "loss": 0.0229, "num_tokens": 14941284.0, "reward": 0.21875, "reward_std": 0.5900968313217163, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 251, "step_time": 52.66949089244008 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1634.0, "completions/max_terminated_length": 1634.0, "completions/mean_length": 704.6875, "completions/mean_terminated_length": 704.6875, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "entropy": 0.7078406568616629, "epoch": 0.14285714285714285, "frac_reward_zero_std": 0.375, "grad_norm": 0.66015625, "learning_rate": 8.577097505668934e-07, "loss": 0.0486, "num_tokens": 14997264.0, "reward": 0.2734375, "reward_std": 0.6780555844306946, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 252, "step_time": 61.867045666091144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1101.0, "completions/max_terminated_length": 1101.0, "completions/mean_length": 730.359375, "completions/mean_terminated_length": 730.359375, "completions/min_length": 419.0, "completions/min_terminated_length": 419.0, "entropy": 0.9553987365216017, "epoch": 0.14342403628117914, "frac_reward_zero_std": 0.5, "grad_norm": 0.63671875, "learning_rate": 8.57142857142857e-07, "loss": 0.0092, "num_tokens": 15053551.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 253, "step_time": 42.036332870833576 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1391.0, "completions/max_terminated_length": 1391.0, "completions/mean_length": 762.953125, "completions/mean_terminated_length": 762.953125, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "entropy": 0.7707686368376017, "epoch": 0.14399092970521543, "frac_reward_zero_std": 0.625, "grad_norm": 0.4296875, "learning_rate": 8.565759637188208e-07, "loss": 0.013, "num_tokens": 15113372.0, "reward": 0.1171875, "reward_std": 0.46922925114631653, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 254, "step_time": 52.543017198331654 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2076.0, "completions/max_terminated_length": 2076.0, "completions/mean_length": 797.25, "completions/mean_terminated_length": 797.25, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "entropy": 0.7859501540660858, "epoch": 0.1445578231292517, "frac_reward_zero_std": 0.625, "grad_norm": 0.45703125, "learning_rate": 8.560090702947845e-07, "loss": 0.0319, "num_tokens": 15174468.0, "reward": 0.15625, "reward_std": 0.49501484632492065, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 255, "step_time": 79.07321002520621 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 712.0625, "completions/mean_terminated_length": 712.0625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "entropy": 0.9650670122355223, "epoch": 0.14512471655328799, "frac_reward_zero_std": 0.5, "grad_norm": 0.55859375, "learning_rate": 8.554421768707483e-07, "loss": 0.0032, "num_tokens": 15232792.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 256, "step_time": 44.803927617147565 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1194.0, "completions/max_terminated_length": 1194.0, "completions/mean_length": 703.984375, "completions/mean_terminated_length": 703.984375, "completions/min_length": 324.0, "completions/min_terminated_length": 324.0, "entropy": 0.9428023956716061, "epoch": 0.14569160997732428, "frac_reward_zero_std": 0.5, "grad_norm": 0.625, "learning_rate": 8.54875283446712e-07, "loss": 0.0024, "num_tokens": 15290367.0, "reward": 0.1875, "reward_std": 0.5455447435379028, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 257, "step_time": 46.41874376870692 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 664.9375, "completions/mean_terminated_length": 664.9375, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "entropy": 0.8272994346916676, "epoch": 0.14625850340136054, "frac_reward_zero_std": 0.75, "grad_norm": 0.40234375, "learning_rate": 8.543083900226758e-07, "loss": -0.0044, "num_tokens": 15341947.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 258, "step_time": 38.47307618521154 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 771.171875, "completions/mean_terminated_length": 771.171875, "completions/min_length": 393.0, "completions/min_terminated_length": 393.0, "entropy": 0.9931114763021469, "epoch": 0.14682539682539683, "frac_reward_zero_std": 0.75, "grad_norm": 0.357421875, "learning_rate": 8.537414965986394e-07, "loss": -0.0203, "num_tokens": 15401766.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 259, "step_time": 56.232815210707486 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1481.0, "completions/max_terminated_length": 1481.0, "completions/mean_length": 723.03125, "completions/mean_terminated_length": 723.03125, "completions/min_length": 391.0, "completions/min_terminated_length": 391.0, "entropy": 0.8141548484563828, "epoch": 0.1473922902494331, "frac_reward_zero_std": 0.625, "grad_norm": 0.44921875, "learning_rate": 8.531746031746031e-07, "loss": 0.0038, "num_tokens": 15457280.0, "reward": 0.3046875, "reward_std": 0.710561990737915, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 260, "step_time": 56.52870470285416 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 658.359375, "completions/mean_terminated_length": 658.359375, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "entropy": 0.6927868071943521, "epoch": 0.14795918367346939, "frac_reward_zero_std": 0.25, "grad_norm": 0.65234375, "learning_rate": 8.526077097505668e-07, "loss": 0.0249, "num_tokens": 15510295.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 261, "step_time": 56.96211295481771 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1513.0, "completions/max_terminated_length": 1513.0, "completions/mean_length": 752.953125, "completions/mean_terminated_length": 752.953125, "completions/min_length": 398.0, "completions/min_terminated_length": 398.0, "entropy": 0.9332785494625568, "epoch": 0.14852607709750568, "frac_reward_zero_std": 0.75, "grad_norm": 0.3671875, "learning_rate": 8.520408163265306e-07, "loss": 0.0098, "num_tokens": 15569868.0, "reward": 0.1015625, "reward_std": 0.3692515790462494, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 262, "step_time": 59.106632246635854 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1880.0, "completions/max_terminated_length": 1880.0, "completions/mean_length": 839.875, "completions/mean_terminated_length": 839.875, "completions/min_length": 340.0, "completions/min_terminated_length": 340.0, "entropy": 0.682667376473546, "epoch": 0.14909297052154194, "frac_reward_zero_std": 0.75, "grad_norm": 0.314453125, "learning_rate": 8.514739229024943e-07, "loss": 0.0144, "num_tokens": 15634316.0, "reward": 0.125, "reward_std": 0.4364357888698578, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 263, "step_time": 71.25024418812245 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1257.0, "completions/max_terminated_length": 1257.0, "completions/mean_length": 722.6875, "completions/mean_terminated_length": 722.6875, "completions/min_length": 391.0, "completions/min_terminated_length": 391.0, "entropy": 0.9179340749979019, "epoch": 0.14965986394557823, "frac_reward_zero_std": 0.625, "grad_norm": 0.55859375, "learning_rate": 8.509070294784581e-07, "loss": 0.018, "num_tokens": 15690792.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 264, "step_time": 47.271880991756916 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1270.0, "completions/max_terminated_length": 1270.0, "completions/mean_length": 637.296875, "completions/mean_terminated_length": 637.296875, "completions/min_length": 277.0, "completions/min_terminated_length": 277.0, "entropy": 1.086241863667965, "epoch": 0.15022675736961452, "frac_reward_zero_std": 0.625, "grad_norm": 0.48828125, "learning_rate": 8.503401360544217e-07, "loss": 0.0132, "num_tokens": 15742243.0, "reward": 0.1875, "reward_std": 0.559903621673584, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 265, "step_time": 47.730507844127715 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1571.0, "completions/max_terminated_length": 1571.0, "completions/mean_length": 706.125, "completions/mean_terminated_length": 706.125, "completions/min_length": 371.0, "completions/min_terminated_length": 371.0, "entropy": 0.9066045880317688, "epoch": 0.15079365079365079, "frac_reward_zero_std": 0.75, "grad_norm": 0.333984375, "learning_rate": 8.497732426303855e-07, "loss": 0.0451, "num_tokens": 15796715.0, "reward": 0.0546875, "reward_std": 0.26899558305740356, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 266, "step_time": 59.063225102610886 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1388.0, "completions/max_terminated_length": 1388.0, "completions/mean_length": 697.734375, "completions/mean_terminated_length": 697.734375, "completions/min_length": 224.0, "completions/min_terminated_length": 224.0, "entropy": 0.8448955230414867, "epoch": 0.15136054421768708, "frac_reward_zero_std": 0.5, "grad_norm": 0.48046875, "learning_rate": 8.492063492063492e-07, "loss": 0.0121, "num_tokens": 15852362.0, "reward": 0.21875, "reward_std": 0.603396475315094, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 267, "step_time": 52.62285229656845 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1680.0, "completions/max_terminated_length": 1680.0, "completions/mean_length": 696.25, "completions/mean_terminated_length": 696.25, "completions/min_length": 143.0, "completions/min_terminated_length": 143.0, "entropy": 0.7814523428678513, "epoch": 0.15192743764172337, "frac_reward_zero_std": 0.75, "grad_norm": 0.400390625, "learning_rate": 8.486394557823128e-07, "loss": 0.0395, "num_tokens": 15910290.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 268, "step_time": 64.97410681564361 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1367.0, "completions/max_terminated_length": 1367.0, "completions/mean_length": 616.078125, "completions/mean_terminated_length": 616.078125, "completions/min_length": 285.0, "completions/min_terminated_length": 285.0, "entropy": 0.8676711972802877, "epoch": 0.15249433106575963, "frac_reward_zero_std": 0.375, "grad_norm": 0.75390625, "learning_rate": 8.480725623582766e-07, "loss": 0.007, "num_tokens": 15961415.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 269, "step_time": 52.289122426882386 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1256.0, "completions/max_terminated_length": 1256.0, "completions/mean_length": 735.8125, "completions/mean_terminated_length": 735.8125, "completions/min_length": 377.0, "completions/min_terminated_length": 377.0, "entropy": 0.8315059375017881, "epoch": 0.15306122448979592, "frac_reward_zero_std": 0.625, "grad_norm": 0.42578125, "learning_rate": 8.475056689342403e-07, "loss": -0.0034, "num_tokens": 16019531.0, "reward": 0.265625, "reward_std": 0.6544642448425293, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 270, "step_time": 48.43833992630243 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2195.0, "completions/max_terminated_length": 2195.0, "completions/mean_length": 753.65625, "completions/mean_terminated_length": 753.65625, "completions/min_length": 7.0, "completions/min_terminated_length": 7.0, "entropy": 1.0189075618982315, "epoch": 0.1536281179138322, "frac_reward_zero_std": 0.625, "grad_norm": 0.49609375, "learning_rate": 8.469387755102041e-07, "loss": 0.0254, "num_tokens": 16078069.0, "reward": 0.1484375, "reward_std": 0.49344661831855774, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 271, "step_time": 84.29482378717512 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1597.0, "completions/max_terminated_length": 1597.0, "completions/mean_length": 761.625, "completions/mean_terminated_length": 761.625, "completions/min_length": 356.0, "completions/min_terminated_length": 356.0, "entropy": 0.9707346335053444, "epoch": 0.15419501133786848, "frac_reward_zero_std": 0.75, "grad_norm": 0.341796875, "learning_rate": 8.463718820861677e-07, "loss": 0.0053, "num_tokens": 16137533.0, "reward": 0.2421875, "reward_std": 0.6900206208229065, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 272, "step_time": 60.21038067340851 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2329.0, "completions/max_terminated_length": 2329.0, "completions/mean_length": 791.125, "completions/mean_terminated_length": 791.125, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "entropy": 0.7534265369176865, "epoch": 0.15476190476190477, "frac_reward_zero_std": 0.375, "grad_norm": 0.578125, "learning_rate": 8.458049886621315e-07, "loss": 0.0448, "num_tokens": 16198797.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 273, "step_time": 89.73313437681645 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1502.0, "completions/max_terminated_length": 1502.0, "completions/mean_length": 731.515625, "completions/mean_terminated_length": 731.515625, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "entropy": 0.7331438101828098, "epoch": 0.15532879818594103, "frac_reward_zero_std": 0.75, "grad_norm": 0.427734375, "learning_rate": 8.452380952380952e-07, "loss": -0.0193, "num_tokens": 16255470.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 274, "step_time": 56.693840723484755 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1401.0, "completions/max_terminated_length": 1401.0, "completions/mean_length": 697.0625, "completions/mean_terminated_length": 697.0625, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "entropy": 0.6416508182883263, "epoch": 0.15589569160997732, "frac_reward_zero_std": 0.375, "grad_norm": 0.578125, "learning_rate": 8.44671201814059e-07, "loss": 0.0094, "num_tokens": 16309378.0, "reward": 0.3125, "reward_std": 0.7210001945495605, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 275, "step_time": 52.967795793898404 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1338.0, "completions/mean_length": 856.109375, "completions/mean_terminated_length": 804.6825561523438, "completions/min_length": 433.0, "completions/min_terminated_length": 433.0, "entropy": 0.6948063969612122, "epoch": 0.1564625850340136, "frac_reward_zero_std": 0.625, "grad_norm": 0.37109375, "learning_rate": 8.441043083900226e-07, "loss": 0.0166, "num_tokens": 16375169.0, "reward": 0.2265625, "reward_std": 0.6167611479759216, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 276, "step_time": 177.92747872509062 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1217.0, "completions/max_terminated_length": 1217.0, "completions/mean_length": 584.125, "completions/mean_terminated_length": 584.125, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "entropy": 0.6641750391572714, "epoch": 0.15702947845804988, "frac_reward_zero_std": 0.375, "grad_norm": 0.6953125, "learning_rate": 8.435374149659864e-07, "loss": 0.0061, "num_tokens": 16423729.0, "reward": 0.3984375, "reward_std": 0.7724404335021973, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 277, "step_time": 46.15311232954264 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1388.0, "completions/max_terminated_length": 1388.0, "completions/mean_length": 868.453125, "completions/mean_terminated_length": 868.453125, "completions/min_length": 425.0, "completions/min_terminated_length": 425.0, "entropy": 0.7224074248224497, "epoch": 0.15759637188208617, "frac_reward_zero_std": 0.75, "grad_norm": 0.322265625, "learning_rate": 8.4297052154195e-07, "loss": -0.0069, "num_tokens": 16491150.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 278, "step_time": 52.73080511763692 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 707.671875, "completions/mean_terminated_length": 707.671875, "completions/min_length": 278.0, "completions/min_terminated_length": 278.0, "entropy": 0.8151501063257456, "epoch": 0.15816326530612246, "frac_reward_zero_std": 0.75, "grad_norm": 0.353515625, "learning_rate": 8.424036281179138e-07, "loss": -0.0161, "num_tokens": 16546033.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 279, "step_time": 44.939537251368165 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2440.0, "completions/max_terminated_length": 2440.0, "completions/mean_length": 784.25, "completions/mean_terminated_length": 784.25, "completions/min_length": 351.0, "completions/min_terminated_length": 351.0, "entropy": 1.0181800108402967, "epoch": 0.15873015873015872, "frac_reward_zero_std": 0.625, "grad_norm": 0.484375, "learning_rate": 8.418367346938775e-07, "loss": 0.0263, "num_tokens": 16605617.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 280, "step_time": 94.32256285753101 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1421.0, "completions/max_terminated_length": 1421.0, "completions/mean_length": 723.46875, "completions/mean_terminated_length": 723.46875, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "entropy": 0.6927716005593538, "epoch": 0.159297052154195, "frac_reward_zero_std": 0.5, "grad_norm": 0.57421875, "learning_rate": 8.412698412698413e-07, "loss": -0.0006, "num_tokens": 16662055.0, "reward": 0.265625, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 281, "step_time": 54.639340850524604 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1401.0, "completions/max_terminated_length": 1401.0, "completions/mean_length": 784.578125, "completions/mean_terminated_length": 784.578125, "completions/min_length": 376.0, "completions/min_terminated_length": 376.0, "entropy": 0.7328203655779362, "epoch": 0.1598639455782313, "frac_reward_zero_std": 0.75, "grad_norm": 0.384765625, "learning_rate": 8.40702947845805e-07, "loss": 0.0139, "num_tokens": 16722388.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 282, "step_time": 53.27179315127432 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1107.0, "completions/mean_length": 722.609375, "completions/mean_terminated_length": 669.0635375976562, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "entropy": 0.7031647693365812, "epoch": 0.16043083900226757, "frac_reward_zero_std": 0.375, "grad_norm": 0.51953125, "learning_rate": 8.401360544217688e-07, "loss": 0.0005, "num_tokens": 16780995.0, "reward": 0.28125, "reward_std": 0.6777355074882507, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 283, "step_time": 179.04967124108225 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1520.0, "completions/max_terminated_length": 1520.0, "completions/mean_length": 677.578125, "completions/mean_terminated_length": 677.578125, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "entropy": 0.9004981126636267, "epoch": 0.16099773242630386, "frac_reward_zero_std": 0.625, "grad_norm": 0.470703125, "learning_rate": 8.395691609977323e-07, "loss": 0.0095, "num_tokens": 16834080.0, "reward": 0.2734375, "reward_std": 0.6780555844306946, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 284, "step_time": 57.85565869230777 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1098.0, "completions/max_terminated_length": 1098.0, "completions/mean_length": 682.59375, "completions/mean_terminated_length": 682.59375, "completions/min_length": 326.0, "completions/min_terminated_length": 326.0, "entropy": 0.8715041428804398, "epoch": 0.16156462585034015, "frac_reward_zero_std": 0.25, "grad_norm": 0.67578125, "learning_rate": 8.39002267573696e-07, "loss": 0.0372, "num_tokens": 16887894.0, "reward": 0.3203125, "reward_std": 0.6978833675384521, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 285, "step_time": 43.965926623903215 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 683.46875, "completions/mean_terminated_length": 683.46875, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "entropy": 1.0438571982085705, "epoch": 0.1621315192743764, "frac_reward_zero_std": 0.625, "grad_norm": 0.50390625, "learning_rate": 8.384353741496598e-07, "loss": 0.003, "num_tokens": 16942412.0, "reward": 0.328125, "reward_std": 0.6970388293266296, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 286, "step_time": 50.78270637895912 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2196.0, "completions/max_terminated_length": 2196.0, "completions/mean_length": 741.359375, "completions/mean_terminated_length": 741.359375, "completions/min_length": 374.0, "completions/min_terminated_length": 374.0, "entropy": 0.883393669500947, "epoch": 0.1626984126984127, "frac_reward_zero_std": 0.625, "grad_norm": 0.45703125, "learning_rate": 8.378684807256235e-07, "loss": 0.0164, "num_tokens": 17000619.0, "reward": 0.15625, "reward_std": 0.5261042714118958, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 287, "step_time": 79.346072871238 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1470.0, "completions/max_terminated_length": 1470.0, "completions/mean_length": 860.609375, "completions/mean_terminated_length": 860.609375, "completions/min_length": 327.0, "completions/min_terminated_length": 327.0, "entropy": 0.7544728927314281, "epoch": 0.16326530612244897, "frac_reward_zero_std": 0.875, "grad_norm": 0.267578125, "learning_rate": 8.373015873015873e-07, "loss": -0.0007, "num_tokens": 17067378.0, "reward": 0.0546875, "reward_std": 0.15728822350502014, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 288, "step_time": 52.59334731567651 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1235.0, "completions/max_terminated_length": 1235.0, "completions/mean_length": 705.0, "completions/mean_terminated_length": 705.0, "completions/min_length": 259.0, "completions/min_terminated_length": 259.0, "entropy": 0.8788765668869019, "epoch": 0.16383219954648526, "frac_reward_zero_std": 0.875, "grad_norm": 0.279296875, "learning_rate": 8.36734693877551e-07, "loss": -0.0093, "num_tokens": 17123122.0, "reward": 0.0390625, "reward_std": 0.13524484634399414, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 289, "step_time": 43.87046900205314 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1443.0, "completions/max_terminated_length": 1443.0, "completions/mean_length": 705.03125, "completions/mean_terminated_length": 705.03125, "completions/min_length": 355.0, "completions/min_terminated_length": 355.0, "entropy": 0.9619238749146461, "epoch": 0.16439909297052155, "frac_reward_zero_std": 0.625, "grad_norm": 0.56640625, "learning_rate": 8.361678004535147e-07, "loss": -0.0029, "num_tokens": 17178516.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 290, "step_time": 50.73440036177635 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1172.0, "completions/max_terminated_length": 1172.0, "completions/mean_length": 677.921875, "completions/mean_terminated_length": 677.921875, "completions/min_length": 366.0, "completions/min_terminated_length": 366.0, "entropy": 0.9480736441910267, "epoch": 0.1649659863945578, "frac_reward_zero_std": 0.75, "grad_norm": 0.373046875, "learning_rate": 8.356009070294783e-07, "loss": -0.0041, "num_tokens": 17233575.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 291, "step_time": 41.97542067337781 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 695.1875, "completions/mean_terminated_length": 695.1875, "completions/min_length": 310.0, "completions/min_terminated_length": 310.0, "entropy": 1.0936996825039387, "epoch": 0.1655328798185941, "frac_reward_zero_std": 0.875, "grad_norm": 0.1845703125, "learning_rate": 8.350340136054421e-07, "loss": 0.0019, "num_tokens": 17292555.0, "reward": 0.0859375, "reward_std": 0.3398725986480713, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 292, "step_time": 64.84023329243064 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1295.0, "completions/max_terminated_length": 1295.0, "completions/mean_length": 668.34375, "completions/mean_terminated_length": 668.34375, "completions/min_length": 367.0, "completions/min_terminated_length": 367.0, "entropy": 0.8729091212153435, "epoch": 0.1660997732426304, "frac_reward_zero_std": 0.625, "grad_norm": 0.474609375, "learning_rate": 8.344671201814058e-07, "loss": -0.0011, "num_tokens": 17346385.0, "reward": 0.1015625, "reward_std": 0.3692515790462494, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 293, "step_time": 46.00880840513855 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1544.0, "completions/max_terminated_length": 1544.0, "completions/mean_length": 830.59375, "completions/mean_terminated_length": 830.59375, "completions/min_length": 405.0, "completions/min_terminated_length": 405.0, "entropy": 0.8223886843770742, "epoch": 0.16666666666666666, "frac_reward_zero_std": 0.5, "grad_norm": 0.443359375, "learning_rate": 8.339002267573696e-07, "loss": -0.0125, "num_tokens": 17412391.0, "reward": 0.203125, "reward_std": 0.6023679971694946, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 294, "step_time": 55.6619848748669 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 3081.0, "completions/mean_length": 821.265625, "completions/mean_terminated_length": 769.2857666015625, "completions/min_length": 359.0, "completions/min_terminated_length": 359.0, "entropy": 0.7634497378021479, "epoch": 0.16723356009070295, "frac_reward_zero_std": 0.375, "grad_norm": 0.5078125, "learning_rate": 8.333333333333333e-07, "loss": 0.0086, "num_tokens": 17475888.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 295, "step_time": 172.29678527265787 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1131.0, "completions/max_terminated_length": 1131.0, "completions/mean_length": 713.515625, "completions/mean_terminated_length": 713.515625, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "entropy": 0.7846761327236891, "epoch": 0.16780045351473924, "frac_reward_zero_std": 0.25, "grad_norm": 0.6640625, "learning_rate": 8.327664399092971e-07, "loss": -0.002, "num_tokens": 17531785.0, "reward": 0.3515625, "reward_std": 0.7274640202522278, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 296, "step_time": 40.278807829134166 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1332.0, "completions/max_terminated_length": 1332.0, "completions/mean_length": 781.171875, "completions/mean_terminated_length": 781.171875, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "entropy": 0.6870723441243172, "epoch": 0.1683673469387755, "frac_reward_zero_std": 0.375, "grad_norm": 0.51953125, "learning_rate": 8.321995464852607e-07, "loss": 0.0432, "num_tokens": 17592324.0, "reward": 0.2109375, "reward_std": 0.5896238088607788, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 297, "step_time": 47.24058729223907 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2147.0, "completions/max_terminated_length": 2147.0, "completions/mean_length": 755.96875, "completions/mean_terminated_length": 755.96875, "completions/min_length": 393.0, "completions/min_terminated_length": 393.0, "entropy": 0.8424653597176075, "epoch": 0.1689342403628118, "frac_reward_zero_std": 0.75, "grad_norm": 0.390625, "learning_rate": 8.316326530612245e-07, "loss": 0.0032, "num_tokens": 17650874.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 298, "step_time": 77.23045068792999 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1169.0, "completions/max_terminated_length": 1169.0, "completions/mean_length": 690.859375, "completions/mean_terminated_length": 690.859375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "entropy": 0.9548549987375736, "epoch": 0.16950113378684808, "frac_reward_zero_std": 0.625, "grad_norm": 0.52734375, "learning_rate": 8.310657596371881e-07, "loss": -0.0076, "num_tokens": 17706529.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 299, "step_time": 42.4382833102718 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1243.0, "completions/max_terminated_length": 1243.0, "completions/mean_length": 723.296875, "completions/mean_terminated_length": 723.296875, "completions/min_length": 261.0, "completions/min_terminated_length": 261.0, "entropy": 0.8589233346283436, "epoch": 0.17006802721088435, "frac_reward_zero_std": 0.75, "grad_norm": 0.3984375, "learning_rate": 8.304988662131519e-07, "loss": 0.0106, "num_tokens": 17764780.0, "reward": 0.0625, "reward_std": 0.2745848298072815, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 300, "step_time": 44.784306516870856 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1137.0, "completions/max_terminated_length": 1137.0, "completions/mean_length": 697.21875, "completions/mean_terminated_length": 697.21875, "completions/min_length": 242.0, "completions/min_terminated_length": 242.0, "entropy": 0.9121045172214508, "epoch": 0.17063492063492064, "frac_reward_zero_std": 0.5, "grad_norm": 0.578125, "learning_rate": 8.299319727891156e-07, "loss": -0.0034, "num_tokens": 17820530.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 301, "step_time": 40.97146006766707 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1421.0, "completions/max_terminated_length": 1421.0, "completions/mean_length": 669.296875, "completions/mean_terminated_length": 669.296875, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "entropy": 0.7469332460314035, "epoch": 0.1712018140589569, "frac_reward_zero_std": 0.5, "grad_norm": 0.5390625, "learning_rate": 8.293650793650794e-07, "loss": -0.0241, "num_tokens": 17874285.0, "reward": 0.2265625, "reward_std": 0.5033986568450928, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.1015625, "rewards/format_reward_func/std": 0.20275264978408813, "step": 302, "step_time": 50.62312275916338 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1442.0, "completions/max_terminated_length": 1442.0, "completions/mean_length": 736.5, "completions/mean_terminated_length": 736.5, "completions/min_length": 292.0, "completions/min_terminated_length": 292.0, "entropy": 0.8989239111542702, "epoch": 0.1717687074829932, "frac_reward_zero_std": 0.625, "grad_norm": 0.44921875, "learning_rate": 8.28798185941043e-07, "loss": 0.0223, "num_tokens": 17931741.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 303, "step_time": 51.81090355012566 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1204.0, "completions/max_terminated_length": 1204.0, "completions/mean_length": 676.96875, "completions/mean_terminated_length": 676.96875, "completions/min_length": 298.0, "completions/min_terminated_length": 298.0, "entropy": 0.9559312872588634, "epoch": 0.17233560090702948, "frac_reward_zero_std": 0.625, "grad_norm": 0.54296875, "learning_rate": 8.282312925170068e-07, "loss": -0.0022, "num_tokens": 17983251.0, "reward": 0.140625, "reward_std": 0.4917473793029785, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 304, "step_time": 42.032407058402896 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 815.46875, "completions/mean_terminated_length": 815.46875, "completions/min_length": 204.0, "completions/min_terminated_length": 204.0, "entropy": 0.8452925551682711, "epoch": 0.17290249433106575, "frac_reward_zero_std": 0.5, "grad_norm": 0.498046875, "learning_rate": 8.276643990929705e-07, "loss": 0.0129, "num_tokens": 18046369.0, "reward": 0.21875, "reward_std": 0.5900968313217163, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 305, "step_time": 51.47114187292755 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 743.65625, "completions/mean_terminated_length": 743.65625, "completions/min_length": 285.0, "completions/min_terminated_length": 285.0, "entropy": 0.7419566698372364, "epoch": 0.17346938775510204, "frac_reward_zero_std": 0.625, "grad_norm": 0.416015625, "learning_rate": 8.270975056689343e-07, "loss": 0.0102, "num_tokens": 18106523.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 306, "step_time": 44.13747341185808 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1436.0, "completions/max_terminated_length": 1436.0, "completions/mean_length": 771.5625, "completions/mean_terminated_length": 771.5625, "completions/min_length": 435.0, "completions/min_terminated_length": 435.0, "entropy": 0.9274702072143555, "epoch": 0.17403628117913833, "frac_reward_zero_std": 0.625, "grad_norm": 0.53125, "learning_rate": 8.26530612244898e-07, "loss": -0.0092, "num_tokens": 18166095.0, "reward": 0.2265625, "reward_std": 0.5904644727706909, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 307, "step_time": 51.33585764002055 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1881.0, "completions/max_terminated_length": 1881.0, "completions/mean_length": 726.3125, "completions/mean_terminated_length": 726.3125, "completions/min_length": 323.0, "completions/min_terminated_length": 323.0, "entropy": 0.7567149810492992, "epoch": 0.1746031746031746, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 8.259637188208616e-07, "loss": 0.0355, "num_tokens": 18229043.0, "reward": 0.3203125, "reward_std": 0.6978833675384521, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 308, "step_time": 69.37814402300864 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1125.0, "completions/max_terminated_length": 1125.0, "completions/mean_length": 691.328125, "completions/mean_terminated_length": 691.328125, "completions/min_length": 220.0, "completions/min_terminated_length": 220.0, "entropy": 0.820183027535677, "epoch": 0.17517006802721088, "frac_reward_zero_std": 0.625, "grad_norm": 0.515625, "learning_rate": 8.253968253968253e-07, "loss": 0.0132, "num_tokens": 18284560.0, "reward": 0.21875, "reward_std": 0.6416434645652771, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 309, "step_time": 39.77160437498242 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1591.0, "completions/max_terminated_length": 1591.0, "completions/mean_length": 733.546875, "completions/mean_terminated_length": 733.546875, "completions/min_length": 297.0, "completions/min_terminated_length": 297.0, "entropy": 0.7642414346337318, "epoch": 0.17573696145124718, "frac_reward_zero_std": 0.625, "grad_norm": 0.439453125, "learning_rate": 8.24829931972789e-07, "loss": 0.0167, "num_tokens": 18341827.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 310, "step_time": 56.358331645838916 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1484.0, "completions/max_terminated_length": 1484.0, "completions/mean_length": 762.4375, "completions/mean_terminated_length": 762.4375, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "entropy": 0.8820738904178143, "epoch": 0.17630385487528344, "frac_reward_zero_std": 0.5, "grad_norm": 0.53125, "learning_rate": 8.242630385487528e-07, "loss": 0.0081, "num_tokens": 18400719.0, "reward": 0.25, "reward_std": 0.6299408078193665, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 311, "step_time": 51.908669171854854 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1463.0, "completions/max_terminated_length": 1463.0, "completions/mean_length": 725.21875, "completions/mean_terminated_length": 725.21875, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "entropy": 0.7850626893341541, "epoch": 0.17687074829931973, "frac_reward_zero_std": 0.25, "grad_norm": 0.6953125, "learning_rate": 8.236961451247165e-07, "loss": 0.0387, "num_tokens": 18457485.0, "reward": 0.515625, "reward_std": 0.8681706190109253, "rewards/accuracy_reward_func/mean": 0.5, "rewards/accuracy_reward_func/std": 0.8728715777397156, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 312, "step_time": 51.591500581242144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1707.0, "completions/max_terminated_length": 1707.0, "completions/mean_length": 686.421875, "completions/mean_terminated_length": 686.421875, "completions/min_length": 262.0, "completions/min_terminated_length": 262.0, "entropy": 0.8607305027544498, "epoch": 0.17743764172335602, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 8.231292517006803e-07, "loss": 0.0428, "num_tokens": 18511328.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 313, "step_time": 59.51167070213705 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 766.125, "completions/mean_terminated_length": 766.125, "completions/min_length": 422.0, "completions/min_terminated_length": 422.0, "entropy": 0.9440110493451357, "epoch": 0.17800453514739228, "frac_reward_zero_std": 0.5, "grad_norm": 0.5, "learning_rate": 8.22562358276644e-07, "loss": -0.0035, "num_tokens": 18570056.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 314, "step_time": 46.22242483403534 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1031.0, "completions/max_terminated_length": 1031.0, "completions/mean_length": 611.09375, "completions/mean_terminated_length": 611.09375, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "entropy": 0.8531279899179935, "epoch": 0.17857142857142858, "frac_reward_zero_std": 0.5, "grad_norm": 0.59375, "learning_rate": 8.219954648526076e-07, "loss": 0.0084, "num_tokens": 18620142.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 315, "step_time": 37.25250487215817 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 812.453125, "completions/mean_terminated_length": 812.453125, "completions/min_length": 345.0, "completions/min_terminated_length": 345.0, "entropy": 0.7896588500589132, "epoch": 0.17913832199546487, "frac_reward_zero_std": 0.75, "grad_norm": 0.330078125, "learning_rate": 8.214285714285713e-07, "loss": 0.0044, "num_tokens": 18683939.0, "reward": 0.1640625, "reward_std": 0.5274578928947449, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 316, "step_time": 68.10056507680565 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1211.0, "completions/max_terminated_length": 1211.0, "completions/mean_length": 701.328125, "completions/mean_terminated_length": 701.328125, "completions/min_length": 18.0, "completions/min_terminated_length": 18.0, "entropy": 0.915490660816431, "epoch": 0.17970521541950113, "frac_reward_zero_std": 0.625, "grad_norm": 0.64453125, "learning_rate": 8.208616780045351e-07, "loss": -0.016, "num_tokens": 18739544.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 317, "step_time": 42.40765083394945 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 684.890625, "completions/mean_terminated_length": 684.890625, "completions/min_length": 363.0, "completions/min_terminated_length": 363.0, "entropy": 0.6934215240180492, "epoch": 0.18027210884353742, "frac_reward_zero_std": 0.375, "grad_norm": 0.609375, "learning_rate": 8.202947845804988e-07, "loss": -0.0324, "num_tokens": 18794417.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 318, "step_time": 40.335626715794206 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 653.875, "completions/mean_terminated_length": 653.875, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "entropy": 1.1461282074451447, "epoch": 0.18083900226757368, "frac_reward_zero_std": 0.75, "grad_norm": 0.470703125, "learning_rate": 8.197278911564626e-07, "loss": -0.0109, "num_tokens": 18845817.0, "reward": 0.0703125, "reward_std": 0.27984103560447693, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 319, "step_time": 41.016207785345614 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 781.03125, "completions/mean_terminated_length": 781.03125, "completions/min_length": 359.0, "completions/min_terminated_length": 359.0, "entropy": 0.7973394598811865, "epoch": 0.18140589569160998, "frac_reward_zero_std": 0.625, "grad_norm": 0.484375, "learning_rate": 8.191609977324263e-07, "loss": -0.0217, "num_tokens": 18905923.0, "reward": 0.171875, "reward_std": 0.4292735755443573, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.109375, "rewards/format_reward_func/std": 0.2083333432674408, "step": 320, "step_time": 60.48013670463115 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1551.0, "completions/max_terminated_length": 1551.0, "completions/mean_length": 760.359375, "completions/mean_terminated_length": 760.359375, "completions/min_length": 375.0, "completions/min_terminated_length": 375.0, "entropy": 0.6594368945807219, "epoch": 0.18197278911564627, "frac_reward_zero_std": 0.25, "grad_norm": 0.578125, "learning_rate": 8.1859410430839e-07, "loss": -0.0598, "num_tokens": 18965026.0, "reward": 0.4296875, "reward_std": 0.7759642601013184, "rewards/accuracy_reward_func/mean": 0.375, "rewards/accuracy_reward_func/std": 0.7867957949638367, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 321, "step_time": 54.980602267198265 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1466.0, "completions/max_terminated_length": 1466.0, "completions/mean_length": 709.921875, "completions/mean_terminated_length": 709.921875, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "entropy": 0.7112272661179304, "epoch": 0.18253968253968253, "frac_reward_zero_std": 0.5, "grad_norm": 0.52734375, "learning_rate": 8.180272108843537e-07, "loss": 0.0114, "num_tokens": 19021141.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 322, "step_time": 51.63816947955638 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1665.0, "completions/max_terminated_length": 1665.0, "completions/mean_length": 861.578125, "completions/mean_terminated_length": 861.578125, "completions/min_length": 348.0, "completions/min_terminated_length": 348.0, "entropy": 0.6500704195350409, "epoch": 0.18310657596371882, "frac_reward_zero_std": 0.625, "grad_norm": 0.384765625, "learning_rate": 8.174603174603174e-07, "loss": -0.0154, "num_tokens": 19085874.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 323, "step_time": 58.53681822679937 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1450.0, "completions/max_terminated_length": 1450.0, "completions/mean_length": 767.546875, "completions/mean_terminated_length": 767.546875, "completions/min_length": 301.0, "completions/min_terminated_length": 301.0, "entropy": 0.8385312054306269, "epoch": 0.1836734693877551, "frac_reward_zero_std": 0.625, "grad_norm": 0.41015625, "learning_rate": 8.168934240362811e-07, "loss": -0.0013, "num_tokens": 19144493.0, "reward": 0.3125, "reward_std": 0.7210001945495605, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 324, "step_time": 51.178459596820176 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1679.0, "completions/max_terminated_length": 1679.0, "completions/mean_length": 688.390625, "completions/mean_terminated_length": 688.390625, "completions/min_length": 153.0, "completions/min_terminated_length": 153.0, "entropy": 0.8860599715262651, "epoch": 0.18424036281179137, "frac_reward_zero_std": 0.625, "grad_norm": 0.46875, "learning_rate": 8.163265306122449e-07, "loss": -0.0275, "num_tokens": 19199726.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 325, "step_time": 59.97611542791128 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1238.0, "completions/max_terminated_length": 1238.0, "completions/mean_length": 694.484375, "completions/mean_terminated_length": 694.484375, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "entropy": 0.8037873115390539, "epoch": 0.18480725623582767, "frac_reward_zero_std": 0.5, "grad_norm": 0.58984375, "learning_rate": 8.157596371882086e-07, "loss": 0.0105, "num_tokens": 19255077.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 326, "step_time": 43.73556720651686 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2087.0, "completions/max_terminated_length": 2087.0, "completions/mean_length": 821.625, "completions/mean_terminated_length": 821.625, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "entropy": 0.6347903534770012, "epoch": 0.18537414965986396, "frac_reward_zero_std": 0.5, "grad_norm": 0.4453125, "learning_rate": 8.151927437641724e-07, "loss": -0.0118, "num_tokens": 19319261.0, "reward": 0.2578125, "reward_std": 0.654606282711029, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 327, "step_time": 75.29317637253553 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1285.0, "completions/max_terminated_length": 1285.0, "completions/mean_length": 807.21875, "completions/mean_terminated_length": 807.21875, "completions/min_length": 4.0, "completions/min_terminated_length": 4.0, "entropy": 0.7769613545387983, "epoch": 0.18594104308390022, "frac_reward_zero_std": 0.375, "grad_norm": 0.578125, "learning_rate": 8.14625850340136e-07, "loss": 0.0082, "num_tokens": 19380979.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 328, "step_time": 44.861584201455116 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1734.0, "completions/max_terminated_length": 1734.0, "completions/mean_length": 747.203125, "completions/mean_terminated_length": 747.203125, "completions/min_length": 392.0, "completions/min_terminated_length": 392.0, "entropy": 0.8658007867634296, "epoch": 0.1865079365079365, "frac_reward_zero_std": 0.875, "grad_norm": 0.26171875, "learning_rate": 8.140589569160997e-07, "loss": 0.0011, "num_tokens": 19440032.0, "reward": 0.0625, "reward_std": 0.2745848298072815, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 329, "step_time": 61.10620769113302 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 709.296875, "completions/mean_terminated_length": 709.296875, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "entropy": 0.8043992314487696, "epoch": 0.1870748299319728, "frac_reward_zero_std": 0.125, "grad_norm": 0.75, "learning_rate": 8.134920634920635e-07, "loss": -0.0316, "num_tokens": 19495355.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 330, "step_time": 42.660171108320355 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1563.0, "completions/max_terminated_length": 1563.0, "completions/mean_length": 638.28125, "completions/mean_terminated_length": 638.28125, "completions/min_length": 9.0, "completions/min_terminated_length": 9.0, "entropy": 1.0374234952032566, "epoch": 0.18764172335600907, "frac_reward_zero_std": 0.625, "grad_norm": 0.51953125, "learning_rate": 8.129251700680271e-07, "loss": -0.0039, "num_tokens": 19545749.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 331, "step_time": 54.728871271014214 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 864.421875, "completions/mean_terminated_length": 864.421875, "completions/min_length": 291.0, "completions/min_terminated_length": 291.0, "entropy": 0.7442141883075237, "epoch": 0.18820861678004536, "frac_reward_zero_std": 0.875, "grad_norm": 0.197265625, "learning_rate": 8.123582766439909e-07, "loss": -0.0118, "num_tokens": 19614520.0, "reward": 0.046875, "reward_std": 0.14689241349697113, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 332, "step_time": 98.94229808729142 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 883.140625, "completions/mean_terminated_length": 883.140625, "completions/min_length": 434.0, "completions/min_terminated_length": 434.0, "entropy": 0.6293924115598202, "epoch": 0.18877551020408162, "frac_reward_zero_std": 0.5, "grad_norm": 0.421875, "learning_rate": 8.117913832199546e-07, "loss": -0.0028, "num_tokens": 19681265.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 333, "step_time": 52.70649899542332 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1195.0, "completions/max_terminated_length": 1195.0, "completions/mean_length": 748.640625, "completions/mean_terminated_length": 748.640625, "completions/min_length": 314.0, "completions/min_terminated_length": 314.0, "entropy": 0.8612051568925381, "epoch": 0.1893424036281179, "frac_reward_zero_std": 0.625, "grad_norm": 0.408203125, "learning_rate": 8.112244897959183e-07, "loss": 0.0353, "num_tokens": 19738930.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 334, "step_time": 41.320436858572066 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1750.0, "completions/max_terminated_length": 1750.0, "completions/mean_length": 853.046875, "completions/mean_terminated_length": 853.046875, "completions/min_length": 333.0, "completions/min_terminated_length": 333.0, "entropy": 0.5430580489337444, "epoch": 0.1899092970521542, "frac_reward_zero_std": 0.25, "grad_norm": 0.515625, "learning_rate": 8.10657596371882e-07, "loss": 0.0235, "num_tokens": 19804261.0, "reward": 0.515625, "reward_std": 0.9299630522727966, "rewards/accuracy_reward_func/mean": 0.46875, "rewards/accuracy_reward_func/std": 0.8539125919342041, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 335, "step_time": 60.53818303346634 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1238.0, "completions/max_terminated_length": 1238.0, "completions/mean_length": 609.25, "completions/mean_terminated_length": 609.25, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "entropy": 0.9052265919744968, "epoch": 0.19047619047619047, "frac_reward_zero_std": 0.625, "grad_norm": 0.5390625, "learning_rate": 8.100907029478458e-07, "loss": 0.0256, "num_tokens": 19852877.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 336, "step_time": 42.84950289875269 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 850.046875, "completions/mean_terminated_length": 850.046875, "completions/min_length": 405.0, "completions/min_terminated_length": 405.0, "entropy": 0.6594413593411446, "epoch": 0.19104308390022676, "frac_reward_zero_std": 0.5, "grad_norm": 0.443359375, "learning_rate": 8.095238095238095e-07, "loss": 0.0106, "num_tokens": 19917728.0, "reward": 0.3046875, "reward_std": 0.664384126663208, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 337, "step_time": 48.65508953295648 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 688.390625, "completions/mean_terminated_length": 688.390625, "completions/min_length": 198.0, "completions/min_terminated_length": 198.0, "entropy": 0.9500348977744579, "epoch": 0.19160997732426305, "frac_reward_zero_std": 0.875, "grad_norm": 0.423828125, "learning_rate": 8.089569160997733e-07, "loss": 0.002, "num_tokens": 19974329.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 338, "step_time": 41.63581903837621 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1230.0, "completions/max_terminated_length": 1230.0, "completions/mean_length": 739.140625, "completions/mean_terminated_length": 739.140625, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "entropy": 0.8920802474021912, "epoch": 0.1921768707482993, "frac_reward_zero_std": 0.75, "grad_norm": 0.330078125, "learning_rate": 8.083900226757369e-07, "loss": -0.0019, "num_tokens": 20033394.0, "reward": 0.1171875, "reward_std": 0.46922925114631653, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 339, "step_time": 43.43856304511428 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1392.0, "completions/max_terminated_length": 1392.0, "completions/mean_length": 806.859375, "completions/mean_terminated_length": 806.859375, "completions/min_length": 352.0, "completions/min_terminated_length": 352.0, "entropy": 0.8819306213408709, "epoch": 0.1927437641723356, "frac_reward_zero_std": 0.75, "grad_norm": 0.33203125, "learning_rate": 8.078231292517006e-07, "loss": 0.0083, "num_tokens": 20095577.0, "reward": 0.125, "reward_std": 0.5039526224136353, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 340, "step_time": 48.59819230623543 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1377.0, "completions/max_terminated_length": 1377.0, "completions/mean_length": 826.03125, "completions/mean_terminated_length": 826.03125, "completions/min_length": 440.0, "completions/min_terminated_length": 440.0, "entropy": 1.1564702279865742, "epoch": 0.1933106575963719, "frac_reward_zero_std": 0.75, "grad_norm": 0.38671875, "learning_rate": 8.072562358276643e-07, "loss": 0.0051, "num_tokens": 20161203.0, "reward": 0.125, "reward_std": 0.4364357888698578, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 341, "step_time": 48.10737384017557 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1516.0, "completions/max_terminated_length": 1516.0, "completions/mean_length": 849.359375, "completions/mean_terminated_length": 849.359375, "completions/min_length": 399.0, "completions/min_terminated_length": 399.0, "entropy": 0.6847367249429226, "epoch": 0.19387755102040816, "frac_reward_zero_std": 0.625, "grad_norm": 0.41015625, "learning_rate": 8.066893424036281e-07, "loss": -0.02, "num_tokens": 20225370.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 342, "step_time": 52.841353938914835 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1541.0, "completions/max_terminated_length": 1541.0, "completions/mean_length": 911.234375, "completions/mean_terminated_length": 911.234375, "completions/min_length": 434.0, "completions/min_terminated_length": 434.0, "entropy": 0.6147547662258148, "epoch": 0.19444444444444445, "frac_reward_zero_std": 0.5, "grad_norm": 0.416015625, "learning_rate": 8.061224489795918e-07, "loss": 0.0062, "num_tokens": 20294697.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 343, "step_time": 53.379401203244925 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 765.8125, "completions/mean_terminated_length": 765.8125, "completions/min_length": 383.0, "completions/min_terminated_length": 383.0, "entropy": 0.7179710138589144, "epoch": 0.19501133786848074, "frac_reward_zero_std": 0.125, "grad_norm": 0.61328125, "learning_rate": 8.055555555555556e-07, "loss": -0.0035, "num_tokens": 20353541.0, "reward": 0.265625, "reward_std": 0.6664806604385376, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 344, "step_time": 52.01110478769988 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1386.0, "completions/max_terminated_length": 1386.0, "completions/mean_length": 773.6875, "completions/mean_terminated_length": 773.6875, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "entropy": 0.9038583058863878, "epoch": 0.195578231292517, "frac_reward_zero_std": 0.875, "grad_norm": 0.25390625, "learning_rate": 8.049886621315193e-07, "loss": -0.0172, "num_tokens": 20415897.0, "reward": 0.0390625, "reward_std": 0.13524484634399414, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 345, "step_time": 48.716181214898825 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1084.0, "completions/max_terminated_length": 1084.0, "completions/mean_length": 632.28125, "completions/mean_terminated_length": 632.28125, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "entropy": 0.9033645987510681, "epoch": 0.1961451247165533, "frac_reward_zero_std": 0.25, "grad_norm": 0.7578125, "learning_rate": 8.04421768707483e-07, "loss": 0.0293, "num_tokens": 20467507.0, "reward": 0.46875, "reward_std": 0.8158888816833496, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 346, "step_time": 37.71913564391434 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1516.0, "completions/max_terminated_length": 1516.0, "completions/mean_length": 728.0, "completions/mean_terminated_length": 728.0, "completions/min_length": 353.0, "completions/min_terminated_length": 353.0, "entropy": 0.512436656281352, "epoch": 0.19671201814058956, "frac_reward_zero_std": 0.5, "grad_norm": 0.51171875, "learning_rate": 8.038548752834466e-07, "loss": 0.0255, "num_tokens": 20522883.0, "reward": 0.2578125, "reward_std": 0.654606282711029, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 347, "step_time": 52.17997798230499 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1552.0, "completions/max_terminated_length": 1552.0, "completions/mean_length": 782.9375, "completions/mean_terminated_length": 782.9375, "completions/min_length": 337.0, "completions/min_terminated_length": 337.0, "entropy": 0.8477420397102833, "epoch": 0.19727891156462585, "frac_reward_zero_std": 0.75, "grad_norm": 0.33203125, "learning_rate": 8.032879818594103e-07, "loss": -0.0042, "num_tokens": 20582663.0, "reward": 0.0546875, "reward_std": 0.26899558305740356, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 348, "step_time": 52.993523797020316 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1714.0, "completions/max_terminated_length": 1714.0, "completions/mean_length": 686.515625, "completions/mean_terminated_length": 686.515625, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "entropy": 1.0396155789494514, "epoch": 0.19784580498866214, "frac_reward_zero_std": 0.5, "grad_norm": 0.64453125, "learning_rate": 8.027210884353741e-07, "loss": 0.012, "num_tokens": 20637984.0, "reward": 0.3125, "reward_std": 0.7426892518997192, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 349, "step_time": 59.609021263197064 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2002.0, "completions/max_terminated_length": 2002.0, "completions/mean_length": 756.28125, "completions/mean_terminated_length": 756.28125, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "entropy": 0.7620594538748264, "epoch": 0.1984126984126984, "frac_reward_zero_std": 0.75, "grad_norm": 0.412109375, "learning_rate": 8.021541950113378e-07, "loss": -0.0012, "num_tokens": 20701946.0, "reward": 0.1484375, "reward_std": 0.47709178924560547, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 350, "step_time": 74.73007242754102 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 698.890625, "completions/mean_terminated_length": 698.890625, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "entropy": 0.77883305773139, "epoch": 0.1989795918367347, "frac_reward_zero_std": 0.5, "grad_norm": 0.57421875, "learning_rate": 8.015873015873016e-07, "loss": -0.0028, "num_tokens": 20758195.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 351, "step_time": 41.66911055613309 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1134.0, "completions/max_terminated_length": 1134.0, "completions/mean_length": 720.765625, "completions/mean_terminated_length": 720.765625, "completions/min_length": 287.0, "completions/min_terminated_length": 287.0, "entropy": 0.8328768480569124, "epoch": 0.19954648526077098, "frac_reward_zero_std": 0.625, "grad_norm": 0.41796875, "learning_rate": 8.010204081632652e-07, "loss": -0.0443, "num_tokens": 20814124.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 352, "step_time": 38.688570947386324 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 740.203125, "completions/mean_terminated_length": 740.203125, "completions/min_length": 349.0, "completions/min_terminated_length": 349.0, "entropy": 1.0927176661789417, "epoch": 0.20011337868480725, "frac_reward_zero_std": 0.75, "grad_norm": 0.58984375, "learning_rate": 8.00453514739229e-07, "loss": -0.0254, "num_tokens": 20871633.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 353, "step_time": 43.63492278382182 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1773.0, "completions/max_terminated_length": 1773.0, "completions/mean_length": 745.609375, "completions/mean_terminated_length": 745.609375, "completions/min_length": 283.0, "completions/min_terminated_length": 283.0, "entropy": 0.582351628690958, "epoch": 0.20068027210884354, "frac_reward_zero_std": 0.25, "grad_norm": 0.6640625, "learning_rate": 7.998866213151927e-07, "loss": 0.0291, "num_tokens": 20929368.0, "reward": 0.53125, "reward_std": 0.8723031282424927, "rewards/accuracy_reward_func/mean": 0.46875, "rewards/accuracy_reward_func/std": 0.8539125919342041, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 354, "step_time": 60.86191135738045 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1450.0, "completions/max_terminated_length": 1450.0, "completions/mean_length": 767.15625, "completions/mean_terminated_length": 767.15625, "completions/min_length": 353.0, "completions/min_terminated_length": 353.0, "entropy": 0.8429246470332146, "epoch": 0.20124716553287983, "frac_reward_zero_std": 0.5, "grad_norm": 0.546875, "learning_rate": 7.993197278911564e-07, "loss": -0.0084, "num_tokens": 20987434.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 355, "step_time": 49.143246746622026 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1267.0, "completions/max_terminated_length": 1267.0, "completions/mean_length": 740.703125, "completions/mean_terminated_length": 740.703125, "completions/min_length": 289.0, "completions/min_terminated_length": 289.0, "entropy": 0.6812925096601248, "epoch": 0.2018140589569161, "frac_reward_zero_std": 0.75, "grad_norm": 0.337890625, "learning_rate": 7.987528344671201e-07, "loss": 0.0336, "num_tokens": 21049383.0, "reward": 0.1796875, "reward_std": 0.5298036932945251, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 356, "step_time": 44.72687572520226 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 764.78125, "completions/mean_terminated_length": 764.78125, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "entropy": 0.7700738701969385, "epoch": 0.20238095238095238, "frac_reward_zero_std": 0.375, "grad_norm": 0.546875, "learning_rate": 7.981859410430839e-07, "loss": -0.0, "num_tokens": 21107657.0, "reward": 0.3828125, "reward_std": 0.7545743584632874, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 357, "step_time": 56.46399795729667 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2142.0, "completions/max_terminated_length": 2142.0, "completions/mean_length": 775.359375, "completions/mean_terminated_length": 775.359375, "completions/min_length": 180.0, "completions/min_terminated_length": 180.0, "entropy": 0.6098148673772812, "epoch": 0.20294784580498867, "frac_reward_zero_std": 0.5, "grad_norm": 0.4765625, "learning_rate": 7.976190476190476e-07, "loss": 0.0411, "num_tokens": 21168424.0, "reward": 0.2421875, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 358, "step_time": 75.8587139165029 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 665.546875, "completions/mean_terminated_length": 665.546875, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "entropy": 0.8294234424829483, "epoch": 0.20351473922902494, "frac_reward_zero_std": 0.5, "grad_norm": 0.5859375, "learning_rate": 7.970521541950113e-07, "loss": -0.0151, "num_tokens": 21221739.0, "reward": 0.2890625, "reward_std": 0.6655030846595764, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 359, "step_time": 55.06628938857466 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2727.0, "completions/max_terminated_length": 2727.0, "completions/mean_length": 776.59375, "completions/mean_terminated_length": 776.59375, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "entropy": 0.6692650187760592, "epoch": 0.20408163265306123, "frac_reward_zero_std": 0.125, "grad_norm": 0.6875, "learning_rate": 7.96485260770975e-07, "loss": -0.0408, "num_tokens": 21280881.0, "reward": 0.375, "reward_std": 0.7559289336204529, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 360, "step_time": 99.08741320390254 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1522.0, "completions/max_terminated_length": 1522.0, "completions/mean_length": 735.9375, "completions/mean_terminated_length": 735.9375, "completions/min_length": 11.0, "completions/min_terminated_length": 11.0, "entropy": 0.9531947933137417, "epoch": 0.2046485260770975, "frac_reward_zero_std": 0.75, "grad_norm": 0.39453125, "learning_rate": 7.959183673469388e-07, "loss": 0.0084, "num_tokens": 21339093.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 361, "step_time": 52.08562697377056 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1211.0, "completions/max_terminated_length": 1211.0, "completions/mean_length": 733.140625, "completions/mean_terminated_length": 733.140625, "completions/min_length": 297.0, "completions/min_terminated_length": 297.0, "entropy": 0.937473900616169, "epoch": 0.20521541950113378, "frac_reward_zero_std": 0.75, "grad_norm": 0.400390625, "learning_rate": 7.953514739229025e-07, "loss": 0.0093, "num_tokens": 21395806.0, "reward": 0.0703125, "reward_std": 0.27984103560447693, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 362, "step_time": 41.219106522388756 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 679.3125, "completions/mean_terminated_length": 679.3125, "completions/min_length": 201.0, "completions/min_terminated_length": 201.0, "entropy": 0.9143323600292206, "epoch": 0.20578231292517007, "frac_reward_zero_std": 0.75, "grad_norm": 0.40625, "learning_rate": 7.947845804988662e-07, "loss": 0.0205, "num_tokens": 21449338.0, "reward": 0.265625, "reward_std": 0.6544642448425293, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 363, "step_time": 38.76798083074391 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1269.0, "completions/max_terminated_length": 1269.0, "completions/mean_length": 633.0, "completions/mean_terminated_length": 633.0, "completions/min_length": 24.0, "completions/min_terminated_length": 24.0, "entropy": 0.8506261613219976, "epoch": 0.20634920634920634, "frac_reward_zero_std": 0.625, "grad_norm": 0.50390625, "learning_rate": 7.942176870748299e-07, "loss": -0.0036, "num_tokens": 21502546.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 364, "step_time": 43.8399096513167 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1347.0, "completions/max_terminated_length": 1347.0, "completions/mean_length": 703.875, "completions/mean_terminated_length": 703.875, "completions/min_length": 297.0, "completions/min_terminated_length": 297.0, "entropy": 0.8162322528660297, "epoch": 0.20691609977324263, "frac_reward_zero_std": 0.375, "grad_norm": 0.671875, "learning_rate": 7.936507936507936e-07, "loss": 0.0111, "num_tokens": 21556562.0, "reward": 0.3515625, "reward_std": 0.7274640202522278, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 365, "step_time": 46.54994601197541 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1237.0, "completions/max_terminated_length": 1237.0, "completions/mean_length": 737.890625, "completions/mean_terminated_length": 737.890625, "completions/min_length": 392.0, "completions/min_terminated_length": 392.0, "entropy": 0.6952655706554651, "epoch": 0.20748299319727892, "frac_reward_zero_std": 0.75, "grad_norm": 0.37890625, "learning_rate": 7.930839002267573e-07, "loss": 0.0004, "num_tokens": 21616107.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 366, "step_time": 42.6978315198794 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2991.0, "completions/max_terminated_length": 2991.0, "completions/mean_length": 821.578125, "completions/mean_terminated_length": 821.578125, "completions/min_length": 362.0, "completions/min_terminated_length": 362.0, "entropy": 0.8010297678411007, "epoch": 0.20804988662131518, "frac_reward_zero_std": 0.5, "grad_norm": 0.45703125, "learning_rate": 7.92517006802721e-07, "loss": -0.0124, "num_tokens": 21679152.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 367, "step_time": 110.2668658932671 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1148.0, "completions/max_terminated_length": 1148.0, "completions/mean_length": 692.90625, "completions/mean_terminated_length": 692.90625, "completions/min_length": 24.0, "completions/min_terminated_length": 24.0, "entropy": 0.9686912037432194, "epoch": 0.20861678004535147, "frac_reward_zero_std": 0.5, "grad_norm": 0.56640625, "learning_rate": 7.919501133786848e-07, "loss": -0.0076, "num_tokens": 21734586.0, "reward": 0.4375, "reward_std": 0.774084210395813, "rewards/accuracy_reward_func/mean": 0.375, "rewards/accuracy_reward_func/std": 0.7867957949638367, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 368, "step_time": 41.391304216347635 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1621.0, "completions/max_terminated_length": 1621.0, "completions/mean_length": 831.9375, "completions/mean_terminated_length": 831.9375, "completions/min_length": 390.0, "completions/min_terminated_length": 390.0, "entropy": 0.8207549266517162, "epoch": 0.20918367346938777, "frac_reward_zero_std": 0.625, "grad_norm": 0.41015625, "learning_rate": 7.913832199546485e-07, "loss": 0.0034, "num_tokens": 21798350.0, "reward": 0.21875, "reward_std": 0.5900968313217163, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 369, "step_time": 57.505966384895146 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2418.0, "completions/max_terminated_length": 2418.0, "completions/mean_length": 753.28125, "completions/mean_terminated_length": 753.28125, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "entropy": 0.897388132289052, "epoch": 0.20975056689342403, "frac_reward_zero_std": 0.75, "grad_norm": 0.37890625, "learning_rate": 7.908163265306123e-07, "loss": -0.0002, "num_tokens": 21857816.0, "reward": 0.1171875, "reward_std": 0.41539984941482544, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 370, "step_time": 86.28427155595273 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 707.015625, "completions/mean_terminated_length": 707.015625, "completions/min_length": 301.0, "completions/min_terminated_length": 301.0, "entropy": 0.9148596245795488, "epoch": 0.21031746031746032, "frac_reward_zero_std": 0.75, "grad_norm": 0.484375, "learning_rate": 7.902494331065758e-07, "loss": 0.0293, "num_tokens": 21913553.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 371, "step_time": 53.49216768797487 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1359.0, "completions/max_terminated_length": 1359.0, "completions/mean_length": 789.859375, "completions/mean_terminated_length": 789.859375, "completions/min_length": 300.0, "completions/min_terminated_length": 300.0, "entropy": 0.7241196781396866, "epoch": 0.2108843537414966, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 7.896825396825396e-07, "loss": -0.0076, "num_tokens": 21973280.0, "reward": 0.2421875, "reward_std": 0.666620135307312, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 372, "step_time": 46.46333046257496 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1379.0, "completions/max_terminated_length": 1379.0, "completions/mean_length": 793.625, "completions/mean_terminated_length": 793.625, "completions/min_length": 273.0, "completions/min_terminated_length": 273.0, "entropy": 0.9126806035637856, "epoch": 0.21145124716553287, "frac_reward_zero_std": 0.75, "grad_norm": 0.333984375, "learning_rate": 7.891156462585033e-07, "loss": 0.0071, "num_tokens": 22033992.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 373, "step_time": 47.039062136784196 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1225.0, "completions/max_terminated_length": 1225.0, "completions/mean_length": 658.15625, "completions/mean_terminated_length": 658.15625, "completions/min_length": 349.0, "completions/min_terminated_length": 349.0, "entropy": 0.9130817167460918, "epoch": 0.21201814058956917, "frac_reward_zero_std": 0.75, "grad_norm": 0.33984375, "learning_rate": 7.885487528344671e-07, "loss": 0.003, "num_tokens": 22086234.0, "reward": 0.1875, "reward_std": 0.559903621673584, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 374, "step_time": 41.82396076992154 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1455.0, "completions/max_terminated_length": 1455.0, "completions/mean_length": 768.046875, "completions/mean_terminated_length": 768.046875, "completions/min_length": 254.0, "completions/min_terminated_length": 254.0, "entropy": 0.6956988330930471, "epoch": 0.21258503401360543, "frac_reward_zero_std": 0.625, "grad_norm": 0.470703125, "learning_rate": 7.879818594104308e-07, "loss": 0.0286, "num_tokens": 22145509.0, "reward": 0.1484375, "reward_std": 0.49344661831855774, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 375, "step_time": 49.9479739703238 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 712.703125, "completions/mean_terminated_length": 712.703125, "completions/min_length": 303.0, "completions/min_terminated_length": 303.0, "entropy": 0.7200741954147816, "epoch": 0.21315192743764172, "frac_reward_zero_std": 0.375, "grad_norm": 0.6171875, "learning_rate": 7.874149659863946e-07, "loss": 0.0405, "num_tokens": 22201378.0, "reward": 0.484375, "reward_std": 0.8589803576469421, "rewards/accuracy_reward_func/mean": 0.4375, "rewards/accuracy_reward_func/std": 0.8333333730697632, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 376, "step_time": 40.888242116197944 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 698.328125, "completions/mean_terminated_length": 698.328125, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "entropy": 0.6802572142332792, "epoch": 0.213718820861678, "frac_reward_zero_std": 0.875, "grad_norm": 0.294921875, "learning_rate": 7.868480725623582e-07, "loss": -0.0053, "num_tokens": 22254615.0, "reward": 0.0703125, "reward_std": 0.33174845576286316, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 377, "step_time": 46.84081286471337 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 697.703125, "completions/mean_terminated_length": 697.703125, "completions/min_length": 379.0, "completions/min_terminated_length": 379.0, "entropy": 0.7381638363003731, "epoch": 0.21428571428571427, "frac_reward_zero_std": 0.25, "grad_norm": 0.640625, "learning_rate": 7.86281179138322e-07, "loss": 0.0413, "num_tokens": 22309500.0, "reward": 0.4765625, "reward_std": 0.8234156966209412, "rewards/accuracy_reward_func/mean": 0.4375, "rewards/accuracy_reward_func/std": 0.8333333730697632, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 378, "step_time": 45.019169168546796 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1311.0, "completions/max_terminated_length": 1311.0, "completions/mean_length": 734.625, "completions/mean_terminated_length": 734.625, "completions/min_length": 384.0, "completions/min_terminated_length": 384.0, "entropy": 1.001370683312416, "epoch": 0.21485260770975056, "frac_reward_zero_std": 0.625, "grad_norm": 0.478515625, "learning_rate": 7.857142857142856e-07, "loss": 0.0238, "num_tokens": 22368140.0, "reward": 0.1015625, "reward_std": 0.3692515790462494, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 379, "step_time": 44.70761434081942 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1632.0, "completions/max_terminated_length": 1632.0, "completions/mean_length": 736.71875, "completions/mean_terminated_length": 736.71875, "completions/min_length": 172.0, "completions/min_terminated_length": 172.0, "entropy": 0.8948388043791056, "epoch": 0.21541950113378686, "frac_reward_zero_std": 0.5, "grad_norm": 0.58203125, "learning_rate": 7.851473922902494e-07, "loss": -0.0129, "num_tokens": 22430162.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 380, "step_time": 56.97926906310022 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1786.0, "completions/max_terminated_length": 1786.0, "completions/mean_length": 808.390625, "completions/mean_terminated_length": 808.390625, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "entropy": 0.6405720394104719, "epoch": 0.21598639455782312, "frac_reward_zero_std": 0.375, "grad_norm": 0.57421875, "learning_rate": 7.845804988662131e-07, "loss": 0.0168, "num_tokens": 22492107.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 381, "step_time": 62.248050440102816 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2058.0, "completions/max_terminated_length": 2058.0, "completions/mean_length": 756.203125, "completions/mean_terminated_length": 756.203125, "completions/min_length": 220.0, "completions/min_terminated_length": 220.0, "entropy": 0.7315251268446445, "epoch": 0.2165532879818594, "frac_reward_zero_std": 0.75, "grad_norm": 0.369140625, "learning_rate": 7.840136054421769e-07, "loss": -0.0089, "num_tokens": 22550776.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 382, "step_time": 71.68382942955941 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1127.0, "completions/max_terminated_length": 1127.0, "completions/mean_length": 559.9375, "completions/mean_terminated_length": 559.9375, "completions/min_length": 210.0, "completions/min_terminated_length": 210.0, "entropy": 0.8677006028592587, "epoch": 0.2171201814058957, "frac_reward_zero_std": 0.25, "grad_norm": 0.7578125, "learning_rate": 7.834467120181405e-07, "loss": 0.0009, "num_tokens": 22597796.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 383, "step_time": 38.73852313775569 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1409.0, "completions/max_terminated_length": 1409.0, "completions/mean_length": 873.34375, "completions/mean_terminated_length": 873.34375, "completions/min_length": 374.0, "completions/min_terminated_length": 374.0, "entropy": 1.0035483818501234, "epoch": 0.21768707482993196, "frac_reward_zero_std": 0.875, "grad_norm": 0.185546875, "learning_rate": 7.828798185941043e-07, "loss": 0.0003, "num_tokens": 22666370.0, "reward": 0.078125, "reward_std": 0.33592739701271057, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 384, "step_time": 48.65996228437871 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1265.0, "completions/mean_length": 731.875, "completions/mean_terminated_length": 678.4762573242188, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 1.0197479017078876, "epoch": 0.21825396825396826, "frac_reward_zero_std": 0.75, "grad_norm": 0.37890625, "learning_rate": 7.82312925170068e-07, "loss": 0.0032, "num_tokens": 22723634.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 385, "step_time": 170.17113096639514 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1078.0, "completions/max_terminated_length": 1078.0, "completions/mean_length": 688.109375, "completions/mean_terminated_length": 688.109375, "completions/min_length": 276.0, "completions/min_terminated_length": 276.0, "entropy": 0.9092415310442448, "epoch": 0.21882086167800455, "frac_reward_zero_std": 0.5, "grad_norm": 0.6015625, "learning_rate": 7.817460317460318e-07, "loss": -0.0048, "num_tokens": 22782593.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 386, "step_time": 39.27197832521051 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1406.0, "completions/max_terminated_length": 1406.0, "completions/mean_length": 674.8125, "completions/mean_terminated_length": 674.8125, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "entropy": 0.8119625337421894, "epoch": 0.2193877551020408, "frac_reward_zero_std": 0.75, "grad_norm": 0.39453125, "learning_rate": 7.811791383219954e-07, "loss": -0.0137, "num_tokens": 22837981.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 387, "step_time": 49.66723500657827 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1304.0, "completions/max_terminated_length": 1304.0, "completions/mean_length": 739.5625, "completions/mean_terminated_length": 739.5625, "completions/min_length": 11.0, "completions/min_terminated_length": 11.0, "entropy": 0.7483879532665014, "epoch": 0.2199546485260771, "frac_reward_zero_std": 0.75, "grad_norm": 0.345703125, "learning_rate": 7.806122448979591e-07, "loss": -0.0148, "num_tokens": 22895241.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 388, "step_time": 46.12076739780605 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1423.0, "completions/max_terminated_length": 1423.0, "completions/mean_length": 783.90625, "completions/mean_terminated_length": 783.90625, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "entropy": 0.9556456692516804, "epoch": 0.2205215419501134, "frac_reward_zero_std": 0.875, "grad_norm": 0.24609375, "learning_rate": 7.800453514739229e-07, "loss": -0.008, "num_tokens": 22957635.0, "reward": 0.109375, "reward_std": 0.41277948021888733, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 389, "step_time": 49.99157304875553 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1285.0, "completions/max_terminated_length": 1285.0, "completions/mean_length": 772.96875, "completions/mean_terminated_length": 772.96875, "completions/min_length": 433.0, "completions/min_terminated_length": 433.0, "entropy": 0.7684030756354332, "epoch": 0.22108843537414966, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 7.794784580498865e-07, "loss": 0.0129, "num_tokens": 23017057.0, "reward": 0.3046875, "reward_std": 0.664384126663208, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 390, "step_time": 43.90013815369457 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1277.0, "completions/max_terminated_length": 1277.0, "completions/mean_length": 611.40625, "completions/mean_terminated_length": 611.40625, "completions/min_length": 303.0, "completions/min_terminated_length": 303.0, "entropy": 0.855836596339941, "epoch": 0.22165532879818595, "frac_reward_zero_std": 0.5, "grad_norm": 0.59765625, "learning_rate": 7.789115646258503e-07, "loss": 0.0222, "num_tokens": 23066291.0, "reward": 0.3984375, "reward_std": 0.7516103982925415, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 391, "step_time": 42.97751526068896 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1177.0, "completions/max_terminated_length": 1177.0, "completions/mean_length": 690.765625, "completions/mean_terminated_length": 690.765625, "completions/min_length": 336.0, "completions/min_terminated_length": 336.0, "entropy": 0.7666698209941387, "epoch": 0.2222222222222222, "frac_reward_zero_std": 0.625, "grad_norm": 0.435546875, "learning_rate": 7.78344671201814e-07, "loss": 0.0101, "num_tokens": 23120540.0, "reward": 0.25, "reward_std": 0.6900655627250671, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 392, "step_time": 40.13614541944116 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1224.0, "completions/max_terminated_length": 1224.0, "completions/mean_length": 711.40625, "completions/mean_terminated_length": 711.40625, "completions/min_length": 350.0, "completions/min_terminated_length": 350.0, "entropy": 0.9489793553948402, "epoch": 0.2227891156462585, "frac_reward_zero_std": 0.75, "grad_norm": 0.494140625, "learning_rate": 7.777777777777778e-07, "loss": -0.0162, "num_tokens": 23175838.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 393, "step_time": 42.049905179068446 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 808.875, "completions/mean_terminated_length": 808.875, "completions/min_length": 405.0, "completions/min_terminated_length": 405.0, "entropy": 0.784135427325964, "epoch": 0.2233560090702948, "frac_reward_zero_std": 0.375, "grad_norm": 0.58203125, "learning_rate": 7.772108843537415e-07, "loss": 0.0124, "num_tokens": 23239006.0, "reward": 0.265625, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 394, "step_time": 50.6201617596671 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1621.0, "completions/max_terminated_length": 1621.0, "completions/mean_length": 824.5, "completions/mean_terminated_length": 824.5, "completions/min_length": 213.0, "completions/min_terminated_length": 213.0, "entropy": 0.6317049227654934, "epoch": 0.22392290249433106, "frac_reward_zero_std": 0.5, "grad_norm": 0.45703125, "learning_rate": 7.766439909297052e-07, "loss": 0.01, "num_tokens": 23302790.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 395, "step_time": 56.24066119547933 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1463.0, "completions/max_terminated_length": 1463.0, "completions/mean_length": 704.796875, "completions/mean_terminated_length": 704.796875, "completions/min_length": 168.0, "completions/min_terminated_length": 168.0, "entropy": 0.9040312692523003, "epoch": 0.22448979591836735, "frac_reward_zero_std": 0.75, "grad_norm": 0.36328125, "learning_rate": 7.760770975056688e-07, "loss": 0.0066, "num_tokens": 23358881.0, "reward": 0.34375, "reward_std": 0.6950790882110596, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 396, "step_time": 50.88810185715556 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 616.921875, "completions/mean_terminated_length": 616.921875, "completions/min_length": 24.0, "completions/min_terminated_length": 24.0, "entropy": 0.7749507799744606, "epoch": 0.22505668934240364, "frac_reward_zero_std": 0.625, "grad_norm": 0.54296875, "learning_rate": 7.755102040816326e-07, "loss": -0.0083, "num_tokens": 23407412.0, "reward": 0.1796875, "reward_std": 0.5589615106582642, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 397, "step_time": 52.4311613375321 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1303.0, "completions/max_terminated_length": 1303.0, "completions/mean_length": 745.265625, "completions/mean_terminated_length": 745.265625, "completions/min_length": 319.0, "completions/min_terminated_length": 319.0, "entropy": 1.037154708057642, "epoch": 0.2256235827664399, "frac_reward_zero_std": 0.625, "grad_norm": 0.51171875, "learning_rate": 7.749433106575963e-07, "loss": -0.005, "num_tokens": 23467309.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 398, "step_time": 44.10899902880192 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1357.0, "completions/max_terminated_length": 1357.0, "completions/mean_length": 698.6875, "completions/mean_terminated_length": 698.6875, "completions/min_length": 315.0, "completions/min_terminated_length": 315.0, "entropy": 0.9808155223727226, "epoch": 0.2261904761904762, "frac_reward_zero_std": 0.5, "grad_norm": 0.578125, "learning_rate": 7.743764172335601e-07, "loss": 0.0022, "num_tokens": 23522145.0, "reward": 0.3671875, "reward_std": 0.7676090598106384, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 399, "step_time": 46.3060958776623 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 683.59375, "completions/mean_terminated_length": 683.59375, "completions/min_length": 66.0, "completions/min_terminated_length": 66.0, "entropy": 0.6724383383989334, "epoch": 0.22675736961451248, "frac_reward_zero_std": 0.5, "grad_norm": 0.55859375, "learning_rate": 7.738095238095238e-07, "loss": 0.0142, "num_tokens": 23574207.0, "reward": 0.3125, "reward_std": 0.6986381411552429, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.09375, "rewards/format_reward_func/std": 0.19669894874095917, "step": 400, "step_time": 40.15125029999763 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1227.0, "completions/max_terminated_length": 1227.0, "completions/mean_length": 771.9375, "completions/mean_terminated_length": 771.9375, "completions/min_length": 370.0, "completions/min_terminated_length": 370.0, "entropy": 0.8765030633658171, "epoch": 0.22732426303854875, "frac_reward_zero_std": 0.75, "grad_norm": 0.328125, "learning_rate": 7.732426303854876e-07, "loss": 0.0157, "num_tokens": 23635267.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 401, "step_time": 42.207340964116156 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2591.0, "completions/max_terminated_length": 2591.0, "completions/mean_length": 801.46875, "completions/mean_terminated_length": 801.46875, "completions/min_length": 310.0, "completions/min_terminated_length": 310.0, "entropy": 0.8337623011320829, "epoch": 0.22789115646258504, "frac_reward_zero_std": 0.625, "grad_norm": 0.447265625, "learning_rate": 7.726757369614512e-07, "loss": 0.0211, "num_tokens": 23699601.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 402, "step_time": 93.77474882733077 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1147.0, "completions/max_terminated_length": 1147.0, "completions/mean_length": 650.578125, "completions/mean_terminated_length": 650.578125, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "entropy": 0.7326683849096298, "epoch": 0.22845804988662133, "frac_reward_zero_std": 0.25, "grad_norm": 0.6796875, "learning_rate": 7.721088435374149e-07, "loss": -0.0558, "num_tokens": 23750414.0, "reward": 0.3359375, "reward_std": 0.7185450792312622, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 403, "step_time": 38.47740055248141 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1746.0, "completions/max_terminated_length": 1746.0, "completions/mean_length": 761.71875, "completions/mean_terminated_length": 761.71875, "completions/min_length": 338.0, "completions/min_terminated_length": 338.0, "entropy": 0.6371756438165903, "epoch": 0.2290249433106576, "frac_reward_zero_std": 0.625, "grad_norm": 0.404296875, "learning_rate": 7.715419501133786e-07, "loss": 0.005, "num_tokens": 23810716.0, "reward": 0.1796875, "reward_std": 0.5298036932945251, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 404, "step_time": 61.017441344447434 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1338.0, "completions/max_terminated_length": 1338.0, "completions/mean_length": 607.6875, "completions/mean_terminated_length": 607.6875, "completions/min_length": 9.0, "completions/min_terminated_length": 9.0, "entropy": 1.2175236158072948, "epoch": 0.22959183673469388, "frac_reward_zero_std": 0.625, "grad_norm": 0.71484375, "learning_rate": 7.709750566893424e-07, "loss": -0.0044, "num_tokens": 23861824.0, "reward": 0.09375, "reward_std": 0.36596253514289856, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 405, "step_time": 45.624282780103385 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 785.640625, "completions/mean_terminated_length": 785.640625, "completions/min_length": 238.0, "completions/min_terminated_length": 238.0, "entropy": 0.659934913739562, "epoch": 0.23015873015873015, "frac_reward_zero_std": 0.5, "grad_norm": 0.49609375, "learning_rate": 7.704081632653061e-07, "loss": 0.004, "num_tokens": 23920985.0, "reward": 0.171875, "reward_std": 0.528690755367279, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 406, "step_time": 48.06194389145821 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1513.0, "completions/max_terminated_length": 1513.0, "completions/mean_length": 768.203125, "completions/mean_terminated_length": 768.203125, "completions/min_length": 327.0, "completions/min_terminated_length": 327.0, "entropy": 0.7393356766551733, "epoch": 0.23072562358276644, "frac_reward_zero_std": 0.75, "grad_norm": 0.380859375, "learning_rate": 7.698412698412699e-07, "loss": 0.0011, "num_tokens": 23981334.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 407, "step_time": 51.93799236509949 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 791.578125, "completions/mean_terminated_length": 791.578125, "completions/min_length": 302.0, "completions/min_terminated_length": 302.0, "entropy": 0.6468864362686872, "epoch": 0.23129251700680273, "frac_reward_zero_std": 0.75, "grad_norm": 0.36328125, "learning_rate": 7.692743764172335e-07, "loss": 0.0047, "num_tokens": 24041171.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 408, "step_time": 56.41847629472613 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1442.0, "completions/max_terminated_length": 1442.0, "completions/mean_length": 863.390625, "completions/mean_terminated_length": 863.390625, "completions/min_length": 394.0, "completions/min_terminated_length": 394.0, "entropy": 0.6300061997026205, "epoch": 0.231859410430839, "frac_reward_zero_std": 0.75, "grad_norm": 0.333984375, "learning_rate": 7.687074829931972e-07, "loss": 0.0041, "num_tokens": 24107764.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 409, "step_time": 49.124212984927 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1315.0, "completions/max_terminated_length": 1315.0, "completions/mean_length": 682.546875, "completions/mean_terminated_length": 682.546875, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "entropy": 0.8103759475052357, "epoch": 0.23242630385487528, "frac_reward_zero_std": 0.25, "grad_norm": 0.76953125, "learning_rate": 7.68140589569161e-07, "loss": 0.0294, "num_tokens": 24164863.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 410, "step_time": 47.03771211300045 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1504.0, "completions/max_terminated_length": 1504.0, "completions/mean_length": 867.53125, "completions/mean_terminated_length": 867.53125, "completions/min_length": 309.0, "completions/min_terminated_length": 309.0, "entropy": 0.6705813817679882, "epoch": 0.23299319727891157, "frac_reward_zero_std": 0.5, "grad_norm": 0.5, "learning_rate": 7.675736961451246e-07, "loss": 0.0019, "num_tokens": 24231129.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 411, "step_time": 51.962798967026174 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1661.0, "completions/max_terminated_length": 1661.0, "completions/mean_length": 765.484375, "completions/mean_terminated_length": 765.484375, "completions/min_length": 289.0, "completions/min_terminated_length": 289.0, "entropy": 0.7321729604154825, "epoch": 0.23356009070294784, "frac_reward_zero_std": 0.625, "grad_norm": 0.478515625, "learning_rate": 7.670068027210884e-07, "loss": 0.027, "num_tokens": 24289264.0, "reward": 0.3046875, "reward_std": 0.7949103116989136, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 412, "step_time": 56.802216346375644 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 775.515625, "completions/mean_terminated_length": 775.515625, "completions/min_length": 313.0, "completions/min_terminated_length": 313.0, "entropy": 0.8015652969479561, "epoch": 0.23412698412698413, "frac_reward_zero_std": 0.5, "grad_norm": 0.5078125, "learning_rate": 7.664399092970521e-07, "loss": 0.0053, "num_tokens": 24347777.0, "reward": 0.3203125, "reward_std": 0.6259417533874512, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.1015625, "rewards/format_reward_func/std": 0.20275264978408813, "step": 413, "step_time": 50.851755070500076 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1296.0, "completions/max_terminated_length": 1296.0, "completions/mean_length": 812.984375, "completions/mean_terminated_length": 812.984375, "completions/min_length": 350.0, "completions/min_terminated_length": 350.0, "entropy": 0.5214714016765356, "epoch": 0.23469387755102042, "frac_reward_zero_std": 0.5, "grad_norm": 0.455078125, "learning_rate": 7.658730158730159e-07, "loss": 0.0112, "num_tokens": 24409896.0, "reward": 0.390625, "reward_std": 0.7531349658966064, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 414, "step_time": 44.565095531754196 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1319.0, "completions/max_terminated_length": 1319.0, "completions/mean_length": 730.359375, "completions/mean_terminated_length": 730.359375, "completions/min_length": 394.0, "completions/min_terminated_length": 394.0, "entropy": 0.8997184168547392, "epoch": 0.23526077097505668, "frac_reward_zero_std": 0.625, "grad_norm": 0.404296875, "learning_rate": 7.653061224489795e-07, "loss": 0.012, "num_tokens": 24467431.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 415, "step_time": 45.2768179923296 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1396.0, "completions/max_terminated_length": 1396.0, "completions/mean_length": 712.03125, "completions/mean_terminated_length": 712.03125, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "entropy": 0.8183233700692654, "epoch": 0.23582766439909297, "frac_reward_zero_std": 0.75, "grad_norm": 0.380859375, "learning_rate": 7.647392290249433e-07, "loss": -0.0093, "num_tokens": 24525633.0, "reward": 0.203125, "reward_std": 0.6529466509819031, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 416, "step_time": 48.1918881079182 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 726.4375, "completions/mean_terminated_length": 726.4375, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "entropy": 0.8129748981446028, "epoch": 0.23639455782312926, "frac_reward_zero_std": 0.75, "grad_norm": 0.296875, "learning_rate": 7.64172335600907e-07, "loss": -0.0106, "num_tokens": 24584173.0, "reward": 0.2109375, "reward_std": 0.576006293296814, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 417, "step_time": 48.32587588019669 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1301.0, "completions/max_terminated_length": 1301.0, "completions/mean_length": 623.875, "completions/mean_terminated_length": 623.875, "completions/min_length": 149.0, "completions/min_terminated_length": 149.0, "entropy": 0.9808018021285534, "epoch": 0.23696145124716553, "frac_reward_zero_std": 0.75, "grad_norm": 0.453125, "learning_rate": 7.636054421768708e-07, "loss": -0.0203, "num_tokens": 24635549.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 418, "step_time": 44.927674242295325 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1632.0, "completions/max_terminated_length": 1632.0, "completions/mean_length": 732.84375, "completions/mean_terminated_length": 732.84375, "completions/min_length": 321.0, "completions/min_terminated_length": 321.0, "entropy": 0.6691880654543638, "epoch": 0.23752834467120182, "frac_reward_zero_std": 0.625, "grad_norm": 0.419921875, "learning_rate": 7.630385487528344e-07, "loss": -0.0094, "num_tokens": 24692691.0, "reward": 0.1484375, "reward_std": 0.47709178924560547, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 419, "step_time": 55.825158819556236 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 767.515625, "completions/mean_terminated_length": 767.515625, "completions/min_length": 414.0, "completions/min_terminated_length": 414.0, "entropy": 0.8011331092566252, "epoch": 0.23809523809523808, "frac_reward_zero_std": 0.75, "grad_norm": 0.3828125, "learning_rate": 7.624716553287982e-07, "loss": -0.0063, "num_tokens": 24752844.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 420, "step_time": 53.255850152112544 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1497.0, "completions/max_terminated_length": 1497.0, "completions/mean_length": 771.734375, "completions/mean_terminated_length": 771.734375, "completions/min_length": 326.0, "completions/min_terminated_length": 326.0, "entropy": 0.8095507547259331, "epoch": 0.23866213151927437, "frac_reward_zero_std": 0.875, "grad_norm": 0.25, "learning_rate": 7.619047619047618e-07, "loss": -0.0171, "num_tokens": 24812779.0, "reward": 0.0390625, "reward_std": 0.13524484634399414, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 421, "step_time": 50.773611281998456 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 681.765625, "completions/mean_terminated_length": 681.765625, "completions/min_length": 376.0, "completions/min_terminated_length": 376.0, "entropy": 0.7349203620105982, "epoch": 0.23922902494331066, "frac_reward_zero_std": 0.75, "grad_norm": 0.3515625, "learning_rate": 7.613378684807256e-07, "loss": -0.0037, "num_tokens": 24865948.0, "reward": 0.28125, "reward_std": 0.7007648944854736, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 422, "step_time": 39.631040232256055 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1510.0, "completions/max_terminated_length": 1510.0, "completions/mean_length": 722.5, "completions/mean_terminated_length": 722.5, "completions/min_length": 354.0, "completions/min_terminated_length": 354.0, "entropy": 0.7520661912858486, "epoch": 0.23979591836734693, "frac_reward_zero_std": 0.5, "grad_norm": 0.51171875, "learning_rate": 7.607709750566893e-07, "loss": -0.0196, "num_tokens": 24922852.0, "reward": 0.3046875, "reward_std": 0.664384126663208, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 423, "step_time": 51.616779503412545 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1843.0, "completions/max_terminated_length": 1843.0, "completions/mean_length": 719.765625, "completions/mean_terminated_length": 719.765625, "completions/min_length": 284.0, "completions/min_terminated_length": 284.0, "entropy": 0.7724943663924932, "epoch": 0.24036281179138322, "frac_reward_zero_std": 0.75, "grad_norm": 0.41015625, "learning_rate": 7.602040816326531e-07, "loss": -0.0035, "num_tokens": 24979133.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 424, "step_time": 62.50873214751482 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1671.0, "completions/max_terminated_length": 1671.0, "completions/mean_length": 736.859375, "completions/mean_terminated_length": 736.859375, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "entropy": 0.8050760552287102, "epoch": 0.2409297052154195, "frac_reward_zero_std": 0.625, "grad_norm": 0.47265625, "learning_rate": 7.596371882086168e-07, "loss": -0.027, "num_tokens": 25037660.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 425, "step_time": 57.295719363726676 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1315.0, "completions/max_terminated_length": 1315.0, "completions/mean_length": 641.71875, "completions/mean_terminated_length": 641.71875, "completions/min_length": 271.0, "completions/min_terminated_length": 271.0, "entropy": 1.2077321372926235, "epoch": 0.24149659863945577, "frac_reward_zero_std": 0.625, "grad_norm": 0.50390625, "learning_rate": 7.590702947845806e-07, "loss": -0.0069, "num_tokens": 25092130.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 426, "step_time": 46.93154688831419 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1814.0, "completions/max_terminated_length": 1814.0, "completions/mean_length": 727.703125, "completions/mean_terminated_length": 727.703125, "completions/min_length": 254.0, "completions/min_terminated_length": 254.0, "entropy": 0.9418045226484537, "epoch": 0.24206349206349206, "frac_reward_zero_std": 0.75, "grad_norm": 0.34375, "learning_rate": 7.585034013605441e-07, "loss": 0.0003, "num_tokens": 25150535.0, "reward": 0.15625, "reward_std": 0.4787135720252991, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 427, "step_time": 63.36056375131011 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1510.0, "completions/max_terminated_length": 1510.0, "completions/mean_length": 753.890625, "completions/mean_terminated_length": 753.890625, "completions/min_length": 206.0, "completions/min_terminated_length": 206.0, "entropy": 0.6319617219269276, "epoch": 0.24263038548752835, "frac_reward_zero_std": 0.75, "grad_norm": 0.34375, "learning_rate": 7.579365079365078e-07, "loss": -0.0097, "num_tokens": 25209680.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 428, "step_time": 52.11038209684193 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1242.0, "completions/max_terminated_length": 1242.0, "completions/mean_length": 728.234375, "completions/mean_terminated_length": 728.234375, "completions/min_length": 4.0, "completions/min_terminated_length": 4.0, "entropy": 0.7340629417449236, "epoch": 0.24319727891156462, "frac_reward_zero_std": 0.875, "grad_norm": 0.275390625, "learning_rate": 7.573696145124716e-07, "loss": 0.0009, "num_tokens": 25265807.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 429, "step_time": 41.60768271610141 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1277.0, "completions/max_terminated_length": 1277.0, "completions/mean_length": 741.234375, "completions/mean_terminated_length": 741.234375, "completions/min_length": 356.0, "completions/min_terminated_length": 356.0, "entropy": 0.8030328117311001, "epoch": 0.2437641723356009, "frac_reward_zero_std": 0.5, "grad_norm": 0.5703125, "learning_rate": 7.568027210884353e-07, "loss": -0.0157, "num_tokens": 25325862.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 430, "step_time": 43.42780246119946 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 676.265625, "completions/mean_terminated_length": 676.265625, "completions/min_length": 220.0, "completions/min_terminated_length": 220.0, "entropy": 0.8896284252405167, "epoch": 0.2443310657596372, "frac_reward_zero_std": 0.25, "grad_norm": 0.69140625, "learning_rate": 7.562358276643991e-07, "loss": -0.0126, "num_tokens": 25378679.0, "reward": 0.4375, "reward_std": 0.8042545318603516, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 431, "step_time": 46.09310553781688 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1472.0, "completions/max_terminated_length": 1472.0, "completions/mean_length": 705.390625, "completions/mean_terminated_length": 705.390625, "completions/min_length": 258.0, "completions/min_terminated_length": 258.0, "entropy": 0.7383911740034819, "epoch": 0.24489795918367346, "frac_reward_zero_std": 0.625, "grad_norm": 0.46875, "learning_rate": 7.556689342403628e-07, "loss": -0.0016, "num_tokens": 25433072.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 432, "step_time": 49.770350963808596 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1634.0, "completions/max_terminated_length": 1634.0, "completions/mean_length": 823.46875, "completions/mean_terminated_length": 823.46875, "completions/min_length": 367.0, "completions/min_terminated_length": 367.0, "entropy": 0.6049559805542231, "epoch": 0.24546485260770975, "frac_reward_zero_std": 0.5, "grad_norm": 0.44921875, "learning_rate": 7.551020408163265e-07, "loss": 0.0026, "num_tokens": 25497070.0, "reward": 0.328125, "reward_std": 0.6970388293266296, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 433, "step_time": 57.097911885008216 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1264.0, "completions/max_terminated_length": 1264.0, "completions/mean_length": 715.09375, "completions/mean_terminated_length": 715.09375, "completions/min_length": 9.0, "completions/min_terminated_length": 9.0, "entropy": 1.0062689557671547, "epoch": 0.24603174603174602, "frac_reward_zero_std": 0.5, "grad_norm": 0.64453125, "learning_rate": 7.545351473922901e-07, "loss": -0.0195, "num_tokens": 25562236.0, "reward": 0.2578125, "reward_std": 0.6784212589263916, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 434, "step_time": 46.80837841890752 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1566.0, "completions/max_terminated_length": 1566.0, "completions/mean_length": 745.890625, "completions/mean_terminated_length": 745.890625, "completions/min_length": 332.0, "completions/min_terminated_length": 332.0, "entropy": 0.8993903696537018, "epoch": 0.2465986394557823, "frac_reward_zero_std": 0.375, "grad_norm": 0.6484375, "learning_rate": 7.539682539682539e-07, "loss": 0.005, "num_tokens": 25619405.0, "reward": 0.2734375, "reward_std": 0.6542273163795471, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 435, "step_time": 53.25824845582247 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1661.0, "completions/max_terminated_length": 1661.0, "completions/mean_length": 792.34375, "completions/mean_terminated_length": 792.34375, "completions/min_length": 216.0, "completions/min_terminated_length": 216.0, "entropy": 0.8956113327294588, "epoch": 0.2471655328798186, "frac_reward_zero_std": 0.875, "grad_norm": 0.32421875, "learning_rate": 7.534013605442176e-07, "loss": 0.0042, "num_tokens": 25679243.0, "reward": 0.0390625, "reward_std": 0.13524484634399414, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 436, "step_time": 56.02351733762771 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1229.0, "completions/max_terminated_length": 1229.0, "completions/mean_length": 709.609375, "completions/mean_terminated_length": 709.609375, "completions/min_length": 325.0, "completions/min_terminated_length": 325.0, "entropy": 0.8032096195966005, "epoch": 0.24773242630385486, "frac_reward_zero_std": 0.5, "grad_norm": 0.5078125, "learning_rate": 7.528344671201814e-07, "loss": 0.0304, "num_tokens": 25734386.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 437, "step_time": 42.16178113408387 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 850.6875, "completions/mean_terminated_length": 850.6875, "completions/min_length": 300.0, "completions/min_terminated_length": 300.0, "entropy": 0.6397589351981878, "epoch": 0.24829931972789115, "frac_reward_zero_std": 0.5, "grad_norm": 0.43359375, "learning_rate": 7.522675736961451e-07, "loss": -0.0119, "num_tokens": 25798718.0, "reward": 0.359375, "reward_std": 0.7263124585151672, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 438, "step_time": 42.7511393353343 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1236.0, "completions/max_terminated_length": 1236.0, "completions/mean_length": 618.953125, "completions/mean_terminated_length": 618.953125, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "entropy": 0.6891420613974333, "epoch": 0.24886621315192745, "frac_reward_zero_std": 0.375, "grad_norm": 0.58984375, "learning_rate": 7.517006802721088e-07, "loss": 0.0294, "num_tokens": 25849811.0, "reward": 0.3984375, "reward_std": 0.7516103982925415, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 439, "step_time": 42.715612224303186 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1131.0, "completions/max_terminated_length": 1131.0, "completions/mean_length": 618.09375, "completions/mean_terminated_length": 618.09375, "completions/min_length": 326.0, "completions/min_terminated_length": 326.0, "entropy": 0.9189383015036583, "epoch": 0.2494331065759637, "frac_reward_zero_std": 0.625, "grad_norm": 0.46484375, "learning_rate": 7.511337868480725e-07, "loss": 0.0251, "num_tokens": 25899305.0, "reward": 0.2890625, "reward_std": 0.6655030846595764, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 440, "step_time": 39.003290640190244 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1666.0, "completions/max_terminated_length": 1666.0, "completions/mean_length": 823.5, "completions/mean_terminated_length": 823.5, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "entropy": 0.7929299920797348, "epoch": 0.25, "frac_reward_zero_std": 0.875, "grad_norm": 0.255859375, "learning_rate": 7.505668934240363e-07, "loss": 0.0022, "num_tokens": 25963049.0, "reward": 0.046875, "reward_std": 0.14689241349697113, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 441, "step_time": 58.207637903280556 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2147.0, "completions/max_terminated_length": 2147.0, "completions/mean_length": 761.40625, "completions/mean_terminated_length": 761.40625, "completions/min_length": 308.0, "completions/min_terminated_length": 308.0, "entropy": 0.7989441230893135, "epoch": 0.25056689342403626, "frac_reward_zero_std": 0.625, "grad_norm": 0.58203125, "learning_rate": 7.5e-07, "loss": -0.0063, "num_tokens": 26020443.0, "reward": 0.265625, "reward_std": 0.6544642448425293, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 442, "step_time": 74.96575433388352 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1505.0, "completions/max_terminated_length": 1505.0, "completions/mean_length": 830.8125, "completions/mean_terminated_length": 830.8125, "completions/min_length": 457.0, "completions/min_terminated_length": 457.0, "entropy": 0.8002884350717068, "epoch": 0.2511337868480726, "frac_reward_zero_std": 0.625, "grad_norm": 0.40234375, "learning_rate": 7.494331065759637e-07, "loss": 0.0268, "num_tokens": 26087447.0, "reward": 0.15625, "reward_std": 0.5261042714118958, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 443, "step_time": 52.967568409629166 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1483.0, "completions/max_terminated_length": 1483.0, "completions/mean_length": 780.5625, "completions/mean_terminated_length": 780.5625, "completions/min_length": 4.0, "completions/min_terminated_length": 4.0, "entropy": 0.8361520990729332, "epoch": 0.25170068027210885, "frac_reward_zero_std": 0.625, "grad_norm": 0.474609375, "learning_rate": 7.488662131519274e-07, "loss": -0.0088, "num_tokens": 26147027.0, "reward": 0.125, "reward_std": 0.417855441570282, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 444, "step_time": 50.80671252217144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 736.25, "completions/mean_terminated_length": 736.25, "completions/min_length": 212.0, "completions/min_terminated_length": 212.0, "entropy": 0.7818868961185217, "epoch": 0.2522675736961451, "frac_reward_zero_std": 0.625, "grad_norm": 0.43359375, "learning_rate": 7.482993197278912e-07, "loss": 0.0169, "num_tokens": 26206291.0, "reward": 0.4609375, "reward_std": 0.7983350157737732, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 445, "step_time": 53.251479097642004 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1633.0, "completions/max_terminated_length": 1633.0, "completions/mean_length": 726.9375, "completions/mean_terminated_length": 726.9375, "completions/min_length": 369.0, "completions/min_terminated_length": 369.0, "entropy": 0.8610232416540384, "epoch": 0.2528344671201814, "frac_reward_zero_std": 0.25, "grad_norm": 0.7421875, "learning_rate": 7.477324263038548e-07, "loss": 0.0077, "num_tokens": 26262095.0, "reward": 0.3671875, "reward_std": 0.725073516368866, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 446, "step_time": 55.98789916560054 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 606.203125, "completions/mean_terminated_length": 606.203125, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "entropy": 0.7205122336745262, "epoch": 0.2534013605442177, "frac_reward_zero_std": 0.375, "grad_norm": 0.68359375, "learning_rate": 7.471655328798186e-07, "loss": -0.036, "num_tokens": 26314428.0, "reward": 0.4375, "reward_std": 0.7943251132965088, "rewards/accuracy_reward_func/mean": 0.375, "rewards/accuracy_reward_func/std": 0.7867957949638367, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 447, "step_time": 41.76088255736977 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1565.0, "completions/max_terminated_length": 1565.0, "completions/mean_length": 799.953125, "completions/mean_terminated_length": 799.953125, "completions/min_length": 454.0, "completions/min_terminated_length": 454.0, "entropy": 0.833087008446455, "epoch": 0.25396825396825395, "frac_reward_zero_std": 0.375, "grad_norm": 0.64453125, "learning_rate": 7.465986394557823e-07, "loss": 0.0343, "num_tokens": 26375929.0, "reward": 0.2890625, "reward_std": 0.6655030846595764, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 448, "step_time": 53.082594824023545 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1251.0, "completions/max_terminated_length": 1251.0, "completions/mean_length": 723.78125, "completions/mean_terminated_length": 723.78125, "completions/min_length": 257.0, "completions/min_terminated_length": 257.0, "entropy": 0.8315397277474403, "epoch": 0.2545351473922903, "frac_reward_zero_std": 0.25, "grad_norm": 0.66015625, "learning_rate": 7.46031746031746e-07, "loss": -0.027, "num_tokens": 26433067.0, "reward": 0.3515625, "reward_std": 0.7274640202522278, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 449, "step_time": 42.740751163102686 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1752.0, "completions/max_terminated_length": 1752.0, "completions/mean_length": 851.140625, "completions/mean_terminated_length": 851.140625, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "entropy": 0.6886486634612083, "epoch": 0.25510204081632654, "frac_reward_zero_std": 0.875, "grad_norm": 0.2578125, "learning_rate": 7.454648526077098e-07, "loss": 0.0061, "num_tokens": 26498692.0, "reward": 0.125, "reward_std": 0.4542567729949951, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 450, "step_time": 60.04871977586299 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1121.0, "completions/max_terminated_length": 1121.0, "completions/mean_length": 693.03125, "completions/mean_terminated_length": 693.03125, "completions/min_length": 231.0, "completions/min_terminated_length": 231.0, "entropy": 0.6884783115237951, "epoch": 0.2556689342403628, "frac_reward_zero_std": 0.375, "grad_norm": 0.54296875, "learning_rate": 7.448979591836734e-07, "loss": 0.0503, "num_tokens": 26552750.0, "reward": 0.296875, "reward_std": 0.6649904847145081, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 451, "step_time": 38.398378430865705 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2723.0, "completions/max_terminated_length": 2723.0, "completions/mean_length": 824.140625, "completions/mean_terminated_length": 824.140625, "completions/min_length": 187.0, "completions/min_terminated_length": 187.0, "entropy": 0.5746772699058056, "epoch": 0.2562358276643991, "frac_reward_zero_std": 0.375, "grad_norm": 0.5, "learning_rate": 7.443310657596371e-07, "loss": -0.016, "num_tokens": 26617207.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 452, "step_time": 100.30535710975528 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 735.796875, "completions/mean_terminated_length": 735.796875, "completions/min_length": 348.0, "completions/min_terminated_length": 348.0, "entropy": 0.822858165949583, "epoch": 0.2568027210884354, "frac_reward_zero_std": 0.625, "grad_norm": 0.470703125, "learning_rate": 7.437641723356008e-07, "loss": 0.0014, "num_tokens": 26674298.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 453, "step_time": 51.56850883085281 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1242.0, "completions/max_terminated_length": 1242.0, "completions/mean_length": 722.234375, "completions/mean_terminated_length": 722.234375, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "entropy": 0.8533442914485931, "epoch": 0.25736961451247165, "frac_reward_zero_std": 0.625, "grad_norm": 0.515625, "learning_rate": 7.431972789115646e-07, "loss": -0.0024, "num_tokens": 26730257.0, "reward": 0.1484375, "reward_std": 0.49344661831855774, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 454, "step_time": 41.675769814290106 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1861.0, "completions/max_terminated_length": 1861.0, "completions/mean_length": 818.953125, "completions/mean_terminated_length": 818.953125, "completions/min_length": 369.0, "completions/min_terminated_length": 369.0, "entropy": 0.8123147170990705, "epoch": 0.25793650793650796, "frac_reward_zero_std": 0.5, "grad_norm": 0.466796875, "learning_rate": 7.426303854875283e-07, "loss": 0.0174, "num_tokens": 26794046.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 455, "step_time": 64.55787387117743 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1733.0, "completions/max_terminated_length": 1733.0, "completions/mean_length": 788.671875, "completions/mean_terminated_length": 788.671875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "entropy": 0.8716447334736586, "epoch": 0.2585034013605442, "frac_reward_zero_std": 0.625, "grad_norm": 0.4609375, "learning_rate": 7.420634920634921e-07, "loss": -0.0143, "num_tokens": 26853513.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 456, "step_time": 59.41806235536933 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1368.0, "completions/max_terminated_length": 1368.0, "completions/mean_length": 806.65625, "completions/mean_terminated_length": 806.65625, "completions/min_length": 366.0, "completions/min_terminated_length": 366.0, "entropy": 0.8578362837433815, "epoch": 0.2590702947845805, "frac_reward_zero_std": 0.875, "grad_norm": 0.287109375, "learning_rate": 7.414965986394558e-07, "loss": -0.0013, "num_tokens": 26916163.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 457, "step_time": 47.249078824184835 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 809.015625, "completions/mean_terminated_length": 809.015625, "completions/min_length": 433.0, "completions/min_terminated_length": 433.0, "entropy": 0.8555043283849955, "epoch": 0.25963718820861675, "frac_reward_zero_std": 0.5, "grad_norm": 0.49609375, "learning_rate": 7.409297052154194e-07, "loss": -0.0103, "num_tokens": 26980372.0, "reward": 0.2265625, "reward_std": 0.5904644727706909, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 458, "step_time": 45.47941083367914 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 731.34375, "completions/mean_terminated_length": 731.34375, "completions/min_length": 476.0, "completions/min_terminated_length": 476.0, "entropy": 0.8994519170373678, "epoch": 0.2602040816326531, "frac_reward_zero_std": 0.625, "grad_norm": 0.482421875, "learning_rate": 7.403628117913831e-07, "loss": 0.0145, "num_tokens": 27037674.0, "reward": 0.1875, "reward_std": 0.5739033818244934, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 459, "step_time": 39.012475188821554 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2292.0, "completions/max_terminated_length": 2292.0, "completions/mean_length": 607.640625, "completions/mean_terminated_length": 607.640625, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "entropy": 0.6151527259498835, "epoch": 0.26077097505668934, "frac_reward_zero_std": 0.625, "grad_norm": 0.609375, "learning_rate": 7.397959183673469e-07, "loss": -0.0369, "num_tokens": 27089427.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 460, "step_time": 82.16680842358619 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 692.390625, "completions/mean_terminated_length": 692.390625, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "entropy": 0.6963042225688696, "epoch": 0.2613378684807256, "frac_reward_zero_std": 0.5, "grad_norm": 0.56640625, "learning_rate": 7.392290249433106e-07, "loss": -0.0166, "num_tokens": 27142900.0, "reward": 0.3046875, "reward_std": 0.664384126663208, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 461, "step_time": 40.6895647700876 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1794.0, "completions/max_terminated_length": 1794.0, "completions/mean_length": 783.640625, "completions/mean_terminated_length": 783.640625, "completions/min_length": 263.0, "completions/min_terminated_length": 263.0, "entropy": 0.7272162567824125, "epoch": 0.2619047619047619, "frac_reward_zero_std": 0.75, "grad_norm": 0.38671875, "learning_rate": 7.386621315192744e-07, "loss": -0.0249, "num_tokens": 27203557.0, "reward": 0.1328125, "reward_std": 0.4734388589859009, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 462, "step_time": 60.710756020620465 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1483.0, "completions/max_terminated_length": 1483.0, "completions/mean_length": 762.0625, "completions/mean_terminated_length": 762.0625, "completions/min_length": 312.0, "completions/min_terminated_length": 312.0, "entropy": 0.8299924489110708, "epoch": 0.2624716553287982, "frac_reward_zero_std": 0.5, "grad_norm": 0.5390625, "learning_rate": 7.380952380952381e-07, "loss": 0.029, "num_tokens": 27262929.0, "reward": 0.2734375, "reward_std": 0.6542273163795471, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 463, "step_time": 50.03462992142886 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 708.78125, "completions/mean_terminated_length": 708.78125, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "entropy": 0.8047873191535473, "epoch": 0.26303854875283444, "frac_reward_zero_std": 0.625, "grad_norm": 0.46484375, "learning_rate": 7.375283446712018e-07, "loss": 0.0096, "num_tokens": 27317939.0, "reward": 0.25, "reward_std": 0.6172134280204773, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 464, "step_time": 42.26102858968079 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1442.0, "completions/max_terminated_length": 1442.0, "completions/mean_length": 829.0, "completions/mean_terminated_length": 829.0, "completions/min_length": 457.0, "completions/min_terminated_length": 457.0, "entropy": 0.7290856074541807, "epoch": 0.26360544217687076, "frac_reward_zero_std": 0.75, "grad_norm": 0.333984375, "learning_rate": 7.369614512471655e-07, "loss": 0.0049, "num_tokens": 27382139.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 465, "step_time": 48.702748977579176 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1627.0, "completions/mean_length": 806.71875, "completions/mean_terminated_length": 754.5079956054688, "completions/min_length": 339.0, "completions/min_terminated_length": 339.0, "entropy": 0.7747468892484903, "epoch": 0.264172335600907, "frac_reward_zero_std": 0.5, "grad_norm": 0.4921875, "learning_rate": 7.363945578231292e-07, "loss": -0.0051, "num_tokens": 27444345.0, "reward": 0.1875, "reward_std": 0.5455447435379028, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 466, "step_time": 167.59687008708715 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1456.0, "completions/max_terminated_length": 1456.0, "completions/mean_length": 695.59375, "completions/mean_terminated_length": 695.59375, "completions/min_length": 179.0, "completions/min_terminated_length": 179.0, "entropy": 0.8323264606297016, "epoch": 0.2647392290249433, "frac_reward_zero_std": 0.375, "grad_norm": 0.66796875, "learning_rate": 7.358276643990929e-07, "loss": -0.0008, "num_tokens": 27501527.0, "reward": 0.28125, "reward_std": 0.71200031042099, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 467, "step_time": 49.73436775896698 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1436.0, "completions/mean_length": 799.671875, "completions/mean_terminated_length": 747.3492431640625, "completions/min_length": 418.0, "completions/min_terminated_length": 418.0, "entropy": 0.864380856975913, "epoch": 0.2653061224489796, "frac_reward_zero_std": 0.5, "grad_norm": 0.54296875, "learning_rate": 7.352607709750567e-07, "loss": -0.0071, "num_tokens": 27565346.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 468, "step_time": 168.72100699786097 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1620.0, "completions/max_terminated_length": 1620.0, "completions/mean_length": 656.609375, "completions/mean_terminated_length": 656.609375, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "entropy": 1.0048782378435135, "epoch": 0.26587301587301587, "frac_reward_zero_std": 0.25, "grad_norm": 0.734375, "learning_rate": 7.346938775510204e-07, "loss": 0.0027, "num_tokens": 27617289.0, "reward": 0.421875, "reward_std": 0.7777600884437561, "rewards/accuracy_reward_func/mean": 0.375, "rewards/accuracy_reward_func/std": 0.7867957949638367, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 469, "step_time": 54.89007071964443 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1607.0, "completions/max_terminated_length": 1607.0, "completions/mean_length": 850.734375, "completions/mean_terminated_length": 850.734375, "completions/min_length": 302.0, "completions/min_terminated_length": 302.0, "entropy": 0.6241070311516523, "epoch": 0.26643990929705214, "frac_reward_zero_std": 0.75, "grad_norm": 0.255859375, "learning_rate": 7.34126984126984e-07, "loss": -0.0052, "num_tokens": 27683528.0, "reward": 0.15625, "reward_std": 0.4443204402923584, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 470, "step_time": 54.522400417365134 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1613.0, "completions/max_terminated_length": 1613.0, "completions/mean_length": 817.828125, "completions/mean_terminated_length": 817.828125, "completions/min_length": 354.0, "completions/min_terminated_length": 354.0, "entropy": 0.6263278722763062, "epoch": 0.26700680272108845, "frac_reward_zero_std": 0.25, "grad_norm": 0.5703125, "learning_rate": 7.335600907029478e-07, "loss": -0.0038, "num_tokens": 27745725.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 471, "step_time": 54.59293780755252 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1464.0, "completions/max_terminated_length": 1464.0, "completions/mean_length": 676.078125, "completions/mean_terminated_length": 676.078125, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "entropy": 0.9488864932209253, "epoch": 0.2675736961451247, "frac_reward_zero_std": 0.625, "grad_norm": 0.453125, "learning_rate": 7.329931972789115e-07, "loss": 0.0229, "num_tokens": 27800218.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 472, "step_time": 50.35415969602764 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 610.6875, "completions/mean_terminated_length": 610.6875, "completions/min_length": 348.0, "completions/min_terminated_length": 348.0, "entropy": 0.9655479695647955, "epoch": 0.268140589569161, "frac_reward_zero_std": 0.5, "grad_norm": 0.58203125, "learning_rate": 7.324263038548753e-07, "loss": 0.0249, "num_tokens": 27849822.0, "reward": 0.4140625, "reward_std": 0.8191879987716675, "rewards/accuracy_reward_func/mean": 0.375, "rewards/accuracy_reward_func/std": 0.7867957949638367, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 473, "step_time": 36.97484754025936 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1558.0, "completions/mean_length": 865.90625, "completions/mean_terminated_length": 814.6349487304688, "completions/min_length": 402.0, "completions/min_terminated_length": 402.0, "entropy": 0.7376750633120537, "epoch": 0.2687074829931973, "frac_reward_zero_std": 0.75, "grad_norm": 0.326171875, "learning_rate": 7.318594104308389e-07, "loss": 0.0001, "num_tokens": 27916312.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 474, "step_time": 168.47027471847832 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1392.0, "completions/max_terminated_length": 1392.0, "completions/mean_length": 762.484375, "completions/mean_terminated_length": 762.484375, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "entropy": 0.8245761692523956, "epoch": 0.26927437641723356, "frac_reward_zero_std": 0.75, "grad_norm": 0.359375, "learning_rate": 7.312925170068027e-07, "loss": 0.0077, "num_tokens": 27976319.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 475, "step_time": 47.97840517479926 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1287.0, "completions/max_terminated_length": 1287.0, "completions/mean_length": 744.515625, "completions/mean_terminated_length": 744.515625, "completions/min_length": 355.0, "completions/min_terminated_length": 355.0, "entropy": 0.8160289656370878, "epoch": 0.2698412698412698, "frac_reward_zero_std": 0.625, "grad_norm": 0.376953125, "learning_rate": 7.307256235827664e-07, "loss": -0.0182, "num_tokens": 28037056.0, "reward": 0.25, "reward_std": 0.6299408078193665, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 476, "step_time": 45.09215791430324 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1155.0, "completions/max_terminated_length": 1155.0, "completions/mean_length": 685.59375, "completions/mean_terminated_length": 685.59375, "completions/min_length": 367.0, "completions/min_terminated_length": 367.0, "entropy": 0.9145476706326008, "epoch": 0.27040816326530615, "frac_reward_zero_std": 0.5, "grad_norm": 0.59375, "learning_rate": 7.301587301587301e-07, "loss": -0.0174, "num_tokens": 28091230.0, "reward": 0.265625, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 477, "step_time": 42.135603019967675 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1386.0, "completions/max_terminated_length": 1386.0, "completions/mean_length": 745.109375, "completions/mean_terminated_length": 745.109375, "completions/min_length": 282.0, "completions/min_terminated_length": 282.0, "entropy": 0.7281346414238214, "epoch": 0.2709750566893424, "frac_reward_zero_std": 0.5, "grad_norm": 0.416015625, "learning_rate": 7.295918367346938e-07, "loss": -0.0439, "num_tokens": 28151277.0, "reward": 0.359375, "reward_std": 0.7478474378585815, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 478, "step_time": 49.57602191157639 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 777.015625, "completions/mean_terminated_length": 777.015625, "completions/min_length": 375.0, "completions/min_terminated_length": 375.0, "entropy": 0.8162975553423166, "epoch": 0.27154195011337867, "frac_reward_zero_std": 0.375, "grad_norm": 0.5546875, "learning_rate": 7.290249433106576e-07, "loss": -0.0057, "num_tokens": 28213142.0, "reward": 0.3125, "reward_std": 0.6986381411552429, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 479, "step_time": 51.75872884038836 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 683.921875, "completions/mean_terminated_length": 683.921875, "completions/min_length": 279.0, "completions/min_terminated_length": 279.0, "entropy": 0.7339695952832699, "epoch": 0.272108843537415, "frac_reward_zero_std": 0.5, "grad_norm": 0.490234375, "learning_rate": 7.284580498866213e-07, "loss": 0.0168, "num_tokens": 28269049.0, "reward": 0.3125, "reward_std": 0.7319250702857971, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 480, "step_time": 48.186254777014256 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2134.0, "completions/max_terminated_length": 2134.0, "completions/mean_length": 858.9375, "completions/mean_terminated_length": 858.9375, "completions/min_length": 373.0, "completions/min_terminated_length": 373.0, "entropy": 0.8460619766265154, "epoch": 0.27267573696145125, "frac_reward_zero_std": 0.625, "grad_norm": 0.435546875, "learning_rate": 7.278911564625851e-07, "loss": 0.0113, "num_tokens": 28334053.0, "reward": 0.2265625, "reward_std": 0.6167611479759216, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 481, "step_time": 74.01236078422517 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1600.0, "completions/max_terminated_length": 1600.0, "completions/mean_length": 711.5, "completions/mean_terminated_length": 711.5, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "entropy": 1.0337553285062313, "epoch": 0.2732426303854875, "frac_reward_zero_std": 0.625, "grad_norm": 0.51953125, "learning_rate": 7.273242630385487e-07, "loss": 0.0149, "num_tokens": 28390757.0, "reward": 0.1015625, "reward_std": 0.3692515790462494, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 482, "step_time": 55.5663038874045 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1871.0, "completions/max_terminated_length": 1871.0, "completions/mean_length": 743.484375, "completions/mean_terminated_length": 743.484375, "completions/min_length": 374.0, "completions/min_terminated_length": 374.0, "entropy": 0.6883541475981474, "epoch": 0.27380952380952384, "frac_reward_zero_std": 0.375, "grad_norm": 0.578125, "learning_rate": 7.267573696145124e-07, "loss": 0.0269, "num_tokens": 28447796.0, "reward": 0.328125, "reward_std": 0.6970388293266296, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 483, "step_time": 63.777127023786306 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 683.71875, "completions/mean_terminated_length": 683.71875, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "entropy": 0.9639867581427097, "epoch": 0.2743764172335601, "frac_reward_zero_std": 0.625, "grad_norm": 0.482421875, "learning_rate": 7.261904761904761e-07, "loss": 0.0036, "num_tokens": 28501322.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 484, "step_time": 39.343465231359005 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 747.109375, "completions/mean_terminated_length": 747.109375, "completions/min_length": 311.0, "completions/min_terminated_length": 311.0, "entropy": 0.7536885496228933, "epoch": 0.27494331065759636, "frac_reward_zero_std": 0.625, "grad_norm": 0.451171875, "learning_rate": 7.256235827664399e-07, "loss": -0.0134, "num_tokens": 28560233.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 485, "step_time": 48.31962897256017 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1981.0, "completions/max_terminated_length": 1981.0, "completions/mean_length": 643.703125, "completions/mean_terminated_length": 643.703125, "completions/min_length": 282.0, "completions/min_terminated_length": 282.0, "entropy": 1.0371637865900993, "epoch": 0.2755102040816326, "frac_reward_zero_std": 0.75, "grad_norm": 0.484375, "learning_rate": 7.250566893424036e-07, "loss": 0.0181, "num_tokens": 28610966.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 486, "step_time": 68.0433338470757 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 752.03125, "completions/mean_terminated_length": 752.03125, "completions/min_length": 397.0, "completions/min_terminated_length": 397.0, "entropy": 0.9433378670364618, "epoch": 0.27607709750566894, "frac_reward_zero_std": 0.75, "grad_norm": 0.37890625, "learning_rate": 7.244897959183674e-07, "loss": -0.0137, "num_tokens": 28668464.0, "reward": 0.0625, "reward_std": 0.2745848298072815, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 487, "step_time": 45.741264135576785 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1221.0, "completions/mean_length": 745.34375, "completions/mean_terminated_length": 692.1587524414062, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "entropy": 0.9549574814736843, "epoch": 0.2766439909297052, "frac_reward_zero_std": 0.625, "grad_norm": 0.46875, "learning_rate": 7.239229024943311e-07, "loss": 0.0035, "num_tokens": 28726022.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 488, "step_time": 167.2989202560857 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2674.0, "completions/max_terminated_length": 2674.0, "completions/mean_length": 736.71875, "completions/mean_terminated_length": 736.71875, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "entropy": 0.7453363202512264, "epoch": 0.27721088435374147, "frac_reward_zero_std": 0.125, "grad_norm": 0.6796875, "learning_rate": 7.233560090702947e-07, "loss": 0.0189, "num_tokens": 28782500.0, "reward": 0.4921875, "reward_std": 0.8188852071762085, "rewards/accuracy_reward_func/mean": 0.4375, "rewards/accuracy_reward_func/std": 0.8333333730697632, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 489, "step_time": 95.1553288353607 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2410.0, "completions/max_terminated_length": 2410.0, "completions/mean_length": 697.984375, "completions/mean_terminated_length": 697.984375, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 0.8135911952704191, "epoch": 0.2777777777777778, "frac_reward_zero_std": 0.5, "grad_norm": 0.58984375, "learning_rate": 7.227891156462584e-07, "loss": 0.0218, "num_tokens": 28836923.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 490, "step_time": 85.06263784877956 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1386.0, "completions/max_terminated_length": 1386.0, "completions/mean_length": 707.46875, "completions/mean_terminated_length": 707.46875, "completions/min_length": 339.0, "completions/min_terminated_length": 339.0, "entropy": 0.7418342307209969, "epoch": 0.27834467120181405, "frac_reward_zero_std": 0.5, "grad_norm": 0.5859375, "learning_rate": 7.222222222222221e-07, "loss": 0.0243, "num_tokens": 28891929.0, "reward": 0.375, "reward_std": 0.7867957949638367, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 491, "step_time": 46.50618886575103 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 776.40625, "completions/mean_terminated_length": 776.40625, "completions/min_length": 382.0, "completions/min_terminated_length": 382.0, "entropy": 0.7166824005544186, "epoch": 0.2789115646258503, "frac_reward_zero_std": 0.5, "grad_norm": 0.52734375, "learning_rate": 7.216553287981859e-07, "loss": 0.0466, "num_tokens": 28952427.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 492, "step_time": 46.96396184246987 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1663.0, "completions/max_terminated_length": 1663.0, "completions/mean_length": 819.046875, "completions/mean_terminated_length": 819.046875, "completions/min_length": 381.0, "completions/min_terminated_length": 381.0, "entropy": 0.6781535688787699, "epoch": 0.27947845804988664, "frac_reward_zero_std": 0.625, "grad_norm": 0.41015625, "learning_rate": 7.210884353741496e-07, "loss": 0.0083, "num_tokens": 29014574.0, "reward": 0.265625, "reward_std": 0.6782841682434082, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 493, "step_time": 56.07055533491075 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1452.0, "completions/max_terminated_length": 1452.0, "completions/mean_length": 765.6875, "completions/mean_terminated_length": 765.6875, "completions/min_length": 301.0, "completions/min_terminated_length": 301.0, "entropy": 0.8276981525123119, "epoch": 0.2800453514739229, "frac_reward_zero_std": 0.75, "grad_norm": 0.353515625, "learning_rate": 7.205215419501134e-07, "loss": 0.0084, "num_tokens": 29074026.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 494, "step_time": 49.20934342965484 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1183.0, "completions/max_terminated_length": 1183.0, "completions/mean_length": 709.609375, "completions/mean_terminated_length": 709.609375, "completions/min_length": 316.0, "completions/min_terminated_length": 316.0, "entropy": 1.0416027531027794, "epoch": 0.28061224489795916, "frac_reward_zero_std": 0.625, "grad_norm": 0.498046875, "learning_rate": 7.19954648526077e-07, "loss": -0.0205, "num_tokens": 29131681.0, "reward": 0.3515625, "reward_std": 0.7489659190177917, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 495, "step_time": 40.29257717169821 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3069.0, "completions/max_terminated_length": 3069.0, "completions/mean_length": 841.96875, "completions/mean_terminated_length": 841.96875, "completions/min_length": 211.0, "completions/min_terminated_length": 211.0, "entropy": 1.0242768581956625, "epoch": 0.2811791383219955, "frac_reward_zero_std": 0.875, "grad_norm": 0.275390625, "learning_rate": 7.193877551020408e-07, "loss": 0.0039, "num_tokens": 29196975.0, "reward": 0.046875, "reward_std": 0.14689241349697113, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 496, "step_time": 115.6932109799236 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 748.203125, "completions/mean_terminated_length": 748.203125, "completions/min_length": 299.0, "completions/min_terminated_length": 299.0, "entropy": 0.7642087694257498, "epoch": 0.28174603174603174, "frac_reward_zero_std": 0.375, "grad_norm": 0.546875, "learning_rate": 7.188208616780045e-07, "loss": 0.0442, "num_tokens": 29257428.0, "reward": 0.359375, "reward_std": 0.7263124585151672, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 497, "step_time": 45.00180886685848 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1546.0, "completions/max_terminated_length": 1546.0, "completions/mean_length": 781.859375, "completions/mean_terminated_length": 781.859375, "completions/min_length": 359.0, "completions/min_terminated_length": 359.0, "entropy": 0.9330013245344162, "epoch": 0.282312925170068, "frac_reward_zero_std": 0.625, "grad_norm": 0.45703125, "learning_rate": 7.182539682539682e-07, "loss": -0.0062, "num_tokens": 29319491.0, "reward": 0.078125, "reward_std": 0.3587760329246521, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 498, "step_time": 52.68529424164444 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 793.0625, "completions/mean_terminated_length": 793.0625, "completions/min_length": 7.0, "completions/min_terminated_length": 7.0, "entropy": 0.6890045292675495, "epoch": 0.2828798185941043, "frac_reward_zero_std": 0.75, "grad_norm": 0.353515625, "learning_rate": 7.176870748299319e-07, "loss": 0.0057, "num_tokens": 29380775.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 499, "step_time": 49.33592140302062 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1679.0, "completions/max_terminated_length": 1679.0, "completions/mean_length": 749.6875, "completions/mean_terminated_length": 749.6875, "completions/min_length": 19.0, "completions/min_terminated_length": 19.0, "entropy": 0.9031962808221579, "epoch": 0.2834467120181406, "frac_reward_zero_std": 0.75, "grad_norm": 0.359375, "learning_rate": 7.171201814058957e-07, "loss": -0.0009, "num_tokens": 29440547.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 500, "step_time": 57.61708065588027 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1302.0, "completions/max_terminated_length": 1302.0, "completions/mean_length": 669.671875, "completions/mean_terminated_length": 669.671875, "completions/min_length": 266.0, "completions/min_terminated_length": 266.0, "entropy": 0.8602342400699854, "epoch": 0.28401360544217685, "frac_reward_zero_std": 0.625, "grad_norm": 0.47265625, "learning_rate": 7.165532879818594e-07, "loss": 0.0134, "num_tokens": 29493126.0, "reward": 0.34375, "reward_std": 0.6950790882110596, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 501, "step_time": 47.38943115156144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 768.4375, "completions/mean_terminated_length": 768.4375, "completions/min_length": 367.0, "completions/min_terminated_length": 367.0, "entropy": 0.7709177304059267, "epoch": 0.28458049886621317, "frac_reward_zero_std": 0.5, "grad_norm": 0.51171875, "learning_rate": 7.159863945578231e-07, "loss": 0.0025, "num_tokens": 29553514.0, "reward": 0.3125, "reward_std": 0.7210001945495605, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 502, "step_time": 41.020229395478964 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1671.0, "completions/max_terminated_length": 1671.0, "completions/mean_length": 762.71875, "completions/mean_terminated_length": 762.71875, "completions/min_length": 231.0, "completions/min_terminated_length": 231.0, "entropy": 0.8904999010264874, "epoch": 0.28514739229024944, "frac_reward_zero_std": 0.625, "grad_norm": 0.404296875, "learning_rate": 7.154195011337868e-07, "loss": 0.0084, "num_tokens": 29613840.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 503, "step_time": 56.85954328905791 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1644.0, "completions/max_terminated_length": 1644.0, "completions/mean_length": 879.484375, "completions/mean_terminated_length": 879.484375, "completions/min_length": 380.0, "completions/min_terminated_length": 380.0, "entropy": 0.6220353841781616, "epoch": 0.2857142857142857, "frac_reward_zero_std": 0.75, "grad_norm": 0.28125, "learning_rate": 7.148526077097506e-07, "loss": -0.0277, "num_tokens": 29679599.0, "reward": 0.125, "reward_std": 0.417855441570282, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 504, "step_time": 55.55082425381988 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 709.75, "completions/mean_terminated_length": 709.75, "completions/min_length": 309.0, "completions/min_terminated_length": 309.0, "entropy": 0.698782479390502, "epoch": 0.286281179138322, "frac_reward_zero_std": 0.5, "grad_norm": 0.5546875, "learning_rate": 7.142857142857143e-07, "loss": 0.0252, "num_tokens": 29732879.0, "reward": 0.3046875, "reward_std": 0.7539166808128357, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 505, "step_time": 54.8224318632856 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2076.0, "completions/max_terminated_length": 2076.0, "completions/mean_length": 793.15625, "completions/mean_terminated_length": 793.15625, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "entropy": 0.883378142490983, "epoch": 0.2868480725623583, "frac_reward_zero_std": 0.75, "grad_norm": 0.369140625, "learning_rate": 7.13718820861678e-07, "loss": -0.0302, "num_tokens": 29793113.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 506, "step_time": 71.06204910762608 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1252.0, "completions/max_terminated_length": 1252.0, "completions/mean_length": 669.5625, "completions/mean_terminated_length": 669.5625, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "entropy": 0.8932874873280525, "epoch": 0.28741496598639454, "frac_reward_zero_std": 0.625, "grad_norm": 0.54296875, "learning_rate": 7.131519274376417e-07, "loss": -0.0184, "num_tokens": 29848925.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 507, "step_time": 44.22843154706061 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1084.0, "completions/max_terminated_length": 1084.0, "completions/mean_length": 629.3125, "completions/mean_terminated_length": 629.3125, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "entropy": 0.733006564900279, "epoch": 0.28798185941043086, "frac_reward_zero_std": 0.75, "grad_norm": 0.47265625, "learning_rate": 7.125850340136053e-07, "loss": 0.0016, "num_tokens": 29898377.0, "reward": 0.25, "reward_std": 0.6172134280204773, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 508, "step_time": 36.84416067041457 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1565.0, "completions/max_terminated_length": 1565.0, "completions/mean_length": 763.890625, "completions/mean_terminated_length": 763.890625, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "entropy": 0.8635252937674522, "epoch": 0.2885487528344671, "frac_reward_zero_std": 0.375, "grad_norm": 0.67578125, "learning_rate": 7.120181405895691e-07, "loss": 0.0158, "num_tokens": 29959178.0, "reward": 0.3203125, "reward_std": 0.7419792413711548, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 509, "step_time": 53.838072184473276 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1169.0, "completions/max_terminated_length": 1169.0, "completions/mean_length": 704.0625, "completions/mean_terminated_length": 704.0625, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "entropy": 0.8978853225708008, "epoch": 0.2891156462585034, "frac_reward_zero_std": 0.875, "grad_norm": 0.267578125, "learning_rate": 7.114512471655328e-07, "loss": -0.0063, "num_tokens": 30012262.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 510, "step_time": 39.582871126011014 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1647.0, "completions/max_terminated_length": 1647.0, "completions/mean_length": 783.859375, "completions/mean_terminated_length": 783.859375, "completions/min_length": 411.0, "completions/min_terminated_length": 411.0, "entropy": 0.799968246370554, "epoch": 0.2896825396825397, "frac_reward_zero_std": 0.5, "grad_norm": 0.490234375, "learning_rate": 7.108843537414966e-07, "loss": 0.0272, "num_tokens": 30074093.0, "reward": 0.203125, "reward_std": 0.5754139423370361, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 511, "step_time": 57.25170711334795 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1238.0, "completions/max_terminated_length": 1238.0, "completions/mean_length": 717.03125, "completions/mean_terminated_length": 717.03125, "completions/min_length": 310.0, "completions/min_terminated_length": 310.0, "entropy": 0.8698889967054129, "epoch": 0.29024943310657597, "frac_reward_zero_std": 0.625, "grad_norm": 0.447265625, "learning_rate": 7.103174603174603e-07, "loss": 0.0059, "num_tokens": 30131095.0, "reward": 0.3125, "reward_std": 0.6636837720870972, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 512, "step_time": 42.829598689451814 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1260.0, "completions/max_terminated_length": 1260.0, "completions/mean_length": 771.09375, "completions/mean_terminated_length": 771.09375, "completions/min_length": 384.0, "completions/min_terminated_length": 384.0, "entropy": 0.7532858662307262, "epoch": 0.29081632653061223, "frac_reward_zero_std": 0.625, "grad_norm": 0.46875, "learning_rate": 7.097505668934241e-07, "loss": -0.003, "num_tokens": 30190981.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 513, "step_time": 42.77709187101573 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 754.40625, "completions/mean_terminated_length": 754.40625, "completions/min_length": 157.0, "completions/min_terminated_length": 157.0, "entropy": 0.8620533980429173, "epoch": 0.29138321995464855, "frac_reward_zero_std": 0.625, "grad_norm": 0.47265625, "learning_rate": 7.091836734693876e-07, "loss": 0.023, "num_tokens": 30250655.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 514, "step_time": 45.86416158825159 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1073.0, "completions/max_terminated_length": 1073.0, "completions/mean_length": 645.046875, "completions/mean_terminated_length": 645.046875, "completions/min_length": 274.0, "completions/min_terminated_length": 274.0, "entropy": 0.8946611173450947, "epoch": 0.2919501133786848, "frac_reward_zero_std": 0.5, "grad_norm": 0.546875, "learning_rate": 7.086167800453514e-07, "loss": 0.029, "num_tokens": 30305274.0, "reward": 0.296875, "reward_std": 0.6649904847145081, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 515, "step_time": 38.31303133163601 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1503.0, "completions/max_terminated_length": 1503.0, "completions/mean_length": 831.09375, "completions/mean_terminated_length": 831.09375, "completions/min_length": 326.0, "completions/min_terminated_length": 326.0, "entropy": 0.7219805084168911, "epoch": 0.2925170068027211, "frac_reward_zero_std": 0.875, "grad_norm": 0.2421875, "learning_rate": 7.080498866213151e-07, "loss": -0.0237, "num_tokens": 30370616.0, "reward": 0.0546875, "reward_std": 0.26899558305740356, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 516, "step_time": 51.80148950777948 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1328.0, "completions/max_terminated_length": 1328.0, "completions/mean_length": 776.765625, "completions/mean_terminated_length": 776.765625, "completions/min_length": 285.0, "completions/min_terminated_length": 285.0, "entropy": 0.7232088558375835, "epoch": 0.29308390022675734, "frac_reward_zero_std": 0.875, "grad_norm": 0.27734375, "learning_rate": 7.074829931972789e-07, "loss": -0.0008, "num_tokens": 30430017.0, "reward": 0.046875, "reward_std": 0.14689241349697113, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 517, "step_time": 45.39673104416579 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1546.0, "completions/max_terminated_length": 1546.0, "completions/mean_length": 677.609375, "completions/mean_terminated_length": 677.609375, "completions/min_length": 267.0, "completions/min_terminated_length": 267.0, "entropy": 0.6846966706216335, "epoch": 0.29365079365079366, "frac_reward_zero_std": 0.5, "grad_norm": 0.57421875, "learning_rate": 7.069160997732426e-07, "loss": 0.016, "num_tokens": 30483504.0, "reward": 0.359375, "reward_std": 0.7478474378585815, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 518, "step_time": 52.81310435663909 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1577.0, "completions/max_terminated_length": 1577.0, "completions/mean_length": 750.40625, "completions/mean_terminated_length": 750.40625, "completions/min_length": 285.0, "completions/min_terminated_length": 285.0, "entropy": 0.8029123768210411, "epoch": 0.2942176870748299, "frac_reward_zero_std": 0.625, "grad_norm": 0.42578125, "learning_rate": 7.063492063492064e-07, "loss": 0.0076, "num_tokens": 30540378.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 519, "step_time": 53.39374658744782 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1460.0, "completions/max_terminated_length": 1460.0, "completions/mean_length": 910.828125, "completions/mean_terminated_length": 910.828125, "completions/min_length": 452.0, "completions/min_terminated_length": 452.0, "entropy": 0.6312018167227507, "epoch": 0.2947845804988662, "frac_reward_zero_std": 0.625, "grad_norm": 0.359375, "learning_rate": 7.0578231292517e-07, "loss": 0.0107, "num_tokens": 30611967.0, "reward": 0.1875, "reward_std": 0.5455447435379028, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 520, "step_time": 51.096797372214496 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 810.53125, "completions/mean_terminated_length": 810.53125, "completions/min_length": 407.0, "completions/min_terminated_length": 407.0, "entropy": 0.811397809535265, "epoch": 0.2953514739229025, "frac_reward_zero_std": 0.5, "grad_norm": 0.46875, "learning_rate": 7.052154195011338e-07, "loss": 0.0079, "num_tokens": 30676281.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 521, "step_time": 42.19156005140394 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2165.0, "completions/max_terminated_length": 2165.0, "completions/mean_length": 684.265625, "completions/mean_terminated_length": 684.265625, "completions/min_length": 356.0, "completions/min_terminated_length": 356.0, "entropy": 0.7709549944847822, "epoch": 0.29591836734693877, "frac_reward_zero_std": 0.5, "grad_norm": 0.578125, "learning_rate": 7.046485260770974e-07, "loss": 0.0225, "num_tokens": 30730258.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 522, "step_time": 74.87661859579384 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1502.0, "completions/max_terminated_length": 1502.0, "completions/mean_length": 756.203125, "completions/mean_terminated_length": 756.203125, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "entropy": 0.7819712497293949, "epoch": 0.29648526077097503, "frac_reward_zero_std": 0.875, "grad_norm": 0.23828125, "learning_rate": 7.040816326530612e-07, "loss": 0.0001, "num_tokens": 30789095.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 523, "step_time": 51.73434377089143 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1584.0, "completions/max_terminated_length": 1584.0, "completions/mean_length": 696.265625, "completions/mean_terminated_length": 696.265625, "completions/min_length": 284.0, "completions/min_terminated_length": 284.0, "entropy": 0.6800563894212246, "epoch": 0.29705215419501135, "frac_reward_zero_std": 0.75, "grad_norm": 0.3828125, "learning_rate": 7.035147392290249e-07, "loss": 0.0017, "num_tokens": 30845672.0, "reward": 0.1484375, "reward_std": 0.5092765688896179, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 524, "step_time": 55.245374178513885 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1404.0, "completions/max_terminated_length": 1404.0, "completions/mean_length": 750.734375, "completions/mean_terminated_length": 750.734375, "completions/min_length": 407.0, "completions/min_terminated_length": 407.0, "entropy": 0.7135310284793377, "epoch": 0.2976190476190476, "frac_reward_zero_std": 0.625, "grad_norm": 0.47265625, "learning_rate": 7.029478458049887e-07, "loss": -0.0067, "num_tokens": 30904439.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 525, "step_time": 47.8006643243134 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1607.0, "completions/max_terminated_length": 1607.0, "completions/mean_length": 836.84375, "completions/mean_terminated_length": 836.84375, "completions/min_length": 430.0, "completions/min_terminated_length": 430.0, "entropy": 0.7955428976565599, "epoch": 0.2981859410430839, "frac_reward_zero_std": 0.625, "grad_norm": 0.44140625, "learning_rate": 7.023809523809523e-07, "loss": -0.0231, "num_tokens": 30968933.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 526, "step_time": 55.37982105277479 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 773.140625, "completions/mean_terminated_length": 773.140625, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "entropy": 0.7125669363886118, "epoch": 0.2987528344671202, "frac_reward_zero_std": 0.625, "grad_norm": 0.439453125, "learning_rate": 7.018140589569161e-07, "loss": 0.0143, "num_tokens": 31029470.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 527, "step_time": 45.26804610528052 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1274.0, "completions/max_terminated_length": 1274.0, "completions/mean_length": 675.0, "completions/mean_terminated_length": 675.0, "completions/min_length": 230.0, "completions/min_terminated_length": 230.0, "entropy": 0.8840137869119644, "epoch": 0.29931972789115646, "frac_reward_zero_std": 0.625, "grad_norm": 0.54296875, "learning_rate": 7.012471655328798e-07, "loss": -0.0026, "num_tokens": 31086014.0, "reward": 0.1015625, "reward_std": 0.3692515790462494, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 528, "step_time": 43.771443992853165 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 843.234375, "completions/mean_terminated_length": 843.234375, "completions/min_length": 487.0, "completions/min_terminated_length": 487.0, "entropy": 0.8427235707640648, "epoch": 0.2998866213151927, "frac_reward_zero_std": 0.75, "grad_norm": 0.36328125, "learning_rate": 7.006802721088436e-07, "loss": -0.0055, "num_tokens": 31150757.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 529, "step_time": 47.427061734721065 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 727.515625, "completions/mean_terminated_length": 727.515625, "completions/min_length": 374.0, "completions/min_terminated_length": 374.0, "entropy": 0.7390955612063408, "epoch": 0.30045351473922904, "frac_reward_zero_std": 0.625, "grad_norm": 0.482421875, "learning_rate": 7.001133786848072e-07, "loss": 0.0034, "num_tokens": 31209062.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 530, "step_time": 41.13703847210854 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1342.0, "completions/max_terminated_length": 1342.0, "completions/mean_length": 743.859375, "completions/mean_terminated_length": 743.859375, "completions/min_length": 362.0, "completions/min_terminated_length": 362.0, "entropy": 0.8050366118550301, "epoch": 0.3010204081632653, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 6.995464852607709e-07, "loss": -0.0348, "num_tokens": 31267789.0, "reward": 0.3046875, "reward_std": 0.664384126663208, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 531, "step_time": 46.66273658722639 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1820.0, "completions/max_terminated_length": 1820.0, "completions/mean_length": 713.109375, "completions/mean_terminated_length": 713.109375, "completions/min_length": 335.0, "completions/min_terminated_length": 335.0, "entropy": 0.8603307083249092, "epoch": 0.30158730158730157, "frac_reward_zero_std": 0.375, "grad_norm": 0.69140625, "learning_rate": 6.989795918367347e-07, "loss": 0.0163, "num_tokens": 31324644.0, "reward": 0.1875, "reward_std": 0.5739033818244934, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 532, "step_time": 62.559358855709434 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 673.40625, "completions/mean_terminated_length": 673.40625, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "entropy": 0.8667759243398905, "epoch": 0.3021541950113379, "frac_reward_zero_std": 0.5, "grad_norm": 0.5546875, "learning_rate": 6.984126984126983e-07, "loss": 0.0014, "num_tokens": 31377774.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 533, "step_time": 36.42737630009651 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 813.75, "completions/mean_terminated_length": 813.75, "completions/min_length": 383.0, "completions/min_terminated_length": 383.0, "entropy": 0.57998451218009, "epoch": 0.30272108843537415, "frac_reward_zero_std": 0.375, "grad_norm": 0.54296875, "learning_rate": 6.978458049886621e-07, "loss": -0.0264, "num_tokens": 31440150.0, "reward": 0.375, "reward_std": 0.7237468957901001, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 534, "step_time": 44.09220365062356 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1368.0, "completions/max_terminated_length": 1368.0, "completions/mean_length": 789.1875, "completions/mean_terminated_length": 789.1875, "completions/min_length": 357.0, "completions/min_terminated_length": 357.0, "entropy": 0.6913237422704697, "epoch": 0.3032879818594104, "frac_reward_zero_std": 0.625, "grad_norm": 0.45703125, "learning_rate": 6.972789115646258e-07, "loss": 0.0284, "num_tokens": 31500722.0, "reward": 0.2578125, "reward_std": 0.654606282711029, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 535, "step_time": 46.49318843707442 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1232.0, "completions/max_terminated_length": 1232.0, "completions/mean_length": 754.171875, "completions/mean_terminated_length": 754.171875, "completions/min_length": 349.0, "completions/min_terminated_length": 349.0, "entropy": 0.7832220420241356, "epoch": 0.30385487528344673, "frac_reward_zero_std": 0.625, "grad_norm": 0.47265625, "learning_rate": 6.967120181405896e-07, "loss": 0.0036, "num_tokens": 31558805.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 536, "step_time": 42.142516100779176 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1386.0, "completions/max_terminated_length": 1386.0, "completions/mean_length": 747.328125, "completions/mean_terminated_length": 747.328125, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "entropy": 0.744468754157424, "epoch": 0.304421768707483, "frac_reward_zero_std": 0.375, "grad_norm": 0.546875, "learning_rate": 6.961451247165533e-07, "loss": 0.0006, "num_tokens": 31621434.0, "reward": 0.3671875, "reward_std": 0.7676090598106384, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 537, "step_time": 50.692077185958624 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1630.0, "completions/max_terminated_length": 1630.0, "completions/mean_length": 696.609375, "completions/mean_terminated_length": 696.609375, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "entropy": 0.7965532336384058, "epoch": 0.30498866213151926, "frac_reward_zero_std": 0.375, "grad_norm": 0.67578125, "learning_rate": 6.95578231292517e-07, "loss": -0.0114, "num_tokens": 31678865.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 538, "step_time": 56.480417009443045 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 849.265625, "completions/mean_terminated_length": 849.265625, "completions/min_length": 437.0, "completions/min_terminated_length": 437.0, "entropy": 0.8315318040549755, "epoch": 0.3055555555555556, "frac_reward_zero_std": 0.625, "grad_norm": 0.421875, "learning_rate": 6.950113378684806e-07, "loss": 0.0009, "num_tokens": 31745010.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 539, "step_time": 59.35373664740473 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1554.0, "completions/max_terminated_length": 1554.0, "completions/mean_length": 796.84375, "completions/mean_terminated_length": 796.84375, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "entropy": 0.7862520944327116, "epoch": 0.30612244897959184, "frac_reward_zero_std": 0.625, "grad_norm": 0.419921875, "learning_rate": 6.944444444444444e-07, "loss": 0.001, "num_tokens": 31807048.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 540, "step_time": 53.72292930446565 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 725.5625, "completions/mean_terminated_length": 725.5625, "completions/min_length": 373.0, "completions/min_terminated_length": 373.0, "entropy": 0.9693691655993462, "epoch": 0.3066893424036281, "frac_reward_zero_std": 0.625, "grad_norm": 0.44140625, "learning_rate": 6.938775510204081e-07, "loss": -0.0146, "num_tokens": 31864004.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 541, "step_time": 51.74457472469658 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 712.453125, "completions/mean_terminated_length": 712.453125, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "entropy": 0.9037486426532269, "epoch": 0.3072562358276644, "frac_reward_zero_std": 0.625, "grad_norm": 0.5078125, "learning_rate": 6.933106575963719e-07, "loss": -0.0017, "num_tokens": 31919913.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 542, "step_time": 51.879045808687806 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1188.0, "completions/max_terminated_length": 1188.0, "completions/mean_length": 651.984375, "completions/mean_terminated_length": 651.984375, "completions/min_length": 4.0, "completions/min_terminated_length": 4.0, "entropy": 1.141062118113041, "epoch": 0.3078231292517007, "frac_reward_zero_std": 0.5, "grad_norm": 0.62109375, "learning_rate": 6.927437641723356e-07, "loss": -0.0188, "num_tokens": 31972304.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 543, "step_time": 40.88096096273512 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1417.0, "completions/max_terminated_length": 1417.0, "completions/mean_length": 826.140625, "completions/mean_terminated_length": 826.140625, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "entropy": 0.7591061405837536, "epoch": 0.30839002267573695, "frac_reward_zero_std": 0.625, "grad_norm": 0.400390625, "learning_rate": 6.921768707482994e-07, "loss": 0.0363, "num_tokens": 32036385.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 544, "step_time": 48.60988793708384 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1401.0, "completions/max_terminated_length": 1401.0, "completions/mean_length": 744.8125, "completions/mean_terminated_length": 744.8125, "completions/min_length": 378.0, "completions/min_terminated_length": 378.0, "entropy": 0.7927939239889383, "epoch": 0.3089569160997732, "frac_reward_zero_std": 0.625, "grad_norm": 0.451171875, "learning_rate": 6.91609977324263e-07, "loss": 0.0235, "num_tokens": 32097493.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 545, "step_time": 48.86637980118394 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1425.0, "completions/max_terminated_length": 1425.0, "completions/mean_length": 732.296875, "completions/mean_terminated_length": 732.296875, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "entropy": 0.7569485493004322, "epoch": 0.30952380952380953, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 6.910430839002267e-07, "loss": 0.0135, "num_tokens": 32156384.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 546, "step_time": 49.739972153678536 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1347.0, "completions/max_terminated_length": 1347.0, "completions/mean_length": 680.53125, "completions/mean_terminated_length": 680.53125, "completions/min_length": 377.0, "completions/min_terminated_length": 377.0, "entropy": 0.6933835446834564, "epoch": 0.3100907029478458, "frac_reward_zero_std": 0.5, "grad_norm": 0.5625, "learning_rate": 6.904761904761904e-07, "loss": 0.0049, "num_tokens": 32209106.0, "reward": 0.21875, "reward_std": 0.5764904618263245, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 547, "step_time": 45.50701228249818 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 768.578125, "completions/mean_terminated_length": 768.578125, "completions/min_length": 331.0, "completions/min_terminated_length": 331.0, "entropy": 0.902186393737793, "epoch": 0.31065759637188206, "frac_reward_zero_std": 0.625, "grad_norm": 0.478515625, "learning_rate": 6.899092970521542e-07, "loss": -0.0052, "num_tokens": 32269119.0, "reward": 0.15625, "reward_std": 0.49501484632492065, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 548, "step_time": 43.688417446799576 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1454.0, "completions/max_terminated_length": 1454.0, "completions/mean_length": 817.125, "completions/mean_terminated_length": 817.125, "completions/min_length": 359.0, "completions/min_terminated_length": 359.0, "entropy": 0.8277923595160246, "epoch": 0.3112244897959184, "frac_reward_zero_std": 0.625, "grad_norm": 0.40625, "learning_rate": 6.893424036281179e-07, "loss": 0.0245, "num_tokens": 32331335.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 549, "step_time": 50.39040738157928 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 746.328125, "completions/mean_terminated_length": 746.328125, "completions/min_length": 176.0, "completions/min_terminated_length": 176.0, "entropy": 0.6842802502214909, "epoch": 0.31179138321995464, "frac_reward_zero_std": 0.375, "grad_norm": 0.56640625, "learning_rate": 6.887755102040817e-07, "loss": 0.014, "num_tokens": 32390716.0, "reward": 0.296875, "reward_std": 0.6649904847145081, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 550, "step_time": 44.9338699048385 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 683.09375, "completions/mean_terminated_length": 683.09375, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "entropy": 0.8729349300265312, "epoch": 0.3123582766439909, "frac_reward_zero_std": 0.625, "grad_norm": 0.5234375, "learning_rate": 6.882086167800453e-07, "loss": 0.0186, "num_tokens": 32445186.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 551, "step_time": 44.33868500776589 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1002.0, "completions/max_terminated_length": 1002.0, "completions/mean_length": 542.546875, "completions/mean_terminated_length": 542.546875, "completions/min_length": 99.0, "completions/min_terminated_length": 99.0, "entropy": 0.8894463591277599, "epoch": 0.3129251700680272, "frac_reward_zero_std": 0.5, "grad_norm": 0.6640625, "learning_rate": 6.87641723356009e-07, "loss": 0.0062, "num_tokens": 32490925.0, "reward": 0.59375, "reward_std": 0.9380620121955872, "rewards/accuracy_reward_func/mean": 0.53125, "rewards/accuracy_reward_func/std": 0.8903138637542725, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 552, "step_time": 34.315698532387614 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1125.0, "completions/max_terminated_length": 1125.0, "completions/mean_length": 669.484375, "completions/mean_terminated_length": 669.484375, "completions/min_length": 182.0, "completions/min_terminated_length": 182.0, "entropy": 0.9448998272418976, "epoch": 0.3134920634920635, "frac_reward_zero_std": 0.5, "grad_norm": 0.65625, "learning_rate": 6.870748299319728e-07, "loss": 0.0165, "num_tokens": 32543484.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 553, "step_time": 38.50360931828618 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1497.0, "completions/max_terminated_length": 1497.0, "completions/mean_length": 745.609375, "completions/mean_terminated_length": 745.609375, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 0.9086252637207508, "epoch": 0.31405895691609975, "frac_reward_zero_std": 0.5, "grad_norm": 0.54296875, "learning_rate": 6.865079365079364e-07, "loss": 0.0064, "num_tokens": 32603171.0, "reward": 0.265625, "reward_std": 0.548943817615509, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.109375, "rewards/format_reward_func/std": 0.2083333432674408, "step": 554, "step_time": 51.281183240935206 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 915.921875, "completions/mean_terminated_length": 915.921875, "completions/min_length": 371.0, "completions/min_terminated_length": 371.0, "entropy": 0.671868585050106, "epoch": 0.31462585034013607, "frac_reward_zero_std": 0.625, "grad_norm": 0.427734375, "learning_rate": 6.859410430839002e-07, "loss": 0.0095, "num_tokens": 32673350.0, "reward": 0.1015625, "reward_std": 0.3692515790462494, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 555, "step_time": 55.37346922699362 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 665.46875, "completions/mean_terminated_length": 665.46875, "completions/min_length": 248.0, "completions/min_terminated_length": 248.0, "entropy": 0.8857528772205114, "epoch": 0.31519274376417233, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 6.853741496598639e-07, "loss": 0.0301, "num_tokens": 32726140.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 556, "step_time": 42.94716017693281 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1541.0, "completions/max_terminated_length": 1541.0, "completions/mean_length": 826.265625, "completions/mean_terminated_length": 826.265625, "completions/min_length": 372.0, "completions/min_terminated_length": 372.0, "entropy": 0.7831900082528591, "epoch": 0.3157596371882086, "frac_reward_zero_std": 0.5, "grad_norm": 0.45703125, "learning_rate": 6.848072562358276e-07, "loss": -0.0003, "num_tokens": 32789085.0, "reward": 0.2421875, "reward_std": 0.666620135307312, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 557, "step_time": 52.977177309803665 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1338.0, "completions/max_terminated_length": 1338.0, "completions/mean_length": 778.609375, "completions/mean_terminated_length": 778.609375, "completions/min_length": 310.0, "completions/min_terminated_length": 310.0, "entropy": 0.7681636288762093, "epoch": 0.3163265306122449, "frac_reward_zero_std": 0.5, "grad_norm": 0.498046875, "learning_rate": 6.842403628117913e-07, "loss": 0.0189, "num_tokens": 32848660.0, "reward": 0.296875, "reward_std": 0.7111289501190186, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 558, "step_time": 46.01967536751181 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1359.0, "completions/max_terminated_length": 1359.0, "completions/mean_length": 695.90625, "completions/mean_terminated_length": 695.90625, "completions/min_length": 205.0, "completions/min_terminated_length": 205.0, "entropy": 0.8793341629207134, "epoch": 0.3168934240362812, "frac_reward_zero_std": 0.875, "grad_norm": 0.294921875, "learning_rate": 6.836734693877551e-07, "loss": -0.0091, "num_tokens": 32903478.0, "reward": 0.1875, "reward_std": 0.5, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 559, "step_time": 46.370093904435635 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1436.0, "completions/max_terminated_length": 1436.0, "completions/mean_length": 724.828125, "completions/mean_terminated_length": 724.828125, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "entropy": 0.6435755137354136, "epoch": 0.31746031746031744, "frac_reward_zero_std": 0.625, "grad_norm": 0.421875, "learning_rate": 6.831065759637188e-07, "loss": -0.0127, "num_tokens": 32960851.0, "reward": 0.15625, "reward_std": 0.4443204402923584, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 560, "step_time": 48.38396042305976 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1861.0, "completions/max_terminated_length": 1861.0, "completions/mean_length": 750.765625, "completions/mean_terminated_length": 750.765625, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "entropy": 0.7577685285359621, "epoch": 0.31802721088435376, "frac_reward_zero_std": 0.625, "grad_norm": 0.373046875, "learning_rate": 6.825396825396826e-07, "loss": 0.0172, "num_tokens": 33020532.0, "reward": 0.1796875, "reward_std": 0.5589615106582642, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 561, "step_time": 64.55832711420953 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 758.71875, "completions/mean_terminated_length": 758.71875, "completions/min_length": 278.0, "completions/min_terminated_length": 278.0, "entropy": 0.6104346979409456, "epoch": 0.31859410430839, "frac_reward_zero_std": 0.5, "grad_norm": 0.484375, "learning_rate": 6.819727891156462e-07, "loss": -0.0132, "num_tokens": 33078402.0, "reward": 0.265625, "reward_std": 0.6782841682434082, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 562, "step_time": 56.82965278439224 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 708.859375, "completions/mean_terminated_length": 708.859375, "completions/min_length": 15.0, "completions/min_terminated_length": 15.0, "entropy": 0.9760242644697428, "epoch": 0.3191609977324263, "frac_reward_zero_std": 0.75, "grad_norm": 0.392578125, "learning_rate": 6.8140589569161e-07, "loss": -0.0077, "num_tokens": 33134577.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 563, "step_time": 39.54180992022157 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2076.0, "completions/max_terminated_length": 2076.0, "completions/mean_length": 790.15625, "completions/mean_terminated_length": 790.15625, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "entropy": 0.7078715600073338, "epoch": 0.3197278911564626, "frac_reward_zero_std": 0.625, "grad_norm": 0.36328125, "learning_rate": 6.808390022675736e-07, "loss": -0.023, "num_tokens": 33196979.0, "reward": 0.40625, "reward_std": 0.7708735466003418, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 564, "step_time": 74.30706621333957 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1752.0, "completions/max_terminated_length": 1752.0, "completions/mean_length": 705.828125, "completions/mean_terminated_length": 705.828125, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "entropy": 1.0469638369977474, "epoch": 0.32029478458049887, "frac_reward_zero_std": 0.75, "grad_norm": 0.470703125, "learning_rate": 6.802721088435374e-07, "loss": 0.0184, "num_tokens": 33253344.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 565, "step_time": 59.81270517874509 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1351.0, "completions/max_terminated_length": 1351.0, "completions/mean_length": 780.859375, "completions/mean_terminated_length": 780.859375, "completions/min_length": 416.0, "completions/min_terminated_length": 416.0, "entropy": 0.9410541523247957, "epoch": 0.32086167800453513, "frac_reward_zero_std": 0.75, "grad_norm": 0.3515625, "learning_rate": 6.797052154195011e-07, "loss": 0.007, "num_tokens": 33314319.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 566, "step_time": 45.85976626165211 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1599.0, "completions/max_terminated_length": 1599.0, "completions/mean_length": 898.859375, "completions/mean_terminated_length": 898.859375, "completions/min_length": 492.0, "completions/min_terminated_length": 492.0, "entropy": 0.7099578492343426, "epoch": 0.32142857142857145, "frac_reward_zero_std": 0.625, "grad_norm": 0.380859375, "learning_rate": 6.791383219954649e-07, "loss": -0.0052, "num_tokens": 33382998.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 567, "step_time": 54.45179197285324 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1702.0, "completions/max_terminated_length": 1702.0, "completions/mean_length": 778.390625, "completions/mean_terminated_length": 778.390625, "completions/min_length": 339.0, "completions/min_terminated_length": 339.0, "entropy": 0.7448876928538084, "epoch": 0.3219954648526077, "frac_reward_zero_std": 0.5, "grad_norm": 0.55078125, "learning_rate": 6.785714285714286e-07, "loss": -0.0014, "num_tokens": 33444663.0, "reward": 0.2265625, "reward_std": 0.6167611479759216, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 568, "step_time": 58.32504893839359 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1237.0, "completions/max_terminated_length": 1237.0, "completions/mean_length": 729.28125, "completions/mean_terminated_length": 729.28125, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "entropy": 0.7822460401803255, "epoch": 0.322562358276644, "frac_reward_zero_std": 0.625, "grad_norm": 0.41796875, "learning_rate": 6.780045351473924e-07, "loss": -0.0207, "num_tokens": 33502609.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 569, "step_time": 42.56434412673116 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1705.0, "completions/max_terminated_length": 1705.0, "completions/mean_length": 750.046875, "completions/mean_terminated_length": 750.046875, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "entropy": 0.7973836399614811, "epoch": 0.3231292517006803, "frac_reward_zero_std": 0.5, "grad_norm": 0.51171875, "learning_rate": 6.774376417233559e-07, "loss": -0.0002, "num_tokens": 33561988.0, "reward": 0.234375, "reward_std": 0.6170124411582947, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 570, "step_time": 59.071029322221875 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1406.0, "completions/max_terminated_length": 1406.0, "completions/mean_length": 794.96875, "completions/mean_terminated_length": 794.96875, "completions/min_length": 349.0, "completions/min_terminated_length": 349.0, "entropy": 0.7202172577381134, "epoch": 0.32369614512471656, "frac_reward_zero_std": 0.75, "grad_norm": 0.375, "learning_rate": 6.768707482993196e-07, "loss": 0.0092, "num_tokens": 33623810.0, "reward": 0.0625, "reward_std": 0.2745848298072815, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 571, "step_time": 48.03664357494563 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1233.0, "completions/max_terminated_length": 1233.0, "completions/mean_length": 762.6875, "completions/mean_terminated_length": 762.6875, "completions/min_length": 418.0, "completions/min_terminated_length": 418.0, "entropy": 0.831648550927639, "epoch": 0.3242630385487528, "frac_reward_zero_std": 0.75, "grad_norm": 0.353515625, "learning_rate": 6.763038548752834e-07, "loss": 0.0042, "num_tokens": 33682734.0, "reward": 0.125, "reward_std": 0.37796446681022644, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 572, "step_time": 42.203618557192385 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1349.0, "completions/mean_length": 783.515625, "completions/mean_terminated_length": 730.9365234375, "completions/min_length": 334.0, "completions/min_terminated_length": 334.0, "entropy": 0.9939813017845154, "epoch": 0.32482993197278914, "frac_reward_zero_std": 0.5, "grad_norm": 0.5234375, "learning_rate": 6.757369614512471e-07, "loss": 0.0058, "num_tokens": 33744327.0, "reward": 0.3515625, "reward_std": 0.7274640202522278, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 573, "step_time": 170.5334463007748 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1742.0, "completions/max_terminated_length": 1742.0, "completions/mean_length": 832.96875, "completions/mean_terminated_length": 832.96875, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "entropy": 0.7291863579303026, "epoch": 0.3253968253968254, "frac_reward_zero_std": 0.5, "grad_norm": 0.48046875, "learning_rate": 6.751700680272109e-07, "loss": -0.0244, "num_tokens": 33807605.0, "reward": 0.2265625, "reward_std": 0.6419816017150879, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 574, "step_time": 59.69654292333871 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3735.0, "completions/max_terminated_length": 3735.0, "completions/mean_length": 877.171875, "completions/mean_terminated_length": 877.171875, "completions/min_length": 348.0, "completions/min_terminated_length": 348.0, "entropy": 1.0489157401025295, "epoch": 0.32596371882086167, "frac_reward_zero_std": 0.75, "grad_norm": 0.384765625, "learning_rate": 6.746031746031746e-07, "loss": 0.0036, "num_tokens": 33877088.0, "reward": 0.0703125, "reward_std": 0.27984103560447693, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 575, "step_time": 151.11537385731936 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1673.0, "completions/max_terminated_length": 1673.0, "completions/mean_length": 745.53125, "completions/mean_terminated_length": 745.53125, "completions/min_length": 233.0, "completions/min_terminated_length": 233.0, "entropy": 0.8816084526479244, "epoch": 0.32653061224489793, "frac_reward_zero_std": 0.5, "grad_norm": 0.55859375, "learning_rate": 6.740362811791383e-07, "loss": 0.0166, "num_tokens": 33934002.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 576, "step_time": 56.5695292269811 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1446.0, "completions/max_terminated_length": 1446.0, "completions/mean_length": 740.953125, "completions/mean_terminated_length": 740.953125, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "entropy": 0.7327089421451092, "epoch": 0.32709750566893425, "frac_reward_zero_std": 0.375, "grad_norm": 0.58203125, "learning_rate": 6.734693877551019e-07, "loss": -0.009, "num_tokens": 33992335.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 577, "step_time": 49.20490532647818 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 738.65625, "completions/mean_terminated_length": 738.65625, "completions/min_length": 321.0, "completions/min_terminated_length": 321.0, "entropy": 0.7540786378085613, "epoch": 0.3276643990929705, "frac_reward_zero_std": 0.375, "grad_norm": 0.5625, "learning_rate": 6.729024943310657e-07, "loss": -0.0208, "num_tokens": 34049585.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 578, "step_time": 45.131568097509444 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1294.0, "completions/max_terminated_length": 1294.0, "completions/mean_length": 670.796875, "completions/mean_terminated_length": 670.796875, "completions/min_length": 302.0, "completions/min_terminated_length": 302.0, "entropy": 0.8483943492174149, "epoch": 0.3282312925170068, "frac_reward_zero_std": 0.75, "grad_norm": 0.484375, "learning_rate": 6.723356009070294e-07, "loss": -0.0137, "num_tokens": 34102884.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 579, "step_time": 43.97271020244807 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2467.0, "completions/max_terminated_length": 2467.0, "completions/mean_length": 831.953125, "completions/mean_terminated_length": 831.953125, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "entropy": 0.6686179060488939, "epoch": 0.3287981859410431, "frac_reward_zero_std": 0.25, "grad_norm": 0.55859375, "learning_rate": 6.717687074829932e-07, "loss": 0.013, "num_tokens": 34167137.0, "reward": 0.421875, "reward_std": 0.7777600884437561, "rewards/accuracy_reward_func/mean": 0.375, "rewards/accuracy_reward_func/std": 0.7867957949638367, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 580, "step_time": 87.3412484433502 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1763.0, "completions/max_terminated_length": 1763.0, "completions/mean_length": 801.453125, "completions/mean_terminated_length": 801.453125, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "entropy": 0.8247601091861725, "epoch": 0.32936507936507936, "frac_reward_zero_std": 0.75, "grad_norm": 0.326171875, "learning_rate": 6.712018140589569e-07, "loss": -0.0017, "num_tokens": 34230310.0, "reward": 0.0625, "reward_std": 0.2745848298072815, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 581, "step_time": 60.85651118867099 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 792.46875, "completions/mean_terminated_length": 792.46875, "completions/min_length": 358.0, "completions/min_terminated_length": 358.0, "entropy": 0.5867024697363377, "epoch": 0.3299319727891156, "frac_reward_zero_std": 0.625, "grad_norm": 0.421875, "learning_rate": 6.706349206349206e-07, "loss": 0.0166, "num_tokens": 34290940.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 582, "step_time": 54.818203169852495 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1397.0, "completions/max_terminated_length": 1397.0, "completions/mean_length": 779.015625, "completions/mean_terminated_length": 779.015625, "completions/min_length": 209.0, "completions/min_terminated_length": 209.0, "entropy": 0.6622234340757132, "epoch": 0.33049886621315194, "frac_reward_zero_std": 0.5, "grad_norm": 0.466796875, "learning_rate": 6.700680272108843e-07, "loss": 0.0088, "num_tokens": 34351981.0, "reward": 0.1875, "reward_std": 0.5, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 583, "step_time": 47.57754575368017 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 693.1875, "completions/mean_terminated_length": 693.1875, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "entropy": 0.8937420938163996, "epoch": 0.3310657596371882, "frac_reward_zero_std": 0.625, "grad_norm": 0.4609375, "learning_rate": 6.695011337868481e-07, "loss": 0.0229, "num_tokens": 34407801.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 584, "step_time": 42.56495556328446 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1394.0, "completions/max_terminated_length": 1394.0, "completions/mean_length": 931.484375, "completions/mean_terminated_length": 931.484375, "completions/min_length": 251.0, "completions/min_terminated_length": 251.0, "entropy": 0.8077186606824398, "epoch": 0.33163265306122447, "frac_reward_zero_std": 0.875, "grad_norm": 0.2099609375, "learning_rate": 6.689342403628118e-07, "loss": 0.0117, "num_tokens": 34478968.0, "reward": 0.15625, "reward_std": 0.4443204402923584, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 585, "step_time": 47.231089257635176 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1672.0, "completions/max_terminated_length": 1672.0, "completions/mean_length": 716.265625, "completions/mean_terminated_length": 716.265625, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "entropy": 0.8019145056605339, "epoch": 0.3321995464852608, "frac_reward_zero_std": 0.875, "grad_norm": 0.30078125, "learning_rate": 6.683673469387755e-07, "loss": 0.0029, "num_tokens": 34546209.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 586, "step_time": 67.00292479712516 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1335.0, "completions/max_terminated_length": 1335.0, "completions/mean_length": 736.53125, "completions/mean_terminated_length": 736.53125, "completions/min_length": 202.0, "completions/min_terminated_length": 202.0, "entropy": 0.7051402106881142, "epoch": 0.33276643990929705, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 6.678004535147392e-07, "loss": 0.0, "num_tokens": 34604299.0, "reward": 0.0625, "reward_std": 0.1666666716337204, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 587, "step_time": 46.34575128182769 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 708.546875, "completions/mean_terminated_length": 708.546875, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "entropy": 0.8992711044847965, "epoch": 0.3333333333333333, "frac_reward_zero_std": 0.625, "grad_norm": 0.546875, "learning_rate": 6.672335600907029e-07, "loss": -0.0137, "num_tokens": 34664494.0, "reward": 0.140625, "reward_std": 0.4753340482711792, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 588, "step_time": 42.4019844615832 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2654.0, "completions/max_terminated_length": 2654.0, "completions/mean_length": 766.90625, "completions/mean_terminated_length": 766.90625, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "entropy": 0.5499698556959629, "epoch": 0.33390022675736963, "frac_reward_zero_std": 0.5, "grad_norm": 0.462890625, "learning_rate": 6.666666666666666e-07, "loss": 0.029, "num_tokens": 34722872.0, "reward": 0.3046875, "reward_std": 0.6993034482002258, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 589, "step_time": 95.65712734311819 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1403.0, "completions/max_terminated_length": 1403.0, "completions/mean_length": 744.34375, "completions/mean_terminated_length": 744.34375, "completions/min_length": 273.0, "completions/min_terminated_length": 273.0, "entropy": 0.5921298786997795, "epoch": 0.3344671201814059, "frac_reward_zero_std": 0.25, "grad_norm": 0.625, "learning_rate": 6.660997732426304e-07, "loss": -0.0128, "num_tokens": 34780398.0, "reward": 0.34375, "reward_std": 0.7175520658493042, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 590, "step_time": 47.90016901958734 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1064.0, "completions/max_terminated_length": 1064.0, "completions/mean_length": 686.203125, "completions/mean_terminated_length": 686.203125, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "entropy": 1.0365900918841362, "epoch": 0.33503401360544216, "frac_reward_zero_std": 0.75, "grad_norm": 0.43359375, "learning_rate": 6.655328798185941e-07, "loss": 0.0093, "num_tokens": 34835507.0, "reward": 0.125, "reward_std": 0.37796446681022644, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 591, "step_time": 36.734624343924224 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1436.0, "completions/max_terminated_length": 1436.0, "completions/mean_length": 772.140625, "completions/mean_terminated_length": 772.140625, "completions/min_length": 404.0, "completions/min_terminated_length": 404.0, "entropy": 0.806842315942049, "epoch": 0.3356009070294785, "frac_reward_zero_std": 0.75, "grad_norm": 0.462890625, "learning_rate": 6.649659863945578e-07, "loss": 0.0047, "num_tokens": 34895524.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 592, "step_time": 50.1835395488888 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 918.0, "completions/max_terminated_length": 918.0, "completions/mean_length": 660.421875, "completions/mean_terminated_length": 660.421875, "completions/min_length": 318.0, "completions/min_terminated_length": 318.0, "entropy": 0.8969474509358406, "epoch": 0.33616780045351474, "frac_reward_zero_std": 0.625, "grad_norm": 0.52734375, "learning_rate": 6.643990929705216e-07, "loss": -0.0263, "num_tokens": 34947367.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 593, "step_time": 31.30379052646458 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1442.0, "completions/max_terminated_length": 1442.0, "completions/mean_length": 755.59375, "completions/mean_terminated_length": 755.59375, "completions/min_length": 266.0, "completions/min_terminated_length": 266.0, "entropy": 0.690056225284934, "epoch": 0.336734693877551, "frac_reward_zero_std": 0.625, "grad_norm": 0.41015625, "learning_rate": 6.638321995464852e-07, "loss": -0.0176, "num_tokens": 35004309.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 594, "step_time": 49.0545499343425 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1356.0, "completions/max_terminated_length": 1356.0, "completions/mean_length": 795.78125, "completions/mean_terminated_length": 795.78125, "completions/min_length": 290.0, "completions/min_terminated_length": 290.0, "entropy": 0.6606776397675276, "epoch": 0.3373015873015873, "frac_reward_zero_std": 0.75, "grad_norm": 0.357421875, "learning_rate": 6.632653061224489e-07, "loss": 0.0013, "num_tokens": 35065127.0, "reward": 0.15625, "reward_std": 0.4443204402923584, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 595, "step_time": 46.109949179925025 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 769.59375, "completions/mean_terminated_length": 769.59375, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "entropy": 0.647127928212285, "epoch": 0.3378684807256236, "frac_reward_zero_std": 0.625, "grad_norm": 0.427734375, "learning_rate": 6.626984126984126e-07, "loss": 0.0099, "num_tokens": 35124485.0, "reward": 0.1796875, "reward_std": 0.5298036932945251, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 596, "step_time": 51.34399641864002 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1460.0, "completions/max_terminated_length": 1460.0, "completions/mean_length": 785.75, "completions/mean_terminated_length": 785.75, "completions/min_length": 132.0, "completions/min_terminated_length": 132.0, "entropy": 0.8237004764378071, "epoch": 0.33843537414965985, "frac_reward_zero_std": 0.5, "grad_norm": 0.54296875, "learning_rate": 6.621315192743764e-07, "loss": -0.0073, "num_tokens": 35183677.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 597, "step_time": 49.71723074000329 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1332.0, "completions/max_terminated_length": 1332.0, "completions/mean_length": 763.015625, "completions/mean_terminated_length": 763.015625, "completions/min_length": 381.0, "completions/min_terminated_length": 381.0, "entropy": 0.7137241717427969, "epoch": 0.33900226757369617, "frac_reward_zero_std": 0.375, "grad_norm": 0.53125, "learning_rate": 6.615646258503401e-07, "loss": 0.0363, "num_tokens": 35242974.0, "reward": 0.3359375, "reward_std": 0.6961041688919067, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 598, "step_time": 45.134176816791296 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1623.0, "completions/max_terminated_length": 1623.0, "completions/mean_length": 823.53125, "completions/mean_terminated_length": 823.53125, "completions/min_length": 377.0, "completions/min_terminated_length": 377.0, "entropy": 0.6909013129770756, "epoch": 0.33956916099773243, "frac_reward_zero_std": 0.375, "grad_norm": 0.5625, "learning_rate": 6.609977324263039e-07, "loss": 0.0313, "num_tokens": 35308888.0, "reward": 0.34375, "reward_std": 0.728528618812561, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 599, "step_time": 58.116622598841786 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 772.953125, "completions/mean_terminated_length": 772.953125, "completions/min_length": 409.0, "completions/min_terminated_length": 409.0, "entropy": 0.8247405681759119, "epoch": 0.3401360544217687, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 6.604308390022676e-07, "loss": 0.0078, "num_tokens": 35370845.0, "reward": 0.359375, "reward_std": 0.7478474378585815, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 600, "step_time": 52.88959987089038 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 705.8125, "completions/mean_terminated_length": 705.8125, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "entropy": 0.6509308498352766, "epoch": 0.340702947845805, "frac_reward_zero_std": 0.375, "grad_norm": 0.5625, "learning_rate": 6.598639455782312e-07, "loss": 0.0099, "num_tokens": 35425377.0, "reward": 0.3671875, "reward_std": 0.725073516368866, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 601, "step_time": 44.81054830830544 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1102.0, "completions/max_terminated_length": 1102.0, "completions/mean_length": 680.453125, "completions/mean_terminated_length": 680.453125, "completions/min_length": 314.0, "completions/min_terminated_length": 314.0, "entropy": 0.7608289979398251, "epoch": 0.3412698412698413, "frac_reward_zero_std": 0.5, "grad_norm": 0.6171875, "learning_rate": 6.592970521541949e-07, "loss": 0.0142, "num_tokens": 35478118.0, "reward": 0.265625, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 602, "step_time": 37.77187288366258 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 672.109375, "completions/mean_terminated_length": 672.109375, "completions/min_length": 237.0, "completions/min_terminated_length": 237.0, "entropy": 0.6571997348219156, "epoch": 0.34183673469387754, "frac_reward_zero_std": 0.125, "grad_norm": 0.69921875, "learning_rate": 6.587301587301587e-07, "loss": 0.0388, "num_tokens": 35530517.0, "reward": 0.6875, "reward_std": 0.9236745238304138, "rewards/accuracy_reward_func/mean": 0.625, "rewards/accuracy_reward_func/std": 0.934353232383728, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 603, "step_time": 47.63913592789322 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 659.203125, "completions/mean_terminated_length": 659.203125, "completions/min_length": 245.0, "completions/min_terminated_length": 245.0, "entropy": 0.9442145936191082, "epoch": 0.3424036281179138, "frac_reward_zero_std": 0.625, "grad_norm": 0.53125, "learning_rate": 6.581632653061224e-07, "loss": 0.0245, "num_tokens": 35583354.0, "reward": 0.3125, "reward_std": 0.6636837720870972, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 604, "step_time": 63.76840851083398 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1444.0, "completions/max_terminated_length": 1444.0, "completions/mean_length": 697.515625, "completions/mean_terminated_length": 697.515625, "completions/min_length": 287.0, "completions/min_terminated_length": 287.0, "entropy": 0.8977196142077446, "epoch": 0.3429705215419501, "frac_reward_zero_std": 0.625, "grad_norm": 0.458984375, "learning_rate": 6.575963718820862e-07, "loss": 0.0122, "num_tokens": 35638179.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 605, "step_time": 49.88650964573026 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1533.0, "completions/max_terminated_length": 1533.0, "completions/mean_length": 657.5, "completions/mean_terminated_length": 657.5, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "entropy": 0.8332506120204926, "epoch": 0.3435374149659864, "frac_reward_zero_std": 0.75, "grad_norm": 0.39453125, "learning_rate": 6.570294784580499e-07, "loss": 0.0057, "num_tokens": 35691707.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 606, "step_time": 53.01526586804539 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1416.0, "completions/max_terminated_length": 1416.0, "completions/mean_length": 648.828125, "completions/mean_terminated_length": 648.828125, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "entropy": 1.0154879465699196, "epoch": 0.34410430839002265, "frac_reward_zero_std": 0.375, "grad_norm": 0.73046875, "learning_rate": 6.564625850340136e-07, "loss": 0.0016, "num_tokens": 35742760.0, "reward": 0.4921875, "reward_std": 0.8380447030067444, "rewards/accuracy_reward_func/mean": 0.4375, "rewards/accuracy_reward_func/std": 0.8333333730697632, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 607, "step_time": 48.210856564342976 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1414.0, "completions/max_terminated_length": 1414.0, "completions/mean_length": 725.875, "completions/mean_terminated_length": 725.875, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "entropy": 0.944462962448597, "epoch": 0.34467120181405897, "frac_reward_zero_std": 0.75, "grad_norm": 0.421875, "learning_rate": 6.558956916099773e-07, "loss": 0.0032, "num_tokens": 35799640.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 608, "step_time": 48.60132034402341 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1429.0, "completions/max_terminated_length": 1429.0, "completions/mean_length": 846.515625, "completions/mean_terminated_length": 846.515625, "completions/min_length": 515.0, "completions/min_terminated_length": 515.0, "entropy": 0.8410714901983738, "epoch": 0.34523809523809523, "frac_reward_zero_std": 0.5, "grad_norm": 0.5078125, "learning_rate": 6.55328798185941e-07, "loss": 0.0339, "num_tokens": 35864201.0, "reward": 0.25, "reward_std": 0.6299408078193665, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 609, "step_time": 48.66408775374293 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1459.0, "completions/max_terminated_length": 1459.0, "completions/mean_length": 736.625, "completions/mean_terminated_length": 736.625, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "entropy": 0.7623574510216713, "epoch": 0.3458049886621315, "frac_reward_zero_std": 0.75, "grad_norm": 0.390625, "learning_rate": 6.547619047619047e-07, "loss": 0.0185, "num_tokens": 35922593.0, "reward": 0.15625, "reward_std": 0.4443204402923584, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 610, "step_time": 50.23085585050285 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1625.0, "completions/max_terminated_length": 1625.0, "completions/mean_length": 805.015625, "completions/mean_terminated_length": 805.015625, "completions/min_length": 378.0, "completions/min_terminated_length": 378.0, "entropy": 0.8921775817871094, "epoch": 0.3463718820861678, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 6.541950113378684e-07, "loss": 0.0128, "num_tokens": 35985210.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 611, "step_time": 55.43881920911372 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 772.203125, "completions/mean_terminated_length": 772.203125, "completions/min_length": 301.0, "completions/min_terminated_length": 301.0, "entropy": 0.9786438681185246, "epoch": 0.3469387755102041, "frac_reward_zero_std": 0.625, "grad_norm": 0.451171875, "learning_rate": 6.536281179138322e-07, "loss": 0.0244, "num_tokens": 36045959.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 612, "step_time": 48.973512183874846 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 4096.0, "completions/max_terminated_length": 1547.0, "completions/mean_length": 944.84375, "completions/mean_terminated_length": 843.1935424804688, "completions/min_length": 332.0, "completions/min_terminated_length": 332.0, "entropy": 0.7050858326256275, "epoch": 0.34750566893424034, "frac_reward_zero_std": 0.625, "grad_norm": 0.3515625, "learning_rate": 6.530612244897958e-07, "loss": 0.0271, "num_tokens": 36116885.0, "reward": 0.234375, "reward_std": 0.6664806604385376, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 613, "step_time": 167.45049969945103 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1424.0, "completions/max_terminated_length": 1424.0, "completions/mean_length": 714.84375, "completions/mean_terminated_length": 714.84375, "completions/min_length": 324.0, "completions/min_terminated_length": 324.0, "entropy": 0.99871014803648, "epoch": 0.34807256235827666, "frac_reward_zero_std": 0.875, "grad_norm": 0.3125, "learning_rate": 6.524943310657596e-07, "loss": -0.0004, "num_tokens": 36172699.0, "reward": 0.125, "reward_std": 0.37796446681022644, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 614, "step_time": 47.915361465886235 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2035.0, "completions/max_terminated_length": 2035.0, "completions/mean_length": 719.6875, "completions/mean_terminated_length": 719.6875, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "entropy": 0.8770802449434996, "epoch": 0.3486394557823129, "frac_reward_zero_std": 0.5, "grad_norm": 0.62890625, "learning_rate": 6.519274376417233e-07, "loss": -0.0263, "num_tokens": 36228455.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 615, "step_time": 71.19476261176169 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 780.3125, "completions/mean_terminated_length": 780.3125, "completions/min_length": 341.0, "completions/min_terminated_length": 341.0, "entropy": 0.7977379746735096, "epoch": 0.3492063492063492, "frac_reward_zero_std": 0.625, "grad_norm": 0.396484375, "learning_rate": 6.513605442176871e-07, "loss": 0.03, "num_tokens": 36289075.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 616, "step_time": 46.82819330506027 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2057.0, "completions/max_terminated_length": 2057.0, "completions/mean_length": 898.3125, "completions/mean_terminated_length": 898.3125, "completions/min_length": 407.0, "completions/min_terminated_length": 407.0, "entropy": 0.5450865644961596, "epoch": 0.3497732426303855, "frac_reward_zero_std": 0.625, "grad_norm": 0.345703125, "learning_rate": 6.507936507936507e-07, "loss": -0.01, "num_tokens": 36357823.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 617, "step_time": 73.42446285765618 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1304.0, "completions/max_terminated_length": 1304.0, "completions/mean_length": 753.21875, "completions/mean_terminated_length": 753.21875, "completions/min_length": 288.0, "completions/min_terminated_length": 288.0, "entropy": 0.9117572195827961, "epoch": 0.35034013605442177, "frac_reward_zero_std": 0.625, "grad_norm": 0.466796875, "learning_rate": 6.502267573696145e-07, "loss": -0.0248, "num_tokens": 36416093.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 618, "step_time": 44.56446181051433 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1483.0, "completions/mean_length": 864.9375, "completions/mean_terminated_length": 813.6508178710938, "completions/min_length": 411.0, "completions/min_terminated_length": 411.0, "entropy": 0.7935023289173841, "epoch": 0.35090702947845803, "frac_reward_zero_std": 0.5, "grad_norm": 0.46484375, "learning_rate": 6.496598639455782e-07, "loss": 0.0169, "num_tokens": 36483521.0, "reward": 0.21875, "reward_std": 0.5764904618263245, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 619, "step_time": 170.8946770299226 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1218.0, "completions/max_terminated_length": 1218.0, "completions/mean_length": 710.234375, "completions/mean_terminated_length": 710.234375, "completions/min_length": 394.0, "completions/min_terminated_length": 394.0, "entropy": 0.8544355668127537, "epoch": 0.35147392290249435, "frac_reward_zero_std": 0.5, "grad_norm": 0.51171875, "learning_rate": 6.490929705215419e-07, "loss": 0.0179, "num_tokens": 36539160.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 620, "step_time": 42.336375730112195 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1233.0, "completions/max_terminated_length": 1233.0, "completions/mean_length": 711.09375, "completions/mean_terminated_length": 711.09375, "completions/min_length": 356.0, "completions/min_terminated_length": 356.0, "entropy": 0.9783514887094498, "epoch": 0.3520408163265306, "frac_reward_zero_std": 0.75, "grad_norm": 0.419921875, "learning_rate": 6.485260770975056e-07, "loss": 0.0062, "num_tokens": 36596734.0, "reward": 0.3828125, "reward_std": 0.81524258852005, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 621, "step_time": 42.53386629931629 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 740.265625, "completions/mean_terminated_length": 740.265625, "completions/min_length": 18.0, "completions/min_terminated_length": 18.0, "entropy": 0.9278873167932034, "epoch": 0.3526077097505669, "frac_reward_zero_std": 0.875, "grad_norm": 0.30078125, "learning_rate": 6.479591836734694e-07, "loss": 0.0076, "num_tokens": 36654535.0, "reward": 0.0390625, "reward_std": 0.13524484634399414, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 622, "step_time": 40.763666816055775 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 697.25, "completions/mean_terminated_length": 697.25, "completions/min_length": 318.0, "completions/min_terminated_length": 318.0, "entropy": 0.8700597248971462, "epoch": 0.3531746031746032, "frac_reward_zero_std": 0.5, "grad_norm": 0.5546875, "learning_rate": 6.473922902494331e-07, "loss": -0.0035, "num_tokens": 36713311.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 623, "step_time": 40.7892361106351 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1452.0, "completions/max_terminated_length": 1452.0, "completions/mean_length": 767.1875, "completions/mean_terminated_length": 767.1875, "completions/min_length": 225.0, "completions/min_terminated_length": 225.0, "entropy": 0.7525980323553085, "epoch": 0.35374149659863946, "frac_reward_zero_std": 0.5, "grad_norm": 0.50390625, "learning_rate": 6.468253968253969e-07, "loss": -0.006, "num_tokens": 36772115.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 624, "step_time": 49.61737298686057 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1688.0, "completions/max_terminated_length": 1688.0, "completions/mean_length": 818.109375, "completions/mean_terminated_length": 818.109375, "completions/min_length": 304.0, "completions/min_terminated_length": 304.0, "entropy": 0.7521986719220877, "epoch": 0.3543083900226757, "frac_reward_zero_std": 0.375, "grad_norm": 0.51953125, "learning_rate": 6.462585034013605e-07, "loss": -0.0161, "num_tokens": 36835690.0, "reward": 0.2265625, "reward_std": 0.6167611479759216, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 625, "step_time": 58.91802252270281 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 840.390625, "completions/mean_terminated_length": 840.390625, "completions/min_length": 376.0, "completions/min_terminated_length": 376.0, "entropy": 0.717966303229332, "epoch": 0.35487528344671204, "frac_reward_zero_std": 0.625, "grad_norm": 0.45703125, "learning_rate": 6.456916099773242e-07, "loss": 0.0034, "num_tokens": 36901123.0, "reward": 0.1171875, "reward_std": 0.43408530950546265, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 626, "step_time": 57.560720714740455 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1388.0, "completions/max_terminated_length": 1388.0, "completions/mean_length": 777.40625, "completions/mean_terminated_length": 777.40625, "completions/min_length": 382.0, "completions/min_terminated_length": 382.0, "entropy": 0.8006776142865419, "epoch": 0.3554421768707483, "frac_reward_zero_std": 0.625, "grad_norm": 0.455078125, "learning_rate": 6.451247165532879e-07, "loss": 0.0177, "num_tokens": 36962941.0, "reward": 0.125, "reward_std": 0.4364357888698578, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 627, "step_time": 47.950291788205504 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1351.0, "completions/max_terminated_length": 1351.0, "completions/mean_length": 757.0, "completions/mean_terminated_length": 757.0, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "entropy": 0.6608790792524815, "epoch": 0.35600907029478457, "frac_reward_zero_std": 0.25, "grad_norm": 0.625, "learning_rate": 6.445578231292517e-07, "loss": 0.0192, "num_tokens": 37020725.0, "reward": 0.46875, "reward_std": 0.8158888816833496, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 628, "step_time": 46.2614265633747 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1352.0, "completions/max_terminated_length": 1352.0, "completions/mean_length": 818.3125, "completions/mean_terminated_length": 818.3125, "completions/min_length": 218.0, "completions/min_terminated_length": 218.0, "entropy": 0.6299032997339964, "epoch": 0.3565759637188209, "frac_reward_zero_std": 0.625, "grad_norm": 0.35546875, "learning_rate": 6.439909297052154e-07, "loss": 0.0221, "num_tokens": 37082113.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 629, "step_time": 46.16052764002234 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1500.0, "completions/max_terminated_length": 1500.0, "completions/mean_length": 783.0, "completions/mean_terminated_length": 783.0, "completions/min_length": 411.0, "completions/min_terminated_length": 411.0, "entropy": 0.6570486966520548, "epoch": 0.35714285714285715, "frac_reward_zero_std": 0.625, "grad_norm": 0.419921875, "learning_rate": 6.434240362811792e-07, "loss": -0.0167, "num_tokens": 37143633.0, "reward": 0.4609375, "reward_std": 0.7983350157737732, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 630, "step_time": 51.87512902356684 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 766.234375, "completions/mean_terminated_length": 766.234375, "completions/min_length": 394.0, "completions/min_terminated_length": 394.0, "entropy": 0.9654701687395573, "epoch": 0.3577097505668934, "frac_reward_zero_std": 0.875, "grad_norm": 0.2578125, "learning_rate": 6.428571428571429e-07, "loss": -0.0025, "num_tokens": 37202672.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 631, "step_time": 48.53060559555888 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1952.0, "completions/max_terminated_length": 1952.0, "completions/mean_length": 642.921875, "completions/mean_terminated_length": 642.921875, "completions/min_length": 220.0, "completions/min_terminated_length": 220.0, "entropy": 0.8917891718447208, "epoch": 0.35827664399092973, "frac_reward_zero_std": 0.625, "grad_norm": 0.5546875, "learning_rate": 6.422902494331065e-07, "loss": -0.0179, "num_tokens": 37254427.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 632, "step_time": 68.72827850095928 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1191.0, "completions/max_terminated_length": 1191.0, "completions/mean_length": 758.796875, "completions/mean_terminated_length": 758.796875, "completions/min_length": 447.0, "completions/min_terminated_length": 447.0, "entropy": 0.8271527048200369, "epoch": 0.358843537414966, "frac_reward_zero_std": 0.375, "grad_norm": 0.56640625, "learning_rate": 6.417233560090702e-07, "loss": -0.0232, "num_tokens": 37312790.0, "reward": 0.359375, "reward_std": 0.7478474378585815, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 633, "step_time": 40.42411103658378 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 733.765625, "completions/mean_terminated_length": 733.765625, "completions/min_length": 269.0, "completions/min_terminated_length": 269.0, "entropy": 0.7152926605194807, "epoch": 0.35941043083900226, "frac_reward_zero_std": 0.75, "grad_norm": 0.36328125, "learning_rate": 6.411564625850339e-07, "loss": 0.0052, "num_tokens": 37370775.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 634, "step_time": 43.42288984358311 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 807.0, "completions/mean_terminated_length": 807.0, "completions/min_length": 346.0, "completions/min_terminated_length": 346.0, "entropy": 0.524667102843523, "epoch": 0.3599773242630385, "frac_reward_zero_std": 0.5, "grad_norm": 0.42578125, "learning_rate": 6.405895691609977e-07, "loss": 0.0139, "num_tokens": 37432231.0, "reward": 0.203125, "reward_std": 0.5754139423370361, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 635, "step_time": 50.549458659254014 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1161.0, "completions/max_terminated_length": 1161.0, "completions/mean_length": 664.625, "completions/mean_terminated_length": 664.625, "completions/min_length": 308.0, "completions/min_terminated_length": 308.0, "entropy": 0.8426170144230127, "epoch": 0.36054421768707484, "frac_reward_zero_std": 0.375, "grad_norm": 0.625, "learning_rate": 6.400226757369614e-07, "loss": 0.0249, "num_tokens": 37484247.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 636, "step_time": 38.97255809511989 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1452.0, "completions/max_terminated_length": 1452.0, "completions/mean_length": 751.203125, "completions/mean_terminated_length": 751.203125, "completions/min_length": 419.0, "completions/min_terminated_length": 419.0, "entropy": 0.8785478807985783, "epoch": 0.3611111111111111, "frac_reward_zero_std": 0.625, "grad_norm": 0.49609375, "learning_rate": 6.394557823129252e-07, "loss": -0.0048, "num_tokens": 37543156.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 637, "step_time": 49.94559401087463 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 747.765625, "completions/mean_terminated_length": 747.765625, "completions/min_length": 22.0, "completions/min_terminated_length": 22.0, "entropy": 0.7077295072376728, "epoch": 0.36167800453514737, "frac_reward_zero_std": 0.75, "grad_norm": 0.361328125, "learning_rate": 6.388888888888888e-07, "loss": -0.0281, "num_tokens": 37600861.0, "reward": 0.203125, "reward_std": 0.6023679971694946, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 638, "step_time": 45.29154534637928 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1479.0, "completions/max_terminated_length": 1479.0, "completions/mean_length": 856.625, "completions/mean_terminated_length": 856.625, "completions/min_length": 399.0, "completions/min_terminated_length": 399.0, "entropy": 0.7673860471695662, "epoch": 0.3622448979591837, "frac_reward_zero_std": 0.625, "grad_norm": 0.361328125, "learning_rate": 6.383219954648526e-07, "loss": -0.009, "num_tokens": 37666237.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 639, "step_time": 50.89554338529706 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1670.0, "completions/max_terminated_length": 1670.0, "completions/mean_length": 699.71875, "completions/mean_terminated_length": 699.71875, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "entropy": 0.7384172566235065, "epoch": 0.36281179138321995, "frac_reward_zero_std": 0.375, "grad_norm": 0.60546875, "learning_rate": 6.377551020408163e-07, "loss": -0.0041, "num_tokens": 37721307.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 640, "step_time": 56.23733188863844 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1224.0, "completions/max_terminated_length": 1224.0, "completions/mean_length": 644.109375, "completions/mean_terminated_length": 644.109375, "completions/min_length": 22.0, "completions/min_terminated_length": 22.0, "entropy": 0.7242608647793531, "epoch": 0.3633786848072562, "frac_reward_zero_std": 0.75, "grad_norm": 0.451171875, "learning_rate": 6.3718820861678e-07, "loss": -0.0272, "num_tokens": 37774218.0, "reward": 0.125, "reward_std": 0.4364357888698578, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 641, "step_time": 42.41266290098429 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2104.0, "completions/max_terminated_length": 2104.0, "completions/mean_length": 825.53125, "completions/mean_terminated_length": 825.53125, "completions/min_length": 236.0, "completions/min_terminated_length": 236.0, "entropy": 0.810003375634551, "epoch": 0.36394557823129253, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 6.366213151927437e-07, "loss": -0.0085, "num_tokens": 37838356.0, "reward": 0.3671875, "reward_std": 0.725073516368866, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 642, "step_time": 73.07388027384877 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 803.25, "completions/mean_terminated_length": 803.25, "completions/min_length": 262.0, "completions/min_terminated_length": 262.0, "entropy": 0.7468740493059158, "epoch": 0.3645124716553288, "frac_reward_zero_std": 0.625, "grad_norm": 0.44921875, "learning_rate": 6.360544217687075e-07, "loss": 0.0256, "num_tokens": 37900724.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 643, "step_time": 51.518991655670106 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1696.0, "completions/max_terminated_length": 1696.0, "completions/mean_length": 850.265625, "completions/mean_terminated_length": 850.265625, "completions/min_length": 371.0, "completions/min_terminated_length": 371.0, "entropy": 0.673087963834405, "epoch": 0.36507936507936506, "frac_reward_zero_std": 0.75, "grad_norm": 0.328125, "learning_rate": 6.354875283446711e-07, "loss": -0.0004, "num_tokens": 37964213.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 644, "step_time": 58.71310798358172 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1975.0, "completions/max_terminated_length": 1975.0, "completions/mean_length": 778.71875, "completions/mean_terminated_length": 778.71875, "completions/min_length": 376.0, "completions/min_terminated_length": 376.0, "entropy": 0.9261895325034857, "epoch": 0.3656462585034014, "frac_reward_zero_std": 0.75, "grad_norm": 0.390625, "learning_rate": 6.349206349206349e-07, "loss": 0.004, "num_tokens": 38025019.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 645, "step_time": 68.84365547727793 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1166.0, "completions/max_terminated_length": 1166.0, "completions/mean_length": 729.046875, "completions/mean_terminated_length": 729.046875, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "entropy": 0.8771541491150856, "epoch": 0.36621315192743764, "frac_reward_zero_std": 0.75, "grad_norm": 0.40234375, "learning_rate": 6.343537414965986e-07, "loss": -0.01, "num_tokens": 38083278.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 646, "step_time": 40.352095168083906 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1622.0, "completions/max_terminated_length": 1622.0, "completions/mean_length": 723.34375, "completions/mean_terminated_length": 723.34375, "completions/min_length": 364.0, "completions/min_terminated_length": 364.0, "entropy": 1.0137921087443829, "epoch": 0.3667800453514739, "frac_reward_zero_std": 0.75, "grad_norm": 0.439453125, "learning_rate": 6.337868480725624e-07, "loss": 0.0016, "num_tokens": 38142580.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 647, "step_time": 57.25881090760231 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1098.0, "completions/max_terminated_length": 1098.0, "completions/mean_length": 685.1875, "completions/mean_terminated_length": 685.1875, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "entropy": 0.7464550426229835, "epoch": 0.3673469387755102, "frac_reward_zero_std": 0.625, "grad_norm": 0.515625, "learning_rate": 6.332199546485261e-07, "loss": 0.0115, "num_tokens": 38200488.0, "reward": 0.21875, "reward_std": 0.5764904618263245, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 648, "step_time": 38.424966879189014 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1172.0, "completions/max_terminated_length": 1172.0, "completions/mean_length": 693.640625, "completions/mean_terminated_length": 693.640625, "completions/min_length": 309.0, "completions/min_terminated_length": 309.0, "entropy": 0.8691299445927143, "epoch": 0.3679138321995465, "frac_reward_zero_std": 0.625, "grad_norm": 0.55859375, "learning_rate": 6.326530612244898e-07, "loss": -0.0586, "num_tokens": 38256145.0, "reward": 0.171875, "reward_std": 0.528690755367279, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 649, "step_time": 41.102212926372886 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1296.0, "completions/max_terminated_length": 1296.0, "completions/mean_length": 707.546875, "completions/mean_terminated_length": 707.546875, "completions/min_length": 208.0, "completions/min_terminated_length": 208.0, "entropy": 0.919795224443078, "epoch": 0.36848072562358275, "frac_reward_zero_std": 0.375, "grad_norm": 0.640625, "learning_rate": 6.320861678004535e-07, "loss": 0.0195, "num_tokens": 38313532.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 650, "step_time": 45.144144896417856 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1353.0, "completions/max_terminated_length": 1353.0, "completions/mean_length": 709.59375, "completions/mean_terminated_length": 709.59375, "completions/min_length": 277.0, "completions/min_terminated_length": 277.0, "entropy": 0.7981938906013966, "epoch": 0.36904761904761907, "frac_reward_zero_std": 0.5, "grad_norm": 0.546875, "learning_rate": 6.315192743764171e-07, "loss": -0.0176, "num_tokens": 38370258.0, "reward": 0.40625, "reward_std": 0.75, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 651, "step_time": 47.05556992441416 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1414.0, "completions/max_terminated_length": 1414.0, "completions/mean_length": 764.5, "completions/mean_terminated_length": 764.5, "completions/min_length": 377.0, "completions/min_terminated_length": 377.0, "entropy": 0.7629802487790585, "epoch": 0.36961451247165533, "frac_reward_zero_std": 0.75, "grad_norm": 0.36328125, "learning_rate": 6.309523809523809e-07, "loss": -0.0039, "num_tokens": 38428738.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 652, "step_time": 48.13539344444871 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 655.28125, "completions/mean_terminated_length": 655.28125, "completions/min_length": 266.0, "completions/min_terminated_length": 266.0, "entropy": 0.8418986964970827, "epoch": 0.3701814058956916, "frac_reward_zero_std": 0.375, "grad_norm": 0.671875, "learning_rate": 6.303854875283446e-07, "loss": 0.0087, "num_tokens": 38482532.0, "reward": 0.5390625, "reward_std": 0.89666348695755, "rewards/accuracy_reward_func/mean": 0.5, "rewards/accuracy_reward_func/std": 0.8728715777397156, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 653, "step_time": 38.0635462095961 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1431.0, "completions/max_terminated_length": 1431.0, "completions/mean_length": 750.359375, "completions/mean_terminated_length": 750.359375, "completions/min_length": 9.0, "completions/min_terminated_length": 9.0, "entropy": 0.7542352508753538, "epoch": 0.3707482993197279, "frac_reward_zero_std": 0.25, "grad_norm": 0.64453125, "learning_rate": 6.298185941043084e-07, "loss": -0.0113, "num_tokens": 38541355.0, "reward": 0.4375, "reward_std": 0.8042545318603516, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 654, "step_time": 49.74752385355532 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1817.0, "completions/max_terminated_length": 1817.0, "completions/mean_length": 693.984375, "completions/mean_terminated_length": 693.984375, "completions/min_length": 283.0, "completions/min_terminated_length": 283.0, "entropy": 0.7433206718415022, "epoch": 0.3713151927437642, "frac_reward_zero_std": 0.375, "grad_norm": 0.61328125, "learning_rate": 6.292517006802721e-07, "loss": 0.0248, "num_tokens": 38596466.0, "reward": 0.4140625, "reward_std": 0.7794721722602844, "rewards/accuracy_reward_func/mean": 0.375, "rewards/accuracy_reward_func/std": 0.7867957949638367, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 655, "step_time": 62.506738415919244 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 774.453125, "completions/mean_terminated_length": 774.453125, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "entropy": 0.8884105794131756, "epoch": 0.37188208616780044, "frac_reward_zero_std": 0.5, "grad_norm": 0.51953125, "learning_rate": 6.286848072562359e-07, "loss": -0.0053, "num_tokens": 38656367.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 656, "step_time": 50.6804611440748 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1346.0, "completions/max_terminated_length": 1346.0, "completions/mean_length": 798.75, "completions/mean_terminated_length": 798.75, "completions/min_length": 531.0, "completions/min_terminated_length": 531.0, "entropy": 0.9627105258405209, "epoch": 0.37244897959183676, "frac_reward_zero_std": 0.875, "grad_norm": 0.265625, "learning_rate": 6.281179138321994e-07, "loss": 0.0014, "num_tokens": 38718191.0, "reward": 0.03125, "reward_std": 0.12198750674724579, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 657, "step_time": 46.4745169011876 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1257.0, "completions/max_terminated_length": 1257.0, "completions/mean_length": 671.015625, "completions/mean_terminated_length": 671.015625, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "entropy": 0.6412702202796936, "epoch": 0.373015873015873, "frac_reward_zero_std": 0.5, "grad_norm": 0.50390625, "learning_rate": 6.275510204081632e-07, "loss": 0.0115, "num_tokens": 38770584.0, "reward": 0.40625, "reward_std": 0.75, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 658, "step_time": 43.05835712514818 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1335.0, "completions/max_terminated_length": 1335.0, "completions/mean_length": 660.75, "completions/mean_terminated_length": 660.75, "completions/min_length": 201.0, "completions/min_terminated_length": 201.0, "entropy": 0.9306098241358995, "epoch": 0.3735827664399093, "frac_reward_zero_std": 0.75, "grad_norm": 0.3515625, "learning_rate": 6.269841269841269e-07, "loss": 0.0206, "num_tokens": 38824848.0, "reward": 0.2734375, "reward_std": 0.6542273163795471, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 659, "step_time": 47.23293770570308 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 663.109375, "completions/mean_terminated_length": 663.109375, "completions/min_length": 268.0, "completions/min_terminated_length": 268.0, "entropy": 0.7533489614725113, "epoch": 0.3741496598639456, "frac_reward_zero_std": 0.5, "grad_norm": 0.5859375, "learning_rate": 6.264172335600907e-07, "loss": 0.0452, "num_tokens": 38877199.0, "reward": 0.3046875, "reward_std": 0.664384126663208, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 660, "step_time": 60.841483088210225 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 950.015625, "completions/mean_terminated_length": 950.015625, "completions/min_length": 97.0, "completions/min_terminated_length": 97.0, "entropy": 0.6504107564687729, "epoch": 0.37471655328798187, "frac_reward_zero_std": 0.5, "grad_norm": 0.431640625, "learning_rate": 6.258503401360544e-07, "loss": 0.0196, "num_tokens": 38949080.0, "reward": 0.2109375, "reward_std": 0.576006293296814, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 661, "step_time": 75.70265740901232 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1536.0, "completions/max_terminated_length": 1536.0, "completions/mean_length": 739.921875, "completions/mean_terminated_length": 739.921875, "completions/min_length": 323.0, "completions/min_terminated_length": 323.0, "entropy": 0.7864415291696787, "epoch": 0.37528344671201813, "frac_reward_zero_std": 0.625, "grad_norm": 0.45703125, "learning_rate": 6.252834467120182e-07, "loss": -0.0188, "num_tokens": 39006931.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 662, "step_time": 52.27386832423508 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1872.0, "completions/max_terminated_length": 1872.0, "completions/mean_length": 828.875, "completions/mean_terminated_length": 828.875, "completions/min_length": 380.0, "completions/min_terminated_length": 380.0, "entropy": 0.8523445185273886, "epoch": 0.3758503401360544, "frac_reward_zero_std": 0.75, "grad_norm": 0.375, "learning_rate": 6.247165532879818e-07, "loss": -0.0363, "num_tokens": 39070923.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 663, "step_time": 63.8998099565506 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 973.0, "completions/max_terminated_length": 973.0, "completions/mean_length": 619.65625, "completions/mean_terminated_length": 619.65625, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "entropy": 0.8749851696193218, "epoch": 0.3764172335600907, "frac_reward_zero_std": 0.5, "grad_norm": 0.5546875, "learning_rate": 6.241496598639456e-07, "loss": 0.0023, "num_tokens": 39120301.0, "reward": 0.28125, "reward_std": 0.6538955569267273, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 664, "step_time": 33.00213658250868 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 712.78125, "completions/mean_terminated_length": 712.78125, "completions/min_length": 297.0, "completions/min_terminated_length": 297.0, "entropy": 0.79150197096169, "epoch": 0.376984126984127, "frac_reward_zero_std": 0.5, "grad_norm": 0.55859375, "learning_rate": 6.235827664399092e-07, "loss": 0.0167, "num_tokens": 39175031.0, "reward": 0.3984375, "reward_std": 0.7724404335021973, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 665, "step_time": 39.8501486890018 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 698.96875, "completions/mean_terminated_length": 698.96875, "completions/min_length": 307.0, "completions/min_terminated_length": 307.0, "entropy": 1.027171328663826, "epoch": 0.37755102040816324, "frac_reward_zero_std": 0.25, "grad_norm": 0.7421875, "learning_rate": 6.23015873015873e-07, "loss": -0.0039, "num_tokens": 39232597.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 666, "step_time": 48.249097941443324 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1352.0, "completions/max_terminated_length": 1352.0, "completions/mean_length": 667.390625, "completions/mean_terminated_length": 667.390625, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "entropy": 0.7203786689788103, "epoch": 0.37811791383219956, "frac_reward_zero_std": 0.75, "grad_norm": 0.44140625, "learning_rate": 6.224489795918367e-07, "loss": 0.0172, "num_tokens": 39285686.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 667, "step_time": 45.993100920692086 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2115.0, "completions/max_terminated_length": 2115.0, "completions/mean_length": 832.96875, "completions/mean_terminated_length": 832.96875, "completions/min_length": 309.0, "completions/min_terminated_length": 309.0, "entropy": 0.7268731407821178, "epoch": 0.3786848072562358, "frac_reward_zero_std": 0.75, "grad_norm": 0.302734375, "learning_rate": 6.218820861678005e-07, "loss": -0.0119, "num_tokens": 39352492.0, "reward": 0.15625, "reward_std": 0.4443204402923584, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 668, "step_time": 76.51206331700087 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1736.0, "completions/max_terminated_length": 1736.0, "completions/mean_length": 795.03125, "completions/mean_terminated_length": 795.03125, "completions/min_length": 393.0, "completions/min_terminated_length": 393.0, "entropy": 1.0189481489360332, "epoch": 0.3792517006802721, "frac_reward_zero_std": 0.875, "grad_norm": 0.205078125, "learning_rate": 6.213151927437641e-07, "loss": 0.0107, "num_tokens": 39415294.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 669, "step_time": 61.95392454788089 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 678.34375, "completions/mean_terminated_length": 678.34375, "completions/min_length": 275.0, "completions/min_terminated_length": 275.0, "entropy": 0.7759978659451008, "epoch": 0.3798185941043084, "frac_reward_zero_std": 0.75, "grad_norm": 0.435546875, "learning_rate": 6.207482993197279e-07, "loss": 0.0097, "num_tokens": 39467188.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 670, "step_time": 47.91178299859166 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1682.0, "completions/max_terminated_length": 1682.0, "completions/mean_length": 692.203125, "completions/mean_terminated_length": 692.203125, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 0.7530367765575647, "epoch": 0.38038548752834467, "frac_reward_zero_std": 0.625, "grad_norm": 0.443359375, "learning_rate": 6.201814058956916e-07, "loss": -0.0008, "num_tokens": 39524657.0, "reward": 0.125, "reward_std": 0.4364357888698578, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 671, "step_time": 60.2888721851632 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1403.0, "completions/max_terminated_length": 1403.0, "completions/mean_length": 649.578125, "completions/mean_terminated_length": 649.578125, "completions/min_length": 261.0, "completions/min_terminated_length": 261.0, "entropy": 0.9415013194084167, "epoch": 0.38095238095238093, "frac_reward_zero_std": 0.625, "grad_norm": 0.447265625, "learning_rate": 6.196145124716554e-07, "loss": 0.0153, "num_tokens": 39576438.0, "reward": 0.21875, "reward_std": 0.603396475315094, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 672, "step_time": 47.99685996212065 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 696.28125, "completions/mean_terminated_length": 696.28125, "completions/min_length": 301.0, "completions/min_terminated_length": 301.0, "entropy": 0.9544547498226166, "epoch": 0.38151927437641725, "frac_reward_zero_std": 0.75, "grad_norm": 0.4375, "learning_rate": 6.19047619047619e-07, "loss": -0.0038, "num_tokens": 39630928.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 673, "step_time": 54.064920268021524 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1364.0, "completions/max_terminated_length": 1364.0, "completions/mean_length": 698.296875, "completions/mean_terminated_length": 698.296875, "completions/min_length": 388.0, "completions/min_terminated_length": 388.0, "entropy": 0.8187709338963032, "epoch": 0.3820861678004535, "frac_reward_zero_std": 0.625, "grad_norm": 0.482421875, "learning_rate": 6.184807256235827e-07, "loss": 0.0406, "num_tokens": 39686555.0, "reward": 0.265625, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 674, "step_time": 46.89131363015622 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1380.0, "completions/max_terminated_length": 1380.0, "completions/mean_length": 753.46875, "completions/mean_terminated_length": 753.46875, "completions/min_length": 4.0, "completions/min_terminated_length": 4.0, "entropy": 0.7073788288980722, "epoch": 0.3826530612244898, "frac_reward_zero_std": 0.5, "grad_norm": 0.51171875, "learning_rate": 6.179138321995464e-07, "loss": 0.0104, "num_tokens": 39745129.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 675, "step_time": 47.17899635992944 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1466.0, "completions/max_terminated_length": 1466.0, "completions/mean_length": 805.90625, "completions/mean_terminated_length": 805.90625, "completions/min_length": 377.0, "completions/min_terminated_length": 377.0, "entropy": 0.7654253169894218, "epoch": 0.3832199546485261, "frac_reward_zero_std": 0.5, "grad_norm": 0.55078125, "learning_rate": 6.173469387755101e-07, "loss": -0.0005, "num_tokens": 39806963.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 676, "step_time": 50.05393888242543 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1368.0, "completions/max_terminated_length": 1368.0, "completions/mean_length": 702.984375, "completions/mean_terminated_length": 702.984375, "completions/min_length": 280.0, "completions/min_terminated_length": 280.0, "entropy": 0.8189000021666288, "epoch": 0.38378684807256236, "frac_reward_zero_std": 0.625, "grad_norm": 0.46875, "learning_rate": 6.167800453514739e-07, "loss": -0.0089, "num_tokens": 39863642.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 677, "step_time": 47.03352596703917 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 544.640625, "completions/mean_terminated_length": 544.640625, "completions/min_length": 211.0, "completions/min_terminated_length": 211.0, "entropy": 1.0272091627120972, "epoch": 0.3843537414965986, "frac_reward_zero_std": 0.5, "grad_norm": 0.64453125, "learning_rate": 6.162131519274376e-07, "loss": 0.0088, "num_tokens": 39908995.0, "reward": 0.2109375, "reward_std": 0.576006293296814, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 678, "step_time": 34.143078247085214 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2551.0, "completions/max_terminated_length": 2551.0, "completions/mean_length": 708.828125, "completions/mean_terminated_length": 708.828125, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "entropy": 0.8408221248537302, "epoch": 0.38492063492063494, "frac_reward_zero_std": 0.375, "grad_norm": 0.65234375, "learning_rate": 6.156462585034014e-07, "loss": 0.0159, "num_tokens": 39970608.0, "reward": 0.3046875, "reward_std": 0.6878606677055359, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 679, "step_time": 101.4084204416722 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1166.0, "completions/max_terminated_length": 1166.0, "completions/mean_length": 666.5, "completions/mean_terminated_length": 666.5, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "entropy": 0.720644436776638, "epoch": 0.3854875283446712, "frac_reward_zero_std": 0.25, "grad_norm": 0.68359375, "learning_rate": 6.150793650793651e-07, "loss": 0.009, "num_tokens": 40024016.0, "reward": 0.484375, "reward_std": 0.8402982950210571, "rewards/accuracy_reward_func/mean": 0.4375, "rewards/accuracy_reward_func/std": 0.8333333730697632, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 680, "step_time": 40.70588407944888 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1545.0, "completions/max_terminated_length": 1545.0, "completions/mean_length": 814.765625, "completions/mean_terminated_length": 814.765625, "completions/min_length": 330.0, "completions/min_terminated_length": 330.0, "entropy": 0.5790644586086273, "epoch": 0.38605442176870747, "frac_reward_zero_std": 0.5, "grad_norm": 0.44140625, "learning_rate": 6.145124716553288e-07, "loss": -0.0126, "num_tokens": 40085937.0, "reward": 0.3984375, "reward_std": 0.8506752252578735, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 681, "step_time": 53.31244630552828 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1194.0, "completions/max_terminated_length": 1194.0, "completions/mean_length": 700.921875, "completions/mean_terminated_length": 700.921875, "completions/min_length": 250.0, "completions/min_terminated_length": 250.0, "entropy": 0.8838515989482403, "epoch": 0.3866213151927438, "frac_reward_zero_std": 0.875, "grad_norm": 0.287109375, "learning_rate": 6.139455782312924e-07, "loss": 0.0115, "num_tokens": 40140412.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 682, "step_time": 40.56008638162166 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1335.0, "completions/max_terminated_length": 1335.0, "completions/mean_length": 706.5625, "completions/mean_terminated_length": 706.5625, "completions/min_length": 239.0, "completions/min_terminated_length": 239.0, "entropy": 0.7888701613992453, "epoch": 0.38718820861678005, "frac_reward_zero_std": 0.625, "grad_norm": 0.44140625, "learning_rate": 6.133786848072562e-07, "loss": 0.0037, "num_tokens": 40196552.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 683, "step_time": 45.72881556674838 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 667.0, "completions/mean_terminated_length": 667.0, "completions/min_length": 321.0, "completions/min_terminated_length": 321.0, "entropy": 0.7939537465572357, "epoch": 0.3877551020408163, "frac_reward_zero_std": 0.5, "grad_norm": 0.609375, "learning_rate": 6.128117913832199e-07, "loss": -0.0012, "num_tokens": 40250568.0, "reward": 0.234375, "reward_std": 0.5037065148353577, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.109375, "rewards/format_reward_func/std": 0.2083333432674408, "step": 684, "step_time": 42.73227133601904 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1704.0, "completions/max_terminated_length": 1704.0, "completions/mean_length": 827.609375, "completions/mean_terminated_length": 827.609375, "completions/min_length": 4.0, "completions/min_terminated_length": 4.0, "entropy": 0.6719684153795242, "epoch": 0.38832199546485263, "frac_reward_zero_std": 0.375, "grad_norm": 0.5390625, "learning_rate": 6.122448979591837e-07, "loss": -0.0568, "num_tokens": 40313031.0, "reward": 0.2890625, "reward_std": 0.6655030846595764, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 685, "step_time": 57.538355994969606 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1648.0, "completions/max_terminated_length": 1648.0, "completions/mean_length": 861.515625, "completions/mean_terminated_length": 861.515625, "completions/min_length": 314.0, "completions/min_terminated_length": 314.0, "entropy": 0.5714195594191551, "epoch": 0.3888888888888889, "frac_reward_zero_std": 0.875, "grad_norm": 0.240234375, "learning_rate": 6.116780045351474e-07, "loss": -0.0039, "num_tokens": 40380608.0, "reward": 0.046875, "reward_std": 0.14689241349697113, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 686, "step_time": 57.4830530770123 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 732.8125, "completions/mean_terminated_length": 732.8125, "completions/min_length": 382.0, "completions/min_terminated_length": 382.0, "entropy": 0.8867952115833759, "epoch": 0.38945578231292516, "frac_reward_zero_std": 0.5, "grad_norm": 0.5390625, "learning_rate": 6.111111111111112e-07, "loss": 0.0081, "num_tokens": 40437204.0, "reward": 0.34375, "reward_std": 0.6950790882110596, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 687, "step_time": 39.095103594474494 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 987.0, "completions/max_terminated_length": 987.0, "completions/mean_length": 572.3125, "completions/mean_terminated_length": 572.3125, "completions/min_length": 146.0, "completions/min_terminated_length": 146.0, "entropy": 0.8474721796810627, "epoch": 0.3900226757369615, "frac_reward_zero_std": 0.75, "grad_norm": 0.482421875, "learning_rate": 6.105442176870748e-07, "loss": -0.001, "num_tokens": 40484848.0, "reward": 0.125, "reward_std": 0.37796446681022644, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 688, "step_time": 34.06386910099536 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1576.0, "completions/max_terminated_length": 1576.0, "completions/mean_length": 756.40625, "completions/mean_terminated_length": 756.40625, "completions/min_length": 466.0, "completions/min_terminated_length": 466.0, "entropy": 0.8588584586977959, "epoch": 0.39058956916099774, "frac_reward_zero_std": 0.25, "grad_norm": 0.69140625, "learning_rate": 6.099773242630385e-07, "loss": -0.0057, "num_tokens": 40543778.0, "reward": 0.3984375, "reward_std": 0.7724404335021973, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 689, "step_time": 53.495136791840196 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1401.0, "completions/max_terminated_length": 1401.0, "completions/mean_length": 732.53125, "completions/mean_terminated_length": 732.53125, "completions/min_length": 294.0, "completions/min_terminated_length": 294.0, "entropy": 0.7606502324342728, "epoch": 0.391156462585034, "frac_reward_zero_std": 0.5, "grad_norm": 0.474609375, "learning_rate": 6.094104308390022e-07, "loss": 0.0096, "num_tokens": 40600692.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 690, "step_time": 47.67858843225986 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 734.78125, "completions/mean_terminated_length": 734.78125, "completions/min_length": 351.0, "completions/min_terminated_length": 351.0, "entropy": 0.7689732853323221, "epoch": 0.39172335600907027, "frac_reward_zero_std": 0.625, "grad_norm": 0.388671875, "learning_rate": 6.08843537414966e-07, "loss": 0.0183, "num_tokens": 40657950.0, "reward": 0.140625, "reward_std": 0.4753340482711792, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 691, "step_time": 46.17201771680266 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1816.0, "completions/max_terminated_length": 1816.0, "completions/mean_length": 755.828125, "completions/mean_terminated_length": 755.828125, "completions/min_length": 315.0, "completions/min_terminated_length": 315.0, "entropy": 0.786823371425271, "epoch": 0.3922902494331066, "frac_reward_zero_std": 0.75, "grad_norm": 0.2734375, "learning_rate": 6.082766439909297e-07, "loss": 0.0068, "num_tokens": 40716715.0, "reward": 0.1171875, "reward_std": 0.41539984941482544, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 692, "step_time": 62.187869605608284 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1218.0, "completions/max_terminated_length": 1218.0, "completions/mean_length": 647.65625, "completions/mean_terminated_length": 647.65625, "completions/min_length": 202.0, "completions/min_terminated_length": 202.0, "entropy": 0.9059557132422924, "epoch": 0.39285714285714285, "frac_reward_zero_std": 0.625, "grad_norm": 0.54296875, "learning_rate": 6.077097505668934e-07, "loss": -0.0175, "num_tokens": 40767525.0, "reward": 0.2421875, "reward_std": 0.6900206208229065, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 693, "step_time": 41.2340766787529 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1199.0, "completions/max_terminated_length": 1199.0, "completions/mean_length": 719.21875, "completions/mean_terminated_length": 719.21875, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 0.9956004079431295, "epoch": 0.3934240362811791, "frac_reward_zero_std": 0.625, "grad_norm": 0.515625, "learning_rate": 6.071428571428571e-07, "loss": 0.0411, "num_tokens": 40823899.0, "reward": 0.3203125, "reward_std": 0.7835976481437683, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 694, "step_time": 40.48257102072239 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1414.0, "completions/max_terminated_length": 1414.0, "completions/mean_length": 756.078125, "completions/mean_terminated_length": 756.078125, "completions/min_length": 233.0, "completions/min_terminated_length": 233.0, "entropy": 0.830492103472352, "epoch": 0.39399092970521543, "frac_reward_zero_std": 0.875, "grad_norm": 0.2041015625, "learning_rate": 6.065759637188208e-07, "loss": 0.0043, "num_tokens": 40884488.0, "reward": 0.125, "reward_std": 0.37796446681022644, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 695, "step_time": 49.3948492500931 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1581.0, "completions/max_terminated_length": 1581.0, "completions/mean_length": 739.84375, "completions/mean_terminated_length": 739.84375, "completions/min_length": 407.0, "completions/min_terminated_length": 407.0, "entropy": 0.6636850759387016, "epoch": 0.3945578231292517, "frac_reward_zero_std": 0.375, "grad_norm": 0.55859375, "learning_rate": 6.060090702947846e-07, "loss": -0.0156, "num_tokens": 40943910.0, "reward": 0.40625, "reward_std": 0.7708735466003418, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 696, "step_time": 55.97520697116852 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1650.0, "completions/max_terminated_length": 1650.0, "completions/mean_length": 697.921875, "completions/mean_terminated_length": 697.921875, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "entropy": 0.73609808832407, "epoch": 0.39512471655328796, "frac_reward_zero_std": 0.5, "grad_norm": 0.546875, "learning_rate": 6.054421768707482e-07, "loss": 0.0007, "num_tokens": 40999065.0, "reward": 0.3046875, "reward_std": 0.664384126663208, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 697, "step_time": 55.76855263393372 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1517.0, "completions/max_terminated_length": 1517.0, "completions/mean_length": 726.25, "completions/mean_terminated_length": 726.25, "completions/min_length": 337.0, "completions/min_terminated_length": 337.0, "entropy": 0.784137524664402, "epoch": 0.3956916099773243, "frac_reward_zero_std": 0.5, "grad_norm": 0.5, "learning_rate": 6.04875283446712e-07, "loss": 0.0198, "num_tokens": 41055809.0, "reward": 0.2265625, "reward_std": 0.5904644727706909, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 698, "step_time": 51.082175908610225 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1463.0, "completions/max_terminated_length": 1463.0, "completions/mean_length": 730.53125, "completions/mean_terminated_length": 730.53125, "completions/min_length": 308.0, "completions/min_terminated_length": 308.0, "entropy": 0.7605953272432089, "epoch": 0.39625850340136054, "frac_reward_zero_std": 0.5, "grad_norm": 0.48046875, "learning_rate": 6.043083900226757e-07, "loss": 0.0312, "num_tokens": 41114627.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 699, "step_time": 50.327455514110625 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 744.046875, "completions/mean_terminated_length": 744.046875, "completions/min_length": 319.0, "completions/min_terminated_length": 319.0, "entropy": 0.7873789332807064, "epoch": 0.3968253968253968, "frac_reward_zero_std": 0.75, "grad_norm": 0.3671875, "learning_rate": 6.037414965986394e-07, "loss": -0.0032, "num_tokens": 41172598.0, "reward": 0.1171875, "reward_std": 0.37524792551994324, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 700, "step_time": 50.620650958269835 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 822.015625, "completions/mean_terminated_length": 822.015625, "completions/min_length": 390.0, "completions/min_terminated_length": 390.0, "entropy": 0.6169170867651701, "epoch": 0.3973922902494331, "frac_reward_zero_std": 0.375, "grad_norm": 0.51953125, "learning_rate": 6.031746031746031e-07, "loss": 0.0103, "num_tokens": 41236375.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 701, "step_time": 58.50305821374059 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 702.59375, "completions/mean_terminated_length": 702.59375, "completions/min_length": 244.0, "completions/min_terminated_length": 244.0, "entropy": 0.8059696778655052, "epoch": 0.3979591836734694, "frac_reward_zero_std": 0.5, "grad_norm": 0.56640625, "learning_rate": 6.026077097505669e-07, "loss": 0.0069, "num_tokens": 41293101.0, "reward": 0.1875, "reward_std": 0.5455447435379028, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 702, "step_time": 43.96138078626245 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 648.046875, "completions/mean_terminated_length": 648.046875, "completions/min_length": 280.0, "completions/min_terminated_length": 280.0, "entropy": 0.6702596955001354, "epoch": 0.39852607709750565, "frac_reward_zero_std": 0.5, "grad_norm": 0.55078125, "learning_rate": 6.020408163265306e-07, "loss": -0.0121, "num_tokens": 41346528.0, "reward": 0.375, "reward_std": 0.745356023311615, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 703, "step_time": 42.75202510878444 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1140.0, "completions/max_terminated_length": 1140.0, "completions/mean_length": 687.78125, "completions/mean_terminated_length": 687.78125, "completions/min_length": 251.0, "completions/min_terminated_length": 251.0, "entropy": 0.7277284692972898, "epoch": 0.39909297052154197, "frac_reward_zero_std": 0.375, "grad_norm": 0.63671875, "learning_rate": 6.014739229024944e-07, "loss": 0.0301, "num_tokens": 41401434.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 704, "step_time": 39.50285739637911 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 665.953125, "completions/mean_terminated_length": 665.953125, "completions/min_length": 174.0, "completions/min_terminated_length": 174.0, "entropy": 0.800261564552784, "epoch": 0.39965986394557823, "frac_reward_zero_std": 0.5, "grad_norm": 0.55859375, "learning_rate": 6.00907029478458e-07, "loss": 0.0108, "num_tokens": 41454063.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 705, "step_time": 43.9979402134195 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 719.078125, "completions/mean_terminated_length": 719.078125, "completions/min_length": 339.0, "completions/min_terminated_length": 339.0, "entropy": 0.877237593755126, "epoch": 0.4002267573696145, "frac_reward_zero_std": 0.75, "grad_norm": 0.435546875, "learning_rate": 6.003401360544217e-07, "loss": -0.0067, "num_tokens": 41512276.0, "reward": 0.2734375, "reward_std": 0.701074481010437, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 706, "step_time": 43.27953645866364 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1338.0, "completions/max_terminated_length": 1338.0, "completions/mean_length": 786.453125, "completions/mean_terminated_length": 786.453125, "completions/min_length": 437.0, "completions/min_terminated_length": 437.0, "entropy": 0.6615323107689619, "epoch": 0.4007936507936508, "frac_reward_zero_std": 0.5, "grad_norm": 0.474609375, "learning_rate": 5.997732426303854e-07, "loss": 0.0054, "num_tokens": 41573089.0, "reward": 0.3359375, "reward_std": 0.6961041688919067, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 707, "step_time": 46.12602823320776 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1576.0, "completions/max_terminated_length": 1576.0, "completions/mean_length": 776.859375, "completions/mean_terminated_length": 776.859375, "completions/min_length": 302.0, "completions/min_terminated_length": 302.0, "entropy": 0.7004927899688482, "epoch": 0.4013605442176871, "frac_reward_zero_std": 0.375, "grad_norm": 0.59765625, "learning_rate": 5.992063492063492e-07, "loss": 0.023, "num_tokens": 41632328.0, "reward": 0.4375, "reward_std": 0.774084210395813, "rewards/accuracy_reward_func/mean": 0.375, "rewards/accuracy_reward_func/std": 0.7867957949638367, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 708, "step_time": 54.11020789574832 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 697.984375, "completions/mean_terminated_length": 697.984375, "completions/min_length": 145.0, "completions/min_terminated_length": 145.0, "entropy": 0.8287873566150665, "epoch": 0.40192743764172334, "frac_reward_zero_std": 0.5, "grad_norm": 0.55078125, "learning_rate": 5.986394557823129e-07, "loss": 0.0374, "num_tokens": 41688127.0, "reward": 0.296875, "reward_std": 0.6649904847145081, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 709, "step_time": 42.59092733357102 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1351.0, "completions/max_terminated_length": 1351.0, "completions/mean_length": 758.484375, "completions/mean_terminated_length": 758.484375, "completions/min_length": 392.0, "completions/min_terminated_length": 392.0, "entropy": 0.9395110942423344, "epoch": 0.40249433106575966, "frac_reward_zero_std": 0.75, "grad_norm": 0.373046875, "learning_rate": 5.980725623582767e-07, "loss": -0.0069, "num_tokens": 41746894.0, "reward": 0.125, "reward_std": 0.37796446681022644, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 710, "step_time": 46.619201021268964 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 898.625, "completions/mean_terminated_length": 898.625, "completions/min_length": 363.0, "completions/min_terminated_length": 363.0, "entropy": 0.6419065445661545, "epoch": 0.4030612244897959, "frac_reward_zero_std": 0.375, "grad_norm": 0.486328125, "learning_rate": 5.975056689342404e-07, "loss": 0.0082, "num_tokens": 41813894.0, "reward": 0.4375, "reward_std": 0.8796644806861877, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 711, "step_time": 63.49873128719628 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 734.046875, "completions/mean_terminated_length": 734.046875, "completions/min_length": 266.0, "completions/min_terminated_length": 266.0, "entropy": 1.0022796988487244, "epoch": 0.4036281179138322, "frac_reward_zero_std": 0.75, "grad_norm": 0.423828125, "learning_rate": 5.969387755102042e-07, "loss": 0.003, "num_tokens": 41872089.0, "reward": 0.046875, "reward_std": 0.2630521357059479, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.015625, "rewards/format_reward_func/std": 0.08768405020236969, "step": 712, "step_time": 44.153327917680144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 812.296875, "completions/mean_terminated_length": 812.296875, "completions/min_length": 290.0, "completions/min_terminated_length": 290.0, "entropy": 0.6370504721999168, "epoch": 0.4041950113378685, "frac_reward_zero_std": 0.5, "grad_norm": 0.47265625, "learning_rate": 5.963718820861677e-07, "loss": -0.0073, "num_tokens": 41934788.0, "reward": 0.4765625, "reward_std": 0.763397216796875, "rewards/accuracy_reward_func/mean": 0.375, "rewards/accuracy_reward_func/std": 0.7867957949638367, "rewards/format_reward_func/mean": 0.1015625, "rewards/format_reward_func/std": 0.20275264978408813, "step": 713, "step_time": 101.39591556601226 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1720.0, "completions/max_terminated_length": 1720.0, "completions/mean_length": 701.71875, "completions/mean_terminated_length": 701.71875, "completions/min_length": 156.0, "completions/min_terminated_length": 156.0, "entropy": 0.7805192265659571, "epoch": 0.40476190476190477, "frac_reward_zero_std": 0.625, "grad_norm": 0.5390625, "learning_rate": 5.958049886621314e-07, "loss": 0.0228, "num_tokens": 41992266.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 714, "step_time": 60.487199569121 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 710.25, "completions/mean_terminated_length": 710.25, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "entropy": 0.6370216235518456, "epoch": 0.40532879818594103, "frac_reward_zero_std": 0.5, "grad_norm": 0.4453125, "learning_rate": 5.952380952380952e-07, "loss": 0.0107, "num_tokens": 42047082.0, "reward": 0.375, "reward_std": 0.7867957949638367, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 715, "step_time": 63.83074046112597 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1340.0, "completions/max_terminated_length": 1340.0, "completions/mean_length": 781.703125, "completions/mean_terminated_length": 781.703125, "completions/min_length": 344.0, "completions/min_terminated_length": 344.0, "entropy": 0.6871910858899355, "epoch": 0.40589569160997735, "frac_reward_zero_std": 0.625, "grad_norm": 0.46484375, "learning_rate": 5.946712018140589e-07, "loss": -0.0113, "num_tokens": 42106431.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 716, "step_time": 45.26214159745723 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 712.53125, "completions/mean_terminated_length": 712.53125, "completions/min_length": 281.0, "completions/min_terminated_length": 281.0, "entropy": 0.8684989213943481, "epoch": 0.4064625850340136, "frac_reward_zero_std": 0.875, "grad_norm": 0.330078125, "learning_rate": 5.941043083900227e-07, "loss": 0.0168, "num_tokens": 42163337.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 717, "step_time": 43.39038709178567 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1401.0, "completions/max_terminated_length": 1401.0, "completions/mean_length": 746.5, "completions/mean_terminated_length": 746.5, "completions/min_length": 267.0, "completions/min_terminated_length": 267.0, "entropy": 0.577209796756506, "epoch": 0.4070294784580499, "frac_reward_zero_std": 0.375, "grad_norm": 0.494140625, "learning_rate": 5.935374149659864e-07, "loss": 0.0105, "num_tokens": 42220441.0, "reward": 0.2265625, "reward_std": 0.5904644727706909, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 718, "step_time": 48.270371528342366 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 654.03125, "completions/mean_terminated_length": 654.03125, "completions/min_length": 251.0, "completions/min_terminated_length": 251.0, "entropy": 0.8232200630009174, "epoch": 0.4075963718820862, "frac_reward_zero_std": 0.625, "grad_norm": 0.435546875, "learning_rate": 5.929705215419501e-07, "loss": 0.02, "num_tokens": 42273035.0, "reward": 0.28125, "reward_std": 0.6538955569267273, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 719, "step_time": 41.617827150039375 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1354.0, "completions/max_terminated_length": 1354.0, "completions/mean_length": 725.734375, "completions/mean_terminated_length": 725.734375, "completions/min_length": 419.0, "completions/min_terminated_length": 419.0, "entropy": 0.7096963319927454, "epoch": 0.40816326530612246, "frac_reward_zero_std": 0.625, "grad_norm": 0.40234375, "learning_rate": 5.924036281179138e-07, "loss": 0.0038, "num_tokens": 42334130.0, "reward": 0.25, "reward_std": 0.6172134280204773, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 720, "step_time": 47.07885314244777 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1600.0, "completions/max_terminated_length": 1600.0, "completions/mean_length": 808.625, "completions/mean_terminated_length": 808.625, "completions/min_length": 290.0, "completions/min_terminated_length": 290.0, "entropy": 0.62769565731287, "epoch": 0.4087301587301587, "frac_reward_zero_std": 0.875, "grad_norm": 0.28125, "learning_rate": 5.918367346938775e-07, "loss": -0.004, "num_tokens": 42395810.0, "reward": 0.0546875, "reward_std": 0.15728822350502014, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 721, "step_time": 54.02559690643102 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 740.375, "completions/mean_terminated_length": 740.375, "completions/min_length": 366.0, "completions/min_terminated_length": 366.0, "entropy": 0.8444711156189442, "epoch": 0.409297052154195, "frac_reward_zero_std": 0.625, "grad_norm": 0.431640625, "learning_rate": 5.912698412698412e-07, "loss": -0.0021, "num_tokens": 42454954.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 722, "step_time": 43.73231762833893 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1242.0, "completions/max_terminated_length": 1242.0, "completions/mean_length": 728.609375, "completions/mean_terminated_length": 728.609375, "completions/min_length": 298.0, "completions/min_terminated_length": 298.0, "entropy": 0.8367441222071648, "epoch": 0.4098639455782313, "frac_reward_zero_std": 0.75, "grad_norm": 0.310546875, "learning_rate": 5.90702947845805e-07, "loss": 0.013, "num_tokens": 42519009.0, "reward": 0.1484375, "reward_std": 0.47709178924560547, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 723, "step_time": 45.36817385815084 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1386.0, "completions/max_terminated_length": 1386.0, "completions/mean_length": 816.359375, "completions/mean_terminated_length": 816.359375, "completions/min_length": 313.0, "completions/min_terminated_length": 313.0, "entropy": 0.7131411526352167, "epoch": 0.41043083900226757, "frac_reward_zero_std": 0.375, "grad_norm": 0.5, "learning_rate": 5.901360544217687e-07, "loss": 0.004, "num_tokens": 42582088.0, "reward": 0.28125, "reward_std": 0.6777355074882507, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 724, "step_time": 46.9804265871644 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 737.046875, "completions/mean_terminated_length": 737.046875, "completions/min_length": 375.0, "completions/min_terminated_length": 375.0, "entropy": 0.7451744675636292, "epoch": 0.41099773242630383, "frac_reward_zero_std": 0.75, "grad_norm": 0.6484375, "learning_rate": 5.895691609977324e-07, "loss": 0.0135, "num_tokens": 42641243.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 725, "step_time": 52.16097322292626 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 705.15625, "completions/mean_terminated_length": 705.15625, "completions/min_length": 339.0, "completions/min_terminated_length": 339.0, "entropy": 0.6900949291884899, "epoch": 0.41156462585034015, "frac_reward_zero_std": 0.375, "grad_norm": 0.62109375, "learning_rate": 5.890022675736961e-07, "loss": 0.0551, "num_tokens": 42696653.0, "reward": 0.3671875, "reward_std": 0.7880163192749023, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 726, "step_time": 38.13649579510093 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1466.0, "completions/max_terminated_length": 1466.0, "completions/mean_length": 735.984375, "completions/mean_terminated_length": 735.984375, "completions/min_length": 235.0, "completions/min_terminated_length": 235.0, "entropy": 0.8310915026813745, "epoch": 0.4121315192743764, "frac_reward_zero_std": 0.625, "grad_norm": 0.419921875, "learning_rate": 5.884353741496599e-07, "loss": 0.0281, "num_tokens": 42753748.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 727, "step_time": 49.61503871437162 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 731.25, "completions/mean_terminated_length": 731.25, "completions/min_length": 383.0, "completions/min_terminated_length": 383.0, "entropy": 0.9786618016660213, "epoch": 0.4126984126984127, "frac_reward_zero_std": 0.75, "grad_norm": 0.412109375, "learning_rate": 5.878684807256236e-07, "loss": 0.0006, "num_tokens": 42813036.0, "reward": 0.15625, "reward_std": 0.5107961297035217, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 728, "step_time": 46.887156607583165 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1585.0, "completions/max_terminated_length": 1585.0, "completions/mean_length": 759.34375, "completions/mean_terminated_length": 759.34375, "completions/min_length": 265.0, "completions/min_terminated_length": 265.0, "entropy": 0.919614166021347, "epoch": 0.413265306122449, "frac_reward_zero_std": 0.625, "grad_norm": 0.486328125, "learning_rate": 5.873015873015873e-07, "loss": 0.0107, "num_tokens": 42875178.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 729, "step_time": 55.13844683393836 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1396.0, "completions/max_terminated_length": 1396.0, "completions/mean_length": 728.9375, "completions/mean_terminated_length": 728.9375, "completions/min_length": 295.0, "completions/min_terminated_length": 295.0, "entropy": 0.8809193521738052, "epoch": 0.41383219954648526, "frac_reward_zero_std": 0.625, "grad_norm": 0.44921875, "learning_rate": 5.86734693877551e-07, "loss": 0.0294, "num_tokens": 42931734.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 730, "step_time": 48.121805551461875 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1328.0, "completions/max_terminated_length": 1328.0, "completions/mean_length": 667.140625, "completions/mean_terminated_length": 667.140625, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "entropy": 0.7360779270529747, "epoch": 0.4143990929705215, "frac_reward_zero_std": 0.5, "grad_norm": 0.62109375, "learning_rate": 5.861678004535146e-07, "loss": -0.001, "num_tokens": 42985119.0, "reward": 0.265625, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 731, "step_time": 45.864870502613485 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1301.0, "completions/max_terminated_length": 1301.0, "completions/mean_length": 739.34375, "completions/mean_terminated_length": 739.34375, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "entropy": 0.6361839286983013, "epoch": 0.41496598639455784, "frac_reward_zero_std": 0.25, "grad_norm": 0.7109375, "learning_rate": 5.856009070294784e-07, "loss": 0.0104, "num_tokens": 43043989.0, "reward": 0.4609375, "reward_std": 0.7983350157737732, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 732, "step_time": 44.30827684234828 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1193.0, "completions/max_terminated_length": 1193.0, "completions/mean_length": 679.5625, "completions/mean_terminated_length": 679.5625, "completions/min_length": 326.0, "completions/min_terminated_length": 326.0, "entropy": 1.0744093619287014, "epoch": 0.4155328798185941, "frac_reward_zero_std": 0.5, "grad_norm": 0.66015625, "learning_rate": 5.850340136054421e-07, "loss": 0.0066, "num_tokens": 43098857.0, "reward": 0.2109375, "reward_std": 0.6029338836669922, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 733, "step_time": 40.71691221278161 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 688.671875, "completions/mean_terminated_length": 688.671875, "completions/min_length": 370.0, "completions/min_terminated_length": 370.0, "entropy": 0.9368883110582829, "epoch": 0.41609977324263037, "frac_reward_zero_std": 0.5, "grad_norm": 0.53125, "learning_rate": 5.844671201814059e-07, "loss": 0.0075, "num_tokens": 43153788.0, "reward": 0.265625, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 734, "step_time": 37.3967783851549 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1316.0, "completions/max_terminated_length": 1316.0, "completions/mean_length": 795.484375, "completions/mean_terminated_length": 795.484375, "completions/min_length": 407.0, "completions/min_terminated_length": 407.0, "entropy": 0.8395189568400383, "epoch": 0.4166666666666667, "frac_reward_zero_std": 0.875, "grad_norm": 0.294921875, "learning_rate": 5.839002267573696e-07, "loss": -0.007, "num_tokens": 43214323.0, "reward": 0.125, "reward_std": 0.37796446681022644, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 735, "step_time": 45.11113391537219 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1152.0, "completions/max_terminated_length": 1152.0, "completions/mean_length": 792.1875, "completions/mean_terminated_length": 792.1875, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 0.8356929533183575, "epoch": 0.41723356009070295, "frac_reward_zero_std": 0.625, "grad_norm": 0.470703125, "learning_rate": 5.833333333333334e-07, "loss": 0.0282, "num_tokens": 43276127.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 736, "step_time": 40.03395030833781 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1551.0, "completions/max_terminated_length": 1551.0, "completions/mean_length": 724.71875, "completions/mean_terminated_length": 724.71875, "completions/min_length": 381.0, "completions/min_terminated_length": 381.0, "entropy": 0.735107097774744, "epoch": 0.4178004535147392, "frac_reward_zero_std": 0.625, "grad_norm": 0.47265625, "learning_rate": 5.82766439909297e-07, "loss": -0.0089, "num_tokens": 43336573.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 737, "step_time": 55.217949592508376 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 751.375, "completions/mean_terminated_length": 751.375, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "entropy": 0.8345095999538898, "epoch": 0.41836734693877553, "frac_reward_zero_std": 0.25, "grad_norm": 0.69140625, "learning_rate": 5.821995464852607e-07, "loss": 0.0365, "num_tokens": 43394461.0, "reward": 0.5546875, "reward_std": 0.8734041452407837, "rewards/accuracy_reward_func/mean": 0.5, "rewards/accuracy_reward_func/std": 0.8728715777397156, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 738, "step_time": 48.906748646870255 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1473.0, "completions/max_terminated_length": 1473.0, "completions/mean_length": 708.40625, "completions/mean_terminated_length": 708.40625, "completions/min_length": 368.0, "completions/min_terminated_length": 368.0, "entropy": 0.8683152124285698, "epoch": 0.4189342403628118, "frac_reward_zero_std": 0.5, "grad_norm": 0.5625, "learning_rate": 5.816326530612244e-07, "loss": -0.0016, "num_tokens": 43453271.0, "reward": 0.28125, "reward_std": 0.665922224521637, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 739, "step_time": 51.43727010767907 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1472.0, "completions/max_terminated_length": 1472.0, "completions/mean_length": 810.96875, "completions/mean_terminated_length": 810.96875, "completions/min_length": 353.0, "completions/min_terminated_length": 353.0, "entropy": 0.6964520383626223, "epoch": 0.41950113378684806, "frac_reward_zero_std": 0.75, "grad_norm": 0.3125, "learning_rate": 5.810657596371882e-07, "loss": -0.0012, "num_tokens": 43514821.0, "reward": 0.15625, "reward_std": 0.5107961297035217, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 740, "step_time": 49.67749553266913 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1840.0, "completions/mean_length": 845.1875, "completions/mean_terminated_length": 793.5873413085938, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 0.5888213403522968, "epoch": 0.4200680272108844, "frac_reward_zero_std": 0.5, "grad_norm": 0.47265625, "learning_rate": 5.804988662131519e-07, "loss": 0.0107, "num_tokens": 43579161.0, "reward": 0.2265625, "reward_std": 0.5904644727706909, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 741, "step_time": 168.525633610785 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1242.0, "completions/max_terminated_length": 1242.0, "completions/mean_length": 743.546875, "completions/mean_terminated_length": 743.546875, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "entropy": 0.5299877347424626, "epoch": 0.42063492063492064, "frac_reward_zero_std": 0.375, "grad_norm": 0.515625, "learning_rate": 5.799319727891157e-07, "loss": 0.0112, "num_tokens": 43635908.0, "reward": 0.28125, "reward_std": 0.665922224521637, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 742, "step_time": 42.83841699268669 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1542.0, "completions/max_terminated_length": 1542.0, "completions/mean_length": 720.21875, "completions/mean_terminated_length": 720.21875, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "entropy": 1.0659764669835567, "epoch": 0.4212018140589569, "frac_reward_zero_std": 0.625, "grad_norm": 0.498046875, "learning_rate": 5.793650793650794e-07, "loss": 0.0152, "num_tokens": 43693034.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 743, "step_time": 53.25322303455323 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1190.0, "completions/max_terminated_length": 1190.0, "completions/mean_length": 750.3125, "completions/mean_terminated_length": 750.3125, "completions/min_length": 372.0, "completions/min_terminated_length": 372.0, "entropy": 0.6504237297922373, "epoch": 0.4217687074829932, "frac_reward_zero_std": 0.75, "grad_norm": 0.35546875, "learning_rate": 5.78798185941043e-07, "loss": -0.0099, "num_tokens": 43751678.0, "reward": 0.28125, "reward_std": 0.7007648944854736, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 744, "step_time": 41.41377501748502 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2257.0, "completions/max_terminated_length": 2257.0, "completions/mean_length": 768.09375, "completions/mean_terminated_length": 768.09375, "completions/min_length": 313.0, "completions/min_terminated_length": 313.0, "entropy": 0.694596691057086, "epoch": 0.4223356009070295, "frac_reward_zero_std": 0.625, "grad_norm": 0.5078125, "learning_rate": 5.782312925170067e-07, "loss": 0.06, "num_tokens": 43811748.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 745, "step_time": 79.29288134723902 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 838.765625, "completions/mean_terminated_length": 838.765625, "completions/min_length": 358.0, "completions/min_terminated_length": 358.0, "entropy": 0.5027836207300425, "epoch": 0.42290249433106575, "frac_reward_zero_std": 0.375, "grad_norm": 0.466796875, "learning_rate": 5.776643990929705e-07, "loss": 0.0122, "num_tokens": 43876501.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 746, "step_time": 49.38363088015467 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1466.0, "completions/max_terminated_length": 1466.0, "completions/mean_length": 795.109375, "completions/mean_terminated_length": 795.109375, "completions/min_length": 349.0, "completions/min_terminated_length": 349.0, "entropy": 0.9084397703409195, "epoch": 0.42346938775510207, "frac_reward_zero_std": 0.5, "grad_norm": 0.53125, "learning_rate": 5.770975056689342e-07, "loss": -0.0214, "num_tokens": 43937892.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 747, "step_time": 50.44422111287713 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1628.0, "completions/max_terminated_length": 1628.0, "completions/mean_length": 747.75, "completions/mean_terminated_length": 747.75, "completions/min_length": 380.0, "completions/min_terminated_length": 380.0, "entropy": 0.6967819910496473, "epoch": 0.42403628117913833, "frac_reward_zero_std": 0.625, "grad_norm": 0.443359375, "learning_rate": 5.76530612244898e-07, "loss": 0.0251, "num_tokens": 43999332.0, "reward": 0.3359375, "reward_std": 0.7185450792312622, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 748, "step_time": 57.57845904864371 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1853.0, "completions/max_terminated_length": 1853.0, "completions/mean_length": 806.46875, "completions/mean_terminated_length": 806.46875, "completions/min_length": 425.0, "completions/min_terminated_length": 425.0, "entropy": 0.8299741912633181, "epoch": 0.4246031746031746, "frac_reward_zero_std": 0.75, "grad_norm": 0.376953125, "learning_rate": 5.759637188208617e-07, "loss": 0.0229, "num_tokens": 44061122.0, "reward": 0.1171875, "reward_std": 0.46922925114631653, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 749, "step_time": 63.8022293895483 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2944.0, "completions/max_terminated_length": 2944.0, "completions/mean_length": 783.640625, "completions/mean_terminated_length": 783.640625, "completions/min_length": 272.0, "completions/min_terminated_length": 272.0, "entropy": 0.960544366389513, "epoch": 0.42517006802721086, "frac_reward_zero_std": 0.625, "grad_norm": 0.48046875, "learning_rate": 5.753968253968254e-07, "loss": 0.0002, "num_tokens": 44127067.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 750, "step_time": 118.26526685990393 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1354.0, "completions/max_terminated_length": 1354.0, "completions/mean_length": 724.109375, "completions/mean_terminated_length": 724.109375, "completions/min_length": 222.0, "completions/min_terminated_length": 222.0, "entropy": 0.9830201044678688, "epoch": 0.4257369614512472, "frac_reward_zero_std": 0.625, "grad_norm": 0.486328125, "learning_rate": 5.748299319727891e-07, "loss": 0.0244, "num_tokens": 44184498.0, "reward": 0.2734375, "reward_std": 0.6780555844306946, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 751, "step_time": 45.99378525465727 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1508.0, "completions/max_terminated_length": 1508.0, "completions/mean_length": 784.0625, "completions/mean_terminated_length": 784.0625, "completions/min_length": 397.0, "completions/min_terminated_length": 397.0, "entropy": 0.8110213577747345, "epoch": 0.42630385487528344, "frac_reward_zero_std": 0.375, "grad_norm": 0.5703125, "learning_rate": 5.742630385487527e-07, "loss": 0.0261, "num_tokens": 44246126.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 752, "step_time": 52.3525614021346 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 665.4375, "completions/mean_terminated_length": 665.4375, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "entropy": 0.8430373854935169, "epoch": 0.4268707482993197, "frac_reward_zero_std": 0.625, "grad_norm": 0.400390625, "learning_rate": 5.736961451247165e-07, "loss": 0.032, "num_tokens": 44299570.0, "reward": 0.21875, "reward_std": 0.5764904618263245, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 753, "step_time": 42.80218448769301 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1264.0, "completions/max_terminated_length": 1264.0, "completions/mean_length": 743.78125, "completions/mean_terminated_length": 743.78125, "completions/min_length": 4.0, "completions/min_terminated_length": 4.0, "entropy": 0.8105163779109716, "epoch": 0.427437641723356, "frac_reward_zero_std": 0.5, "grad_norm": 0.546875, "learning_rate": 5.731292517006802e-07, "loss": -0.0031, "num_tokens": 44357388.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 754, "step_time": 42.49784342665225 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 701.390625, "completions/mean_terminated_length": 701.390625, "completions/min_length": 194.0, "completions/min_terminated_length": 194.0, "entropy": 0.668464969843626, "epoch": 0.4280045351473923, "frac_reward_zero_std": 0.625, "grad_norm": 0.4453125, "learning_rate": 5.72562358276644e-07, "loss": 0.0018, "num_tokens": 44411381.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 755, "step_time": 54.76838449109346 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 645.34375, "completions/mean_terminated_length": 645.34375, "completions/min_length": 330.0, "completions/min_terminated_length": 330.0, "entropy": 0.9178903736174107, "epoch": 0.42857142857142855, "frac_reward_zero_std": 0.25, "grad_norm": 0.8125, "learning_rate": 5.719954648526076e-07, "loss": -0.0107, "num_tokens": 44461635.0, "reward": 0.3984375, "reward_std": 0.7516103982925415, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 756, "step_time": 38.94769489020109 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 675.90625, "completions/mean_terminated_length": 675.90625, "completions/min_length": 4.0, "completions/min_terminated_length": 4.0, "entropy": 1.0502110309898853, "epoch": 0.42913832199546487, "frac_reward_zero_std": 0.875, "grad_norm": 0.298828125, "learning_rate": 5.714285714285714e-07, "loss": -0.0109, "num_tokens": 44515069.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 757, "step_time": 42.03664774727076 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1222.0, "completions/max_terminated_length": 1222.0, "completions/mean_length": 719.71875, "completions/mean_terminated_length": 719.71875, "completions/min_length": 14.0, "completions/min_terminated_length": 14.0, "entropy": 0.7671445775777102, "epoch": 0.42970521541950113, "frac_reward_zero_std": 0.375, "grad_norm": 0.59765625, "learning_rate": 5.708616780045351e-07, "loss": -0.0067, "num_tokens": 44570131.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 758, "step_time": 40.8751755123958 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1571.0, "completions/max_terminated_length": 1571.0, "completions/mean_length": 829.53125, "completions/mean_terminated_length": 829.53125, "completions/min_length": 366.0, "completions/min_terminated_length": 366.0, "entropy": 0.7642947994172573, "epoch": 0.4302721088435374, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 5.702947845804989e-07, "loss": 0.0352, "num_tokens": 44634093.0, "reward": 0.296875, "reward_std": 0.6649904847145081, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 759, "step_time": 54.131978866644204 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 744.859375, "completions/mean_terminated_length": 744.859375, "completions/min_length": 349.0, "completions/min_terminated_length": 349.0, "entropy": 0.95208290964365, "epoch": 0.4308390022675737, "frac_reward_zero_std": 0.625, "grad_norm": 0.462890625, "learning_rate": 5.697278911564625e-07, "loss": -0.0062, "num_tokens": 44692580.0, "reward": 0.15625, "reward_std": 0.49501484632492065, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 760, "step_time": 45.60193131119013 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1383.0, "completions/max_terminated_length": 1383.0, "completions/mean_length": 820.046875, "completions/mean_terminated_length": 820.046875, "completions/min_length": 370.0, "completions/min_terminated_length": 370.0, "entropy": 0.7668710723519325, "epoch": 0.43140589569161, "frac_reward_zero_std": 0.75, "grad_norm": 0.3203125, "learning_rate": 5.691609977324263e-07, "loss": -0.0045, "num_tokens": 44759479.0, "reward": 0.078125, "reward_std": 0.2847827076911926, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 761, "step_time": 49.02586789615452 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 860.4375, "completions/mean_terminated_length": 860.4375, "completions/min_length": 480.0, "completions/min_terminated_length": 480.0, "entropy": 0.725027572363615, "epoch": 0.43197278911564624, "frac_reward_zero_std": 0.375, "grad_norm": 0.515625, "learning_rate": 5.685941043083899e-07, "loss": 0.0034, "num_tokens": 44824139.0, "reward": 0.3359375, "reward_std": 0.7185450792312622, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 762, "step_time": 54.259508294053376 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1319.0, "completions/max_terminated_length": 1319.0, "completions/mean_length": 729.15625, "completions/mean_terminated_length": 729.15625, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "entropy": 0.789761122316122, "epoch": 0.43253968253968256, "frac_reward_zero_std": 0.625, "grad_norm": 0.4609375, "learning_rate": 5.680272108843537e-07, "loss": -0.0118, "num_tokens": 44882285.0, "reward": 0.2734375, "reward_std": 0.6780555844306946, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 763, "step_time": 45.12244441639632 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1804.0, "completions/max_terminated_length": 1804.0, "completions/mean_length": 765.96875, "completions/mean_terminated_length": 765.96875, "completions/min_length": 186.0, "completions/min_terminated_length": 186.0, "entropy": 0.8189289644360542, "epoch": 0.4331065759637188, "frac_reward_zero_std": 0.625, "grad_norm": 0.37109375, "learning_rate": 5.674603174603174e-07, "loss": 0.0002, "num_tokens": 44941499.0, "reward": 0.1796875, "reward_std": 0.5298036932945251, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 764, "step_time": 61.34843339771032 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1769.0, "completions/max_terminated_length": 1769.0, "completions/mean_length": 747.53125, "completions/mean_terminated_length": 747.53125, "completions/min_length": 430.0, "completions/min_terminated_length": 430.0, "entropy": 0.8964242096990347, "epoch": 0.4336734693877551, "frac_reward_zero_std": 0.625, "grad_norm": 0.5, "learning_rate": 5.668934240362812e-07, "loss": 0.0091, "num_tokens": 45001397.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 765, "step_time": 60.81541761662811 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 995.0, "completions/max_terminated_length": 995.0, "completions/mean_length": 586.203125, "completions/mean_terminated_length": 586.203125, "completions/min_length": 162.0, "completions/min_terminated_length": 162.0, "entropy": 1.0916917622089386, "epoch": 0.4342403628117914, "frac_reward_zero_std": 0.625, "grad_norm": 0.6328125, "learning_rate": 5.663265306122449e-07, "loss": 0.0114, "num_tokens": 45050914.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 766, "step_time": 34.052269238978624 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1876.0, "completions/max_terminated_length": 1876.0, "completions/mean_length": 671.640625, "completions/mean_terminated_length": 671.640625, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "entropy": 0.8228061571717262, "epoch": 0.43480725623582767, "frac_reward_zero_std": 0.625, "grad_norm": 0.4453125, "learning_rate": 5.657596371882087e-07, "loss": 0.0232, "num_tokens": 45103571.0, "reward": 0.375, "reward_std": 0.7237468957901001, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 767, "step_time": 63.753929750062525 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2744.0, "completions/max_terminated_length": 2744.0, "completions/mean_length": 776.796875, "completions/mean_terminated_length": 776.796875, "completions/min_length": 262.0, "completions/min_terminated_length": 262.0, "entropy": 0.6937824469059706, "epoch": 0.43537414965986393, "frac_reward_zero_std": 0.5, "grad_norm": 0.4921875, "learning_rate": 5.651927437641723e-07, "loss": 0.0839, "num_tokens": 45163486.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 768, "step_time": 99.16287459339947 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1625.0, "completions/max_terminated_length": 1625.0, "completions/mean_length": 919.171875, "completions/mean_terminated_length": 919.171875, "completions/min_length": 533.0, "completions/min_terminated_length": 533.0, "entropy": 0.5660981275141239, "epoch": 0.43594104308390025, "frac_reward_zero_std": 0.5, "grad_norm": 0.416015625, "learning_rate": 5.64625850340136e-07, "loss": 0.0097, "num_tokens": 45231185.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 769, "step_time": 55.40216701198369 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1838.0, "completions/max_terminated_length": 1838.0, "completions/mean_length": 738.65625, "completions/mean_terminated_length": 738.65625, "completions/min_length": 197.0, "completions/min_terminated_length": 197.0, "entropy": 0.6767476070672274, "epoch": 0.4365079365079365, "frac_reward_zero_std": 0.375, "grad_norm": 0.57421875, "learning_rate": 5.640589569160997e-07, "loss": 0.015, "num_tokens": 45289083.0, "reward": 0.359375, "reward_std": 0.7478474378585815, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 770, "step_time": 64.3451043413952 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1220.0, "completions/max_terminated_length": 1220.0, "completions/mean_length": 698.328125, "completions/mean_terminated_length": 698.328125, "completions/min_length": 282.0, "completions/min_terminated_length": 282.0, "entropy": 1.005146386101842, "epoch": 0.4370748299319728, "frac_reward_zero_std": 0.875, "grad_norm": 0.359375, "learning_rate": 5.634920634920635e-07, "loss": -0.0048, "num_tokens": 45342728.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 771, "step_time": 41.5515520805493 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1205.0, "completions/max_terminated_length": 1205.0, "completions/mean_length": 756.921875, "completions/mean_terminated_length": 756.921875, "completions/min_length": 354.0, "completions/min_terminated_length": 354.0, "entropy": 0.7958099376410246, "epoch": 0.4376417233560091, "frac_reward_zero_std": 0.5, "grad_norm": 0.578125, "learning_rate": 5.629251700680272e-07, "loss": -0.0216, "num_tokens": 45400107.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 772, "step_time": 41.31011986825615 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1242.0, "completions/max_terminated_length": 1242.0, "completions/mean_length": 751.96875, "completions/mean_terminated_length": 751.96875, "completions/min_length": 392.0, "completions/min_terminated_length": 392.0, "entropy": 0.9773392900824547, "epoch": 0.43820861678004536, "frac_reward_zero_std": 0.75, "grad_norm": 0.421875, "learning_rate": 5.62358276643991e-07, "loss": -0.0111, "num_tokens": 45458601.0, "reward": 0.0703125, "reward_std": 0.27984103560447693, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 773, "step_time": 42.69651668332517 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1407.0, "completions/max_terminated_length": 1407.0, "completions/mean_length": 806.125, "completions/mean_terminated_length": 806.125, "completions/min_length": 338.0, "completions/min_terminated_length": 338.0, "entropy": 0.9133709631860256, "epoch": 0.4387755102040816, "frac_reward_zero_std": 0.5, "grad_norm": 0.55859375, "learning_rate": 5.617913832199547e-07, "loss": -0.0011, "num_tokens": 45522753.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 774, "step_time": 49.31818789802492 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 812.515625, "completions/mean_terminated_length": 812.515625, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "entropy": 0.5701410993933678, "epoch": 0.43934240362811794, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 5.612244897959183e-07, "loss": 0.0478, "num_tokens": 45586578.0, "reward": 0.53125, "reward_std": 0.8351171612739563, "rewards/accuracy_reward_func/mean": 0.46875, "rewards/accuracy_reward_func/std": 0.8539125919342041, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 775, "step_time": 47.40858102682978 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 826.625, "completions/mean_terminated_length": 826.625, "completions/min_length": 364.0, "completions/min_terminated_length": 364.0, "entropy": 0.7575418595224619, "epoch": 0.4399092970521542, "frac_reward_zero_std": 0.375, "grad_norm": 0.53125, "learning_rate": 5.60657596371882e-07, "loss": 0.0051, "num_tokens": 45652490.0, "reward": 0.2734375, "reward_std": 0.6780555844306946, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 776, "step_time": 52.221214167773724 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1353.0, "completions/max_terminated_length": 1353.0, "completions/mean_length": 707.953125, "completions/mean_terminated_length": 707.953125, "completions/min_length": 15.0, "completions/min_terminated_length": 15.0, "entropy": 0.887224867939949, "epoch": 0.44047619047619047, "frac_reward_zero_std": 0.625, "grad_norm": 0.482421875, "learning_rate": 5.600907029478457e-07, "loss": 0.0102, "num_tokens": 45708751.0, "reward": 0.1875, "reward_std": 0.5, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 777, "step_time": 46.73268238455057 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1164.0, "completions/max_terminated_length": 1164.0, "completions/mean_length": 673.015625, "completions/mean_terminated_length": 673.015625, "completions/min_length": 9.0, "completions/min_terminated_length": 9.0, "entropy": 0.8263971842825413, "epoch": 0.4410430839002268, "frac_reward_zero_std": 0.5, "grad_norm": 0.6171875, "learning_rate": 5.595238095238095e-07, "loss": 0.0043, "num_tokens": 45761592.0, "reward": 0.3828125, "reward_std": 0.7545743584632874, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 778, "step_time": 39.5863815927878 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1420.0, "completions/max_terminated_length": 1420.0, "completions/mean_length": 768.234375, "completions/mean_terminated_length": 768.234375, "completions/min_length": 130.0, "completions/min_terminated_length": 130.0, "entropy": 0.6901191417127848, "epoch": 0.44160997732426305, "frac_reward_zero_std": 0.25, "grad_norm": 0.59375, "learning_rate": 5.589569160997732e-07, "loss": -0.0535, "num_tokens": 45819247.0, "reward": 0.2890625, "reward_std": 0.6655030846595764, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 779, "step_time": 48.00127135589719 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1542.0, "completions/max_terminated_length": 1542.0, "completions/mean_length": 713.59375, "completions/mean_terminated_length": 713.59375, "completions/min_length": 114.0, "completions/min_terminated_length": 114.0, "entropy": 0.7308294903486967, "epoch": 0.4421768707482993, "frac_reward_zero_std": 0.625, "grad_norm": 0.439453125, "learning_rate": 5.58390022675737e-07, "loss": -0.0285, "num_tokens": 45875405.0, "reward": 0.375, "reward_std": 0.7237468957901001, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 780, "step_time": 52.49343588668853 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 807.234375, "completions/mean_terminated_length": 807.234375, "completions/min_length": 412.0, "completions/min_terminated_length": 412.0, "entropy": 0.7114958334714174, "epoch": 0.4427437641723356, "frac_reward_zero_std": 0.5, "grad_norm": 0.482421875, "learning_rate": 5.578231292517006e-07, "loss": 0.0114, "num_tokens": 45936076.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 781, "step_time": 42.76526521053165 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1711.0, "completions/max_terminated_length": 1711.0, "completions/mean_length": 731.65625, "completions/mean_terminated_length": 731.65625, "completions/min_length": 239.0, "completions/min_terminated_length": 239.0, "entropy": 0.8945471830666065, "epoch": 0.4433106575963719, "frac_reward_zero_std": 0.625, "grad_norm": 0.5625, "learning_rate": 5.572562358276644e-07, "loss": -0.0008, "num_tokens": 45993326.0, "reward": 0.171875, "reward_std": 0.528690755367279, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 782, "step_time": 58.49189604818821 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 668.390625, "completions/mean_terminated_length": 668.390625, "completions/min_length": 285.0, "completions/min_terminated_length": 285.0, "entropy": 0.9769992753863335, "epoch": 0.44387755102040816, "frac_reward_zero_std": 0.875, "grad_norm": 0.359375, "learning_rate": 5.566893424036281e-07, "loss": 0.0043, "num_tokens": 46045031.0, "reward": 0.0546875, "reward_std": 0.15728822350502014, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 783, "step_time": 38.11941893771291 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1154.0, "completions/max_terminated_length": 1154.0, "completions/mean_length": 618.921875, "completions/mean_terminated_length": 618.921875, "completions/min_length": 249.0, "completions/min_terminated_length": 249.0, "entropy": 0.8294611759483814, "epoch": 0.4444444444444444, "frac_reward_zero_std": 0.5, "grad_norm": 0.6484375, "learning_rate": 5.561224489795918e-07, "loss": 0.0218, "num_tokens": 46095178.0, "reward": 0.328125, "reward_std": 0.6970388293266296, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 784, "step_time": 39.68534651771188 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 697.859375, "completions/mean_terminated_length": 697.859375, "completions/min_length": 247.0, "completions/min_terminated_length": 247.0, "entropy": 0.813194153830409, "epoch": 0.44501133786848074, "frac_reward_zero_std": 0.5, "grad_norm": 0.5703125, "learning_rate": 5.555555555555555e-07, "loss": -0.0149, "num_tokens": 46149177.0, "reward": 0.3125, "reward_std": 0.6636837720870972, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 785, "step_time": 45.37823559436947 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 724.953125, "completions/mean_terminated_length": 724.953125, "completions/min_length": 287.0, "completions/min_terminated_length": 287.0, "entropy": 0.9080928973853588, "epoch": 0.445578231292517, "frac_reward_zero_std": 0.625, "grad_norm": 0.515625, "learning_rate": 5.549886621315193e-07, "loss": 0.01, "num_tokens": 46209878.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 786, "step_time": 44.43291334249079 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1106.0, "completions/max_terminated_length": 1106.0, "completions/mean_length": 631.15625, "completions/mean_terminated_length": 631.15625, "completions/min_length": 11.0, "completions/min_terminated_length": 11.0, "entropy": 0.857480525970459, "epoch": 0.44614512471655327, "frac_reward_zero_std": 0.375, "grad_norm": 0.6015625, "learning_rate": 5.544217687074829e-07, "loss": 0.0463, "num_tokens": 46260248.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 787, "step_time": 37.46732331532985 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2864.0, "completions/max_terminated_length": 2864.0, "completions/mean_length": 785.546875, "completions/mean_terminated_length": 785.546875, "completions/min_length": 262.0, "completions/min_terminated_length": 262.0, "entropy": 0.6720971409231424, "epoch": 0.4467120181405896, "frac_reward_zero_std": 0.5, "grad_norm": 0.44140625, "learning_rate": 5.538548752834467e-07, "loss": -0.0103, "num_tokens": 46320851.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 788, "step_time": 106.06966610439122 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1913.0, "completions/max_terminated_length": 1913.0, "completions/mean_length": 817.0, "completions/mean_terminated_length": 817.0, "completions/min_length": 313.0, "completions/min_terminated_length": 313.0, "entropy": 0.6204305812716484, "epoch": 0.44727891156462585, "frac_reward_zero_std": 0.375, "grad_norm": 0.484375, "learning_rate": 5.532879818594104e-07, "loss": -0.0054, "num_tokens": 46382315.0, "reward": 0.3359375, "reward_std": 0.7185450792312622, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 789, "step_time": 65.26288800500333 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1375.0, "completions/max_terminated_length": 1375.0, "completions/mean_length": 758.734375, "completions/mean_terminated_length": 758.734375, "completions/min_length": 277.0, "completions/min_terminated_length": 277.0, "entropy": 0.7401899695396423, "epoch": 0.4478458049886621, "frac_reward_zero_std": 0.375, "grad_norm": 0.57421875, "learning_rate": 5.527210884353742e-07, "loss": -0.0334, "num_tokens": 46443322.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 790, "step_time": 48.78369199298322 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 714.921875, "completions/mean_terminated_length": 714.921875, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "entropy": 0.7558332365006208, "epoch": 0.44841269841269843, "frac_reward_zero_std": 0.75, "grad_norm": 0.326171875, "learning_rate": 5.521541950113379e-07, "loss": 0.004, "num_tokens": 46499925.0, "reward": 0.25, "reward_std": 0.6900655627250671, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 791, "step_time": 47.22886026557535 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1854.0, "completions/max_terminated_length": 1854.0, "completions/mean_length": 864.578125, "completions/mean_terminated_length": 864.578125, "completions/min_length": 390.0, "completions/min_terminated_length": 390.0, "entropy": 0.6858900133520365, "epoch": 0.4489795918367347, "frac_reward_zero_std": 0.5, "grad_norm": 0.421875, "learning_rate": 5.515873015873016e-07, "loss": 0.0086, "num_tokens": 46565514.0, "reward": 0.5234375, "reward_std": 0.8927827477455139, "rewards/accuracy_reward_func/mean": 0.46875, "rewards/accuracy_reward_func/std": 0.8539125919342041, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 792, "step_time": 64.46240985859185 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1498.0, "completions/max_terminated_length": 1498.0, "completions/mean_length": 851.265625, "completions/mean_terminated_length": 851.265625, "completions/min_length": 535.0, "completions/min_terminated_length": 535.0, "entropy": 0.7471986319869757, "epoch": 0.44954648526077096, "frac_reward_zero_std": 0.625, "grad_norm": 0.390625, "learning_rate": 5.510204081632652e-07, "loss": -0.0267, "num_tokens": 46630859.0, "reward": 0.2109375, "reward_std": 0.576006293296814, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 793, "step_time": 51.35069953557104 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 807.609375, "completions/mean_terminated_length": 807.609375, "completions/min_length": 27.0, "completions/min_terminated_length": 27.0, "entropy": 0.7217428702861071, "epoch": 0.4501133786848073, "frac_reward_zero_std": 0.875, "grad_norm": 0.1494140625, "learning_rate": 5.504535147392289e-07, "loss": -0.0005, "num_tokens": 46697194.0, "reward": 0.125, "reward_std": 0.4542567729949951, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 794, "step_time": 60.17362304404378 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1281.0, "completions/max_terminated_length": 1281.0, "completions/mean_length": 727.15625, "completions/mean_terminated_length": 727.15625, "completions/min_length": 433.0, "completions/min_terminated_length": 433.0, "entropy": 0.9020949602127075, "epoch": 0.45068027210884354, "frac_reward_zero_std": 0.625, "grad_norm": 0.484375, "learning_rate": 5.498866213151927e-07, "loss": -0.0134, "num_tokens": 46754660.0, "reward": 0.34375, "reward_std": 0.7175520658493042, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 795, "step_time": 43.84019714873284 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1260.0, "completions/max_terminated_length": 1260.0, "completions/mean_length": 795.78125, "completions/mean_terminated_length": 795.78125, "completions/min_length": 329.0, "completions/min_terminated_length": 329.0, "entropy": 0.8808070737868547, "epoch": 0.4512471655328798, "frac_reward_zero_std": 0.625, "grad_norm": 0.435546875, "learning_rate": 5.493197278911564e-07, "loss": -0.0046, "num_tokens": 46823054.0, "reward": 0.1796875, "reward_std": 0.5589615106582642, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 796, "step_time": 46.92815616633743 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 835.515625, "completions/mean_terminated_length": 835.515625, "completions/min_length": 476.0, "completions/min_terminated_length": 476.0, "entropy": 0.8159170430153608, "epoch": 0.4518140589569161, "frac_reward_zero_std": 0.625, "grad_norm": 0.40625, "learning_rate": 5.487528344671202e-07, "loss": -0.0259, "num_tokens": 46886367.0, "reward": 0.3125, "reward_std": 0.7319250702857971, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 797, "step_time": 49.25248962920159 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1486.0, "completions/max_terminated_length": 1486.0, "completions/mean_length": 704.609375, "completions/mean_terminated_length": 704.609375, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "entropy": 0.7818028852343559, "epoch": 0.4523809523809524, "frac_reward_zero_std": 0.5, "grad_norm": 0.55078125, "learning_rate": 5.481859410430839e-07, "loss": 0.026, "num_tokens": 46941838.0, "reward": 0.1796875, "reward_std": 0.49894481897354126, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 798, "step_time": 50.434957961551845 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1493.0, "completions/max_terminated_length": 1493.0, "completions/mean_length": 755.640625, "completions/mean_terminated_length": 755.640625, "completions/min_length": 292.0, "completions/min_terminated_length": 292.0, "entropy": 0.6671258062124252, "epoch": 0.45294784580498865, "frac_reward_zero_std": 0.375, "grad_norm": 0.5546875, "learning_rate": 5.476190476190477e-07, "loss": 0.0306, "num_tokens": 47000215.0, "reward": 0.3359375, "reward_std": 0.7403060793876648, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 799, "step_time": 51.44981790706515 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1775.0, "completions/max_terminated_length": 1775.0, "completions/mean_length": 724.5, "completions/mean_terminated_length": 724.5, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "entropy": 0.8277410194277763, "epoch": 0.45351473922902497, "frac_reward_zero_std": 0.75, "grad_norm": 0.361328125, "learning_rate": 5.470521541950112e-07, "loss": -0.0069, "num_tokens": 47056583.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 800, "step_time": 61.59256974142045 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1541.0, "completions/max_terminated_length": 1541.0, "completions/mean_length": 760.0, "completions/mean_terminated_length": 760.0, "completions/min_length": 378.0, "completions/min_terminated_length": 378.0, "entropy": 0.8334639333188534, "epoch": 0.45408163265306123, "frac_reward_zero_std": 0.625, "grad_norm": 0.50390625, "learning_rate": 5.46485260770975e-07, "loss": 0.0134, "num_tokens": 47117263.0, "reward": 0.296875, "reward_std": 0.6884463429450989, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 801, "step_time": 53.386829040013254 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1361.0, "completions/max_terminated_length": 1361.0, "completions/mean_length": 769.96875, "completions/mean_terminated_length": 769.96875, "completions/min_length": 395.0, "completions/min_terminated_length": 395.0, "entropy": 0.7062572371214628, "epoch": 0.4546485260770975, "frac_reward_zero_std": 0.5, "grad_norm": 0.447265625, "learning_rate": 5.459183673469387e-07, "loss": -0.0063, "num_tokens": 47176485.0, "reward": 0.3203125, "reward_std": 0.7202688455581665, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 802, "step_time": 46.564074200578034 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1534.0, "completions/max_terminated_length": 1534.0, "completions/mean_length": 746.078125, "completions/mean_terminated_length": 746.078125, "completions/min_length": 300.0, "completions/min_terminated_length": 300.0, "entropy": 0.7608544230461121, "epoch": 0.4552154195011338, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 5.453514739229025e-07, "loss": 0.0283, "num_tokens": 47236530.0, "reward": 0.2421875, "reward_std": 0.6171631813049316, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 803, "step_time": 53.19833767414093 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1015.0, "completions/max_terminated_length": 1015.0, "completions/mean_length": 679.671875, "completions/mean_terminated_length": 679.671875, "completions/min_length": 417.0, "completions/min_terminated_length": 417.0, "entropy": 0.8538750410079956, "epoch": 0.4557823129251701, "frac_reward_zero_std": 0.5, "grad_norm": 0.57421875, "learning_rate": 5.447845804988662e-07, "loss": 0.0345, "num_tokens": 47288997.0, "reward": 0.3046875, "reward_std": 0.664384126663208, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.1171875, "rewards/format_reward_func/std": 0.21347814798355103, "step": 804, "step_time": 34.811470283195376 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 733.859375, "completions/mean_terminated_length": 733.859375, "completions/min_length": 256.0, "completions/min_terminated_length": 256.0, "entropy": 0.8062970414757729, "epoch": 0.45634920634920634, "frac_reward_zero_std": 0.625, "grad_norm": 0.404296875, "learning_rate": 5.4421768707483e-07, "loss": 0.0118, "num_tokens": 47346980.0, "reward": 0.2109375, "reward_std": 0.6029338836669922, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 805, "step_time": 37.23000880237669 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1310.0, "completions/max_terminated_length": 1310.0, "completions/mean_length": 742.328125, "completions/mean_terminated_length": 742.328125, "completions/min_length": 4.0, "completions/min_terminated_length": 4.0, "entropy": 0.7718549035489559, "epoch": 0.45691609977324266, "frac_reward_zero_std": 0.5, "grad_norm": 0.498046875, "learning_rate": 5.436507936507936e-07, "loss": -0.0047, "num_tokens": 47406801.0, "reward": 0.2421875, "reward_std": 0.6423677802085876, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 806, "step_time": 45.8796949069947 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1474.0, "completions/mean_length": 770.40625, "completions/mean_terminated_length": 717.6190795898438, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "entropy": 0.9605542048811913, "epoch": 0.4574829931972789, "frac_reward_zero_std": 0.625, "grad_norm": 0.396484375, "learning_rate": 5.430839002267574e-07, "loss": 0.0068, "num_tokens": 47466827.0, "reward": 0.15625, "reward_std": 0.4443204402923584, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 807, "step_time": 167.76192534156144 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1547.0, "completions/max_terminated_length": 1547.0, "completions/mean_length": 810.625, "completions/mean_terminated_length": 810.625, "completions/min_length": 321.0, "completions/min_terminated_length": 321.0, "entropy": 0.6971529796719551, "epoch": 0.4580498866213152, "frac_reward_zero_std": 0.625, "grad_norm": 0.421875, "learning_rate": 5.42517006802721e-07, "loss": 0.0369, "num_tokens": 47529947.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 808, "step_time": 52.849554646760225 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1357.0, "completions/max_terminated_length": 1357.0, "completions/mean_length": 643.953125, "completions/mean_terminated_length": 643.953125, "completions/min_length": 203.0, "completions/min_terminated_length": 203.0, "entropy": 0.7967690993100405, "epoch": 0.45861678004535145, "frac_reward_zero_std": 0.5, "grad_norm": 0.578125, "learning_rate": 5.419501133786848e-07, "loss": 0.0332, "num_tokens": 47583520.0, "reward": 0.3359375, "reward_std": 0.6961041688919067, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 809, "step_time": 46.43441923055798 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1437.0, "completions/max_terminated_length": 1437.0, "completions/mean_length": 768.640625, "completions/mean_terminated_length": 768.640625, "completions/min_length": 384.0, "completions/min_terminated_length": 384.0, "entropy": 0.7078898008912802, "epoch": 0.45918367346938777, "frac_reward_zero_std": 0.5, "grad_norm": 0.478515625, "learning_rate": 5.413832199546485e-07, "loss": -0.0138, "num_tokens": 47642713.0, "reward": 0.3046875, "reward_std": 0.710561990737915, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 810, "step_time": 49.29645659215748 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1661.0, "completions/max_terminated_length": 1661.0, "completions/mean_length": 821.890625, "completions/mean_terminated_length": 821.890625, "completions/min_length": 260.0, "completions/min_terminated_length": 260.0, "entropy": 0.5739525891840458, "epoch": 0.45975056689342403, "frac_reward_zero_std": 0.625, "grad_norm": 0.337890625, "learning_rate": 5.408163265306123e-07, "loss": 0.0403, "num_tokens": 47705402.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 811, "step_time": 56.428977597504854 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1159.0, "completions/max_terminated_length": 1159.0, "completions/mean_length": 737.75, "completions/mean_terminated_length": 737.75, "completions/min_length": 337.0, "completions/min_terminated_length": 337.0, "entropy": 0.8854955472052097, "epoch": 0.4603174603174603, "frac_reward_zero_std": 0.75, "grad_norm": 0.3203125, "learning_rate": 5.402494331065759e-07, "loss": 0.0161, "num_tokens": 47764842.0, "reward": 0.1171875, "reward_std": 0.41539984941482544, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 812, "step_time": 40.85682630725205 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1157.0, "completions/max_terminated_length": 1157.0, "completions/mean_length": 693.078125, "completions/mean_terminated_length": 693.078125, "completions/min_length": 51.0, "completions/min_terminated_length": 51.0, "entropy": 0.7058844473212957, "epoch": 0.4608843537414966, "frac_reward_zero_std": 0.375, "grad_norm": 0.58203125, "learning_rate": 5.396825396825396e-07, "loss": -0.0421, "num_tokens": 47819767.0, "reward": 0.2421875, "reward_std": 0.6171631813049316, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 813, "step_time": 39.65966281667352 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 839.90625, "completions/mean_terminated_length": 839.90625, "completions/min_length": 367.0, "completions/min_terminated_length": 367.0, "entropy": 0.8400525860488415, "epoch": 0.4614512471655329, "frac_reward_zero_std": 0.875, "grad_norm": 0.259765625, "learning_rate": 5.391156462585034e-07, "loss": 0.0086, "num_tokens": 47884801.0, "reward": 0.09375, "reward_std": 0.29378482699394226, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 814, "step_time": 57.51062997803092 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1467.0, "completions/max_terminated_length": 1467.0, "completions/mean_length": 711.03125, "completions/mean_terminated_length": 711.03125, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "entropy": 0.6923054903745651, "epoch": 0.46201814058956914, "frac_reward_zero_std": 0.75, "grad_norm": 0.294921875, "learning_rate": 5.385487528344671e-07, "loss": -0.0013, "num_tokens": 47939843.0, "reward": 0.375, "reward_std": 0.8261595964431763, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 815, "step_time": 50.49040416907519 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1228.0, "completions/max_terminated_length": 1228.0, "completions/mean_length": 705.359375, "completions/mean_terminated_length": 705.359375, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "entropy": 0.8476248569786549, "epoch": 0.46258503401360546, "frac_reward_zero_std": 0.625, "grad_norm": 0.46875, "learning_rate": 5.379818594104308e-07, "loss": -0.0071, "num_tokens": 47998610.0, "reward": 0.203125, "reward_std": 0.5471335649490356, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 816, "step_time": 44.53605140373111 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 845.4375, "completions/mean_terminated_length": 845.4375, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "entropy": 0.6319690588861704, "epoch": 0.4631519274376417, "frac_reward_zero_std": 0.625, "grad_norm": 0.396484375, "learning_rate": 5.374149659863945e-07, "loss": 0.0163, "num_tokens": 48063238.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 817, "step_time": 50.90306873712689 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 721.46875, "completions/mean_terminated_length": 721.46875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "entropy": 0.7134434916079044, "epoch": 0.463718820861678, "frac_reward_zero_std": 0.5, "grad_norm": 0.52734375, "learning_rate": 5.368480725623582e-07, "loss": -0.0008, "num_tokens": 48120948.0, "reward": 0.2421875, "reward_std": 0.6171631813049316, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 818, "step_time": 43.22208777628839 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1312.0, "completions/max_terminated_length": 1312.0, "completions/mean_length": 751.03125, "completions/mean_terminated_length": 751.03125, "completions/min_length": 308.0, "completions/min_terminated_length": 308.0, "entropy": 0.7312752678990364, "epoch": 0.4642857142857143, "frac_reward_zero_std": 0.875, "grad_norm": 0.271484375, "learning_rate": 5.362811791383219e-07, "loss": -0.003, "num_tokens": 48179742.0, "reward": 0.2109375, "reward_std": 0.653468668460846, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 819, "step_time": 45.742073049768806 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1257.0, "completions/max_terminated_length": 1257.0, "completions/mean_length": 728.859375, "completions/mean_terminated_length": 728.859375, "completions/min_length": 323.0, "completions/min_terminated_length": 323.0, "entropy": 0.8979786559939384, "epoch": 0.46485260770975056, "frac_reward_zero_std": 0.5, "grad_norm": 0.53515625, "learning_rate": 5.357142857142857e-07, "loss": 0.0156, "num_tokens": 48236013.0, "reward": 0.2734375, "reward_std": 0.7233612537384033, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 820, "step_time": 43.38824546709657 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 799.515625, "completions/mean_terminated_length": 799.515625, "completions/min_length": 307.0, "completions/min_terminated_length": 307.0, "entropy": 0.7565864995121956, "epoch": 0.46541950113378683, "frac_reward_zero_std": 0.625, "grad_norm": 0.435546875, "learning_rate": 5.351473922902494e-07, "loss": 0.018, "num_tokens": 48297742.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 821, "step_time": 46.17744138929993 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2680.0, "completions/max_terminated_length": 2680.0, "completions/mean_length": 796.25, "completions/mean_terminated_length": 796.25, "completions/min_length": 362.0, "completions/min_terminated_length": 362.0, "entropy": 0.7098961528390646, "epoch": 0.46598639455782315, "frac_reward_zero_std": 0.625, "grad_norm": 0.375, "learning_rate": 5.345804988662132e-07, "loss": -0.0113, "num_tokens": 48357630.0, "reward": 0.3125, "reward_std": 0.7099072337150574, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 822, "step_time": 95.88210507016629 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 812.828125, "completions/mean_terminated_length": 812.828125, "completions/min_length": 228.0, "completions/min_terminated_length": 228.0, "entropy": 0.6991224847733974, "epoch": 0.4665532879818594, "frac_reward_zero_std": 0.75, "grad_norm": 0.322265625, "learning_rate": 5.340136054421769e-07, "loss": 0.0055, "num_tokens": 48418683.0, "reward": 0.125, "reward_std": 0.417855441570282, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 823, "step_time": 47.68941381480545 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 721.296875, "completions/mean_terminated_length": 721.296875, "completions/min_length": 329.0, "completions/min_terminated_length": 329.0, "entropy": 0.8589720036834478, "epoch": 0.4671201814058957, "frac_reward_zero_std": 0.5, "grad_norm": 0.5, "learning_rate": 5.334467120181405e-07, "loss": 0.0059, "num_tokens": 48474638.0, "reward": 0.34375, "reward_std": 0.6950790882110596, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 824, "step_time": 57.212311525829136 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1902.0, "completions/max_terminated_length": 1902.0, "completions/mean_length": 761.53125, "completions/mean_terminated_length": 761.53125, "completions/min_length": 240.0, "completions/min_terminated_length": 240.0, "entropy": 0.9465934336185455, "epoch": 0.467687074829932, "frac_reward_zero_std": 0.5, "grad_norm": 0.5234375, "learning_rate": 5.328798185941042e-07, "loss": 0.0177, "num_tokens": 48534288.0, "reward": 0.46875, "reward_std": 0.8351171612739563, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 825, "step_time": 65.78962198086083 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 3657.0, "completions/max_terminated_length": 3657.0, "completions/mean_length": 713.09375, "completions/mean_terminated_length": 713.09375, "completions/min_length": 201.0, "completions/min_terminated_length": 201.0, "entropy": 0.7143605556339025, "epoch": 0.46825396825396826, "frac_reward_zero_std": 0.375, "grad_norm": 0.60546875, "learning_rate": 5.32312925170068e-07, "loss": 0.0157, "num_tokens": 48589630.0, "reward": 0.375, "reward_std": 0.7237468957901001, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 826, "step_time": 143.11971341632307 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 657.390625, "completions/mean_terminated_length": 657.390625, "completions/min_length": 141.0, "completions/min_terminated_length": 141.0, "entropy": 0.601087711751461, "epoch": 0.4688208616780045, "frac_reward_zero_std": 0.25, "grad_norm": 0.64453125, "learning_rate": 5.317460317460317e-07, "loss": 0.0268, "num_tokens": 48642895.0, "reward": 0.34375, "reward_std": 0.7393422722816467, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 827, "step_time": 44.84990187827498 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1361.0, "completions/max_terminated_length": 1361.0, "completions/mean_length": 743.296875, "completions/mean_terminated_length": 743.296875, "completions/min_length": 283.0, "completions/min_terminated_length": 283.0, "entropy": 0.7138791903853416, "epoch": 0.46938775510204084, "frac_reward_zero_std": 0.5, "grad_norm": 0.515625, "learning_rate": 5.311791383219955e-07, "loss": -0.0084, "num_tokens": 48700538.0, "reward": 0.265625, "reward_std": 0.6297439336776733, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 828, "step_time": 46.322299542836845 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1342.0, "completions/max_terminated_length": 1342.0, "completions/mean_length": 810.390625, "completions/mean_terminated_length": 810.390625, "completions/min_length": 328.0, "completions/min_terminated_length": 328.0, "entropy": 0.7545554023236036, "epoch": 0.4699546485260771, "frac_reward_zero_std": 0.875, "grad_norm": 0.2119140625, "learning_rate": 5.306122448979592e-07, "loss": 0.0128, "num_tokens": 48765875.0, "reward": 0.125, "reward_std": 0.37796446681022644, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 829, "step_time": 46.48511579725891 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 960.0, "completions/max_terminated_length": 960.0, "completions/mean_length": 623.640625, "completions/mean_terminated_length": 623.640625, "completions/min_length": 334.0, "completions/min_terminated_length": 334.0, "entropy": 0.9303254708647728, "epoch": 0.47052154195011336, "frac_reward_zero_std": 0.625, "grad_norm": 0.51171875, "learning_rate": 5.30045351473923e-07, "loss": -0.0114, "num_tokens": 48819396.0, "reward": 0.328125, "reward_std": 0.7411849498748779, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 830, "step_time": 35.616590317338705 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1225.0, "completions/max_terminated_length": 1225.0, "completions/mean_length": 755.953125, "completions/mean_terminated_length": 755.953125, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "entropy": 0.7408284954726696, "epoch": 0.4710884353741497, "frac_reward_zero_std": 0.625, "grad_norm": 0.384765625, "learning_rate": 5.294784580498866e-07, "loss": -0.018, "num_tokens": 48877153.0, "reward": 0.34375, "reward_std": 0.7175520658493042, "rewards/accuracy_reward_func/mean": 0.28125, "rewards/accuracy_reward_func/std": 0.7007648944854736, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 831, "step_time": 41.72119013313204 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2270.0, "completions/max_terminated_length": 2270.0, "completions/mean_length": 674.15625, "completions/mean_terminated_length": 674.15625, "completions/min_length": 223.0, "completions/min_terminated_length": 223.0, "entropy": 0.8973985128104687, "epoch": 0.47165532879818595, "frac_reward_zero_std": 0.5, "grad_norm": 0.57421875, "learning_rate": 5.289115646258502e-07, "loss": 0.0078, "num_tokens": 48929299.0, "reward": 0.3125, "reward_std": 0.7099072337150574, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 832, "step_time": 78.77375955600291 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2071.0, "completions/max_terminated_length": 2071.0, "completions/mean_length": 877.78125, "completions/mean_terminated_length": 877.78125, "completions/min_length": 229.0, "completions/min_terminated_length": 229.0, "entropy": 0.6091449167579412, "epoch": 0.4722222222222222, "frac_reward_zero_std": 0.75, "grad_norm": 0.29296875, "learning_rate": 5.28344671201814e-07, "loss": -0.0131, "num_tokens": 48994805.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 833, "step_time": 71.98725274205208 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1381.0, "completions/max_terminated_length": 1381.0, "completions/mean_length": 705.6875, "completions/mean_terminated_length": 705.6875, "completions/min_length": 270.0, "completions/min_terminated_length": 270.0, "entropy": 0.8348685037344694, "epoch": 0.47278911564625853, "frac_reward_zero_std": 0.5, "grad_norm": 0.6015625, "learning_rate": 5.277777777777777e-07, "loss": -0.0051, "num_tokens": 49050097.0, "reward": 0.2734375, "reward_std": 0.6294977068901062, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 834, "step_time": 48.15299692656845 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 761.5, "completions/mean_terminated_length": 761.5, "completions/min_length": 214.0, "completions/min_terminated_length": 214.0, "entropy": 0.8017330877482891, "epoch": 0.4733560090702948, "frac_reward_zero_std": 0.625, "grad_norm": 0.455078125, "learning_rate": 5.272108843537415e-07, "loss": -0.0032, "num_tokens": 49109337.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 835, "step_time": 45.73853251710534 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1237.0, "completions/max_terminated_length": 1237.0, "completions/mean_length": 691.53125, "completions/mean_terminated_length": 691.53125, "completions/min_length": 363.0, "completions/min_terminated_length": 363.0, "entropy": 0.7132383771240711, "epoch": 0.47392290249433106, "frac_reward_zero_std": 0.75, "grad_norm": 0.439453125, "learning_rate": 5.266439909297052e-07, "loss": 0.0144, "num_tokens": 49163435.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 836, "step_time": 42.433180400170386 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 780.0625, "completions/mean_terminated_length": 780.0625, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "entropy": 0.8584453668445349, "epoch": 0.4744897959183674, "frac_reward_zero_std": 0.375, "grad_norm": 0.5625, "learning_rate": 5.260770975056689e-07, "loss": 0.0071, "num_tokens": 49223639.0, "reward": 0.2421875, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 837, "step_time": 56.13675133977085 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1243.0, "completions/max_terminated_length": 1243.0, "completions/mean_length": 686.5625, "completions/mean_terminated_length": 686.5625, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "entropy": 0.7876420617103577, "epoch": 0.47505668934240364, "frac_reward_zero_std": 0.375, "grad_norm": 0.6328125, "learning_rate": 5.255102040816326e-07, "loss": 0.0108, "num_tokens": 49276699.0, "reward": 0.53125, "reward_std": 0.8351171612739563, "rewards/accuracy_reward_func/mean": 0.46875, "rewards/accuracy_reward_func/std": 0.8539125919342041, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 838, "step_time": 42.725214873440564 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1267.0, "completions/max_terminated_length": 1267.0, "completions/mean_length": 661.171875, "completions/mean_terminated_length": 661.171875, "completions/min_length": 255.0, "completions/min_terminated_length": 255.0, "entropy": 0.8045499790459871, "epoch": 0.4756235827664399, "frac_reward_zero_std": 0.625, "grad_norm": 0.51171875, "learning_rate": 5.249433106575964e-07, "loss": 0.0008, "num_tokens": 49330302.0, "reward": 0.3046875, "reward_std": 0.710561990737915, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 839, "step_time": 44.34828465804458 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 679.234375, "completions/mean_terminated_length": 679.234375, "completions/min_length": 357.0, "completions/min_terminated_length": 357.0, "entropy": 0.954418022185564, "epoch": 0.47619047619047616, "frac_reward_zero_std": 0.625, "grad_norm": 0.53125, "learning_rate": 5.2437641723356e-07, "loss": 0.007, "num_tokens": 49386573.0, "reward": 0.2265625, "reward_std": 0.5904644727706909, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 840, "step_time": 42.85355908796191 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 733.1875, "completions/mean_terminated_length": 733.1875, "completions/min_length": 278.0, "completions/min_terminated_length": 278.0, "entropy": 0.8145475946366787, "epoch": 0.4767573696145125, "frac_reward_zero_std": 0.625, "grad_norm": 0.494140625, "learning_rate": 5.238095238095238e-07, "loss": 0.0337, "num_tokens": 49444049.0, "reward": 0.2421875, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0234375, "rewards/format_reward_func/std": 0.10652101784944534, "step": 841, "step_time": 53.10136365517974 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2173.0, "completions/max_terminated_length": 2173.0, "completions/mean_length": 740.796875, "completions/mean_terminated_length": 740.796875, "completions/min_length": 123.0, "completions/min_terminated_length": 123.0, "entropy": 0.8114655651152134, "epoch": 0.47732426303854875, "frac_reward_zero_std": 0.75, "grad_norm": 0.34765625, "learning_rate": 5.232426303854875e-07, "loss": 0.0271, "num_tokens": 49502836.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 842, "step_time": 77.42696913983673 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1642.0, "completions/max_terminated_length": 1642.0, "completions/mean_length": 754.03125, "completions/mean_terminated_length": 754.03125, "completions/min_length": 199.0, "completions/min_terminated_length": 199.0, "entropy": 0.7055961042642593, "epoch": 0.477891156462585, "frac_reward_zero_std": 0.5, "grad_norm": 0.42578125, "learning_rate": 5.226757369614512e-07, "loss": 0.017, "num_tokens": 49560110.0, "reward": 0.390625, "reward_std": 0.7739239931106567, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 843, "step_time": 55.73219994083047 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1322.0, "completions/max_terminated_length": 1322.0, "completions/mean_length": 715.125, "completions/mean_terminated_length": 715.125, "completions/min_length": 252.0, "completions/min_terminated_length": 252.0, "entropy": 0.8245087638497353, "epoch": 0.47845804988662133, "frac_reward_zero_std": 0.75, "grad_norm": 0.36328125, "learning_rate": 5.221088435374149e-07, "loss": 0.0008, "num_tokens": 49615438.0, "reward": 0.21875, "reward_std": 0.6538955569267273, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 844, "step_time": 44.68207171745598 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1597.0, "completions/max_terminated_length": 1597.0, "completions/mean_length": 798.625, "completions/mean_terminated_length": 798.625, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "entropy": 0.7979188617318869, "epoch": 0.4790249433106576, "frac_reward_zero_std": 0.75, "grad_norm": 0.333984375, "learning_rate": 5.215419501133787e-07, "loss": -0.019, "num_tokens": 49678918.0, "reward": 0.1640625, "reward_std": 0.496453195810318, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 845, "step_time": 55.64507510140538 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1729.0, "completions/max_terminated_length": 1729.0, "completions/mean_length": 701.15625, "completions/mean_terminated_length": 701.15625, "completions/min_length": 142.0, "completions/min_terminated_length": 142.0, "entropy": 0.8943630382418633, "epoch": 0.47959183673469385, "frac_reward_zero_std": 0.5, "grad_norm": 0.5546875, "learning_rate": 5.209750566893424e-07, "loss": 0.0344, "num_tokens": 49733448.0, "reward": 0.140625, "reward_std": 0.4406772553920746, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 846, "step_time": 59.579053840599954 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2160.0, "completions/max_terminated_length": 2160.0, "completions/mean_length": 817.359375, "completions/mean_terminated_length": 817.359375, "completions/min_length": 273.0, "completions/min_terminated_length": 273.0, "entropy": 0.5982284694910049, "epoch": 0.4801587301587302, "frac_reward_zero_std": 0.5, "grad_norm": 0.4296875, "learning_rate": 5.204081632653062e-07, "loss": -0.0004, "num_tokens": 49795631.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 847, "step_time": 75.91410423349589 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1697.0, "completions/max_terminated_length": 1697.0, "completions/mean_length": 869.609375, "completions/mean_terminated_length": 869.609375, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "entropy": 0.6736322026699781, "epoch": 0.48072562358276644, "frac_reward_zero_std": 0.875, "grad_norm": 0.2236328125, "learning_rate": 5.198412698412698e-07, "loss": 0.0009, "num_tokens": 49860606.0, "reward": 0.0859375, "reward_std": 0.3398725986480713, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 848, "step_time": 58.301382817327976 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1979.0, "completions/max_terminated_length": 1979.0, "completions/mean_length": 634.765625, "completions/mean_terminated_length": 634.765625, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "entropy": 0.8133383151143789, "epoch": 0.4812925170068027, "frac_reward_zero_std": 0.75, "grad_norm": 0.439453125, "learning_rate": 5.192743764172335e-07, "loss": 0.007, "num_tokens": 49912231.0, "reward": 0.109375, "reward_std": 0.37234508991241455, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 849, "step_time": 69.65134237147868 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1630.0, "completions/max_terminated_length": 1630.0, "completions/mean_length": 652.640625, "completions/mean_terminated_length": 652.640625, "completions/min_length": 171.0, "completions/min_terminated_length": 171.0, "entropy": 0.7148718424141407, "epoch": 0.481859410430839, "frac_reward_zero_std": 0.625, "grad_norm": 0.5, "learning_rate": 5.187074829931972e-07, "loss": 0.0006, "num_tokens": 49964464.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 850, "step_time": 57.249084785580635 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1642.0, "completions/max_terminated_length": 1642.0, "completions/mean_length": 751.140625, "completions/mean_terminated_length": 751.140625, "completions/min_length": 211.0, "completions/min_terminated_length": 211.0, "entropy": 0.7809609100222588, "epoch": 0.4824263038548753, "frac_reward_zero_std": 0.75, "grad_norm": 0.30859375, "learning_rate": 5.18140589569161e-07, "loss": 0.0036, "num_tokens": 50023073.0, "reward": 0.0703125, "reward_std": 0.27984103560447693, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 851, "step_time": 56.52398924808949 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 677.6875, "completions/mean_terminated_length": 677.6875, "completions/min_length": 106.0, "completions/min_terminated_length": 106.0, "entropy": 0.9891784265637398, "epoch": 0.48299319727891155, "frac_reward_zero_std": 0.75, "grad_norm": 0.47265625, "learning_rate": 5.175736961451247e-07, "loss": 0.0033, "num_tokens": 50076749.0, "reward": 0.0859375, "reward_std": 0.28942590951919556, "rewards/accuracy_reward_func/mean": 0.03125, "rewards/accuracy_reward_func/std": 0.25, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 852, "step_time": 42.11584579385817 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 4096.0, "completions/max_terminated_length": 1387.0, "completions/mean_length": 783.21875, "completions/mean_terminated_length": 730.6349487304688, "completions/min_length": 329.0, "completions/min_terminated_length": 329.0, "entropy": 0.8308954164385796, "epoch": 0.48356009070294786, "frac_reward_zero_std": 0.5, "grad_norm": 0.478515625, "learning_rate": 5.170068027210885e-07, "loss": -0.0198, "num_tokens": 50137371.0, "reward": 0.3125, "reward_std": 0.7099072337150574, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 853, "step_time": 169.93684468977153 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1159.0, "completions/max_terminated_length": 1159.0, "completions/mean_length": 713.6875, "completions/mean_terminated_length": 713.6875, "completions/min_length": 336.0, "completions/min_terminated_length": 336.0, "entropy": 0.9584300182759762, "epoch": 0.48412698412698413, "frac_reward_zero_std": 0.875, "grad_norm": 0.33203125, "learning_rate": 5.164399092970522e-07, "loss": 0.0098, "num_tokens": 50194703.0, "reward": 0.046875, "reward_std": 0.14689241349697113, "rewards/accuracy_reward_func/mean": 0.0, "rewards/accuracy_reward_func/std": 0.0, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 854, "step_time": 39.95798596646637 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 855.171875, "completions/mean_terminated_length": 855.171875, "completions/min_length": 421.0, "completions/min_terminated_length": 421.0, "entropy": 0.6795213408768177, "epoch": 0.4846938775510204, "frac_reward_zero_std": 0.625, "grad_norm": 0.419921875, "learning_rate": 5.15873015873016e-07, "loss": 0.0094, "num_tokens": 50259994.0, "reward": 0.1953125, "reward_std": 0.5463964939117432, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 855, "step_time": 52.30651016253978 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1287.0, "completions/max_terminated_length": 1287.0, "completions/mean_length": 647.09375, "completions/mean_terminated_length": 647.09375, "completions/min_length": 317.0, "completions/min_terminated_length": 317.0, "entropy": 0.8219214268028736, "epoch": 0.4852607709750567, "frac_reward_zero_std": 0.0, "grad_norm": 0.87109375, "learning_rate": 5.153061224489795e-07, "loss": -0.0243, "num_tokens": 50311496.0, "reward": 0.53125, "reward_std": 0.8351171612739563, "rewards/accuracy_reward_func/mean": 0.46875, "rewards/accuracy_reward_func/std": 0.8539125919342041, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 856, "step_time": 43.74736983794719 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 630.875, "completions/mean_terminated_length": 630.875, "completions/min_length": 314.0, "completions/min_terminated_length": 314.0, "entropy": 0.8346275314688683, "epoch": 0.485827664399093, "frac_reward_zero_std": 0.375, "grad_norm": 0.64453125, "learning_rate": 5.147392290249432e-07, "loss": 0.0067, "num_tokens": 50363360.0, "reward": 0.4609375, "reward_std": 0.7983350157737732, "rewards/accuracy_reward_func/mean": 0.40625, "rewards/accuracy_reward_func/std": 0.8110105991363525, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 857, "step_time": 52.10660233069211 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 737.171875, "completions/mean_terminated_length": 737.171875, "completions/min_length": 377.0, "completions/min_terminated_length": 377.0, "entropy": 0.8310398850589991, "epoch": 0.48639455782312924, "frac_reward_zero_std": 0.625, "grad_norm": 0.470703125, "learning_rate": 5.14172335600907e-07, "loss": -0.0303, "num_tokens": 50421827.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 858, "step_time": 44.48652577307075 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1220.0, "completions/max_terminated_length": 1220.0, "completions/mean_length": 673.171875, "completions/mean_terminated_length": 673.171875, "completions/min_length": 330.0, "completions/min_terminated_length": 330.0, "entropy": 0.9571273773908615, "epoch": 0.48696145124716556, "frac_reward_zero_std": 0.75, "grad_norm": 0.4453125, "learning_rate": 5.136054421768707e-07, "loss": 0.0135, "num_tokens": 50474806.0, "reward": 0.15625, "reward_std": 0.4443204402923584, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 859, "step_time": 41.45254014246166 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 666.890625, "completions/mean_terminated_length": 666.890625, "completions/min_length": 384.0, "completions/min_terminated_length": 384.0, "entropy": 0.92544767819345, "epoch": 0.4875283446712018, "frac_reward_zero_std": 0.5, "grad_norm": 0.578125, "learning_rate": 5.130385487528345e-07, "loss": 0.0019, "num_tokens": 50526975.0, "reward": 0.3046875, "reward_std": 0.710561990737915, "rewards/accuracy_reward_func/mean": 0.25, "rewards/accuracy_reward_func/std": 0.6666666865348816, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 860, "step_time": 40.417448453605175 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1511.0, "completions/max_terminated_length": 1511.0, "completions/mean_length": 820.453125, "completions/mean_terminated_length": 820.453125, "completions/min_length": 381.0, "completions/min_terminated_length": 381.0, "entropy": 0.7471801303327084, "epoch": 0.4880952380952381, "frac_reward_zero_std": 0.625, "grad_norm": 0.392578125, "learning_rate": 5.124716553287982e-07, "loss": -0.0161, "num_tokens": 50590172.0, "reward": 0.09375, "reward_std": 0.36596253514289856, "rewards/accuracy_reward_func/mean": 0.0625, "rewards/accuracy_reward_func/std": 0.35073620080947876, "rewards/format_reward_func/mean": 0.03125, "rewards/format_reward_func/std": 0.12198751419782639, "step": 861, "step_time": 52.75329654570669 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1692.0, "completions/max_terminated_length": 1692.0, "completions/mean_length": 791.78125, "completions/mean_terminated_length": 791.78125, "completions/min_length": 327.0, "completions/min_terminated_length": 327.0, "entropy": 0.6955524003133178, "epoch": 0.4886621315192744, "frac_reward_zero_std": 0.625, "grad_norm": 0.40234375, "learning_rate": 5.119047619047619e-07, "loss": 0.0071, "num_tokens": 50650750.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 862, "step_time": 58.22629368491471 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1316.0, "completions/max_terminated_length": 1316.0, "completions/mean_length": 611.46875, "completions/mean_terminated_length": 611.46875, "completions/min_length": 173.0, "completions/min_terminated_length": 173.0, "entropy": 0.7979342136532068, "epoch": 0.48922902494331066, "frac_reward_zero_std": 0.5, "grad_norm": 0.58984375, "learning_rate": 5.113378684807256e-07, "loss": -0.0108, "num_tokens": 50699988.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 863, "step_time": 45.322847423143685 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 999.0, "completions/max_terminated_length": 999.0, "completions/mean_length": 644.953125, "completions/mean_terminated_length": 644.953125, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "entropy": 1.2054049521684647, "epoch": 0.4897959183673469, "frac_reward_zero_std": 0.5, "grad_norm": 0.67578125, "learning_rate": 5.107709750566893e-07, "loss": -0.0018, "num_tokens": 50752945.0, "reward": 0.2265625, "reward_std": 0.5904644727706909, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 864, "step_time": 34.44790306407958 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2064.0, "completions/max_terminated_length": 2064.0, "completions/mean_length": 805.4375, "completions/mean_terminated_length": 805.4375, "completions/min_length": 343.0, "completions/min_terminated_length": 343.0, "entropy": 0.729393633082509, "epoch": 0.49036281179138325, "frac_reward_zero_std": 0.625, "grad_norm": 0.4375, "learning_rate": 5.10204081632653e-07, "loss": 0.0146, "num_tokens": 50815317.0, "reward": 0.1328125, "reward_std": 0.4386323094367981, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 865, "step_time": 71.76423513516784 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1230.0, "completions/max_terminated_length": 1230.0, "completions/mean_length": 704.65625, "completions/mean_terminated_length": 704.65625, "completions/min_length": 315.0, "completions/min_terminated_length": 315.0, "entropy": 1.010851725935936, "epoch": 0.4909297052154195, "frac_reward_zero_std": 0.75, "grad_norm": 0.419921875, "learning_rate": 5.096371882086168e-07, "loss": -0.0135, "num_tokens": 50871743.0, "reward": 0.203125, "reward_std": 0.6281666159629822, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 866, "step_time": 41.98374019283801 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 740.65625, "completions/mean_terminated_length": 740.65625, "completions/min_length": 365.0, "completions/min_terminated_length": 365.0, "entropy": 0.8532387372106314, "epoch": 0.4914965986394558, "frac_reward_zero_std": 0.5, "grad_norm": 0.5703125, "learning_rate": 5.090702947845805e-07, "loss": -0.0021, "num_tokens": 50935129.0, "reward": 0.2421875, "reward_std": 0.5908843874931335, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 867, "step_time": 46.843373676761985 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1258.0, "completions/max_terminated_length": 1258.0, "completions/mean_length": 708.3125, "completions/mean_terminated_length": 708.3125, "completions/min_length": 273.0, "completions/min_terminated_length": 273.0, "entropy": 0.6399582121521235, "epoch": 0.49206349206349204, "frac_reward_zero_std": 0.625, "grad_norm": 0.486328125, "learning_rate": 5.085034013605442e-07, "loss": -0.0179, "num_tokens": 50990165.0, "reward": 0.25, "reward_std": 0.5909368395805359, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 868, "step_time": 43.424545377492905 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1910.0, "completions/max_terminated_length": 1910.0, "completions/mean_length": 780.140625, "completions/mean_terminated_length": 780.140625, "completions/min_length": 299.0, "completions/min_terminated_length": 299.0, "entropy": 0.810400640591979, "epoch": 0.49263038548752835, "frac_reward_zero_std": 0.625, "grad_norm": 0.455078125, "learning_rate": 5.079365079365079e-07, "loss": 0.0313, "num_tokens": 51049406.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 869, "step_time": 65.45391212496907 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 731.515625, "completions/mean_terminated_length": 731.515625, "completions/min_length": 367.0, "completions/min_terminated_length": 367.0, "entropy": 0.8992546834051609, "epoch": 0.4931972789115646, "frac_reward_zero_std": 0.875, "grad_norm": 0.267578125, "learning_rate": 5.073696145124717e-07, "loss": 0.0113, "num_tokens": 51108247.0, "reward": 0.15625, "reward_std": 0.4443204402923584, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 870, "step_time": 41.870989022776484 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 744.078125, "completions/mean_terminated_length": 744.078125, "completions/min_length": 380.0, "completions/min_terminated_length": 380.0, "entropy": 0.8174138925969601, "epoch": 0.4937641723356009, "frac_reward_zero_std": 0.625, "grad_norm": 0.4375, "learning_rate": 5.068027210884354e-07, "loss": -0.0256, "num_tokens": 51167020.0, "reward": 0.1484375, "reward_std": 0.44257262349128723, "rewards/accuracy_reward_func/mean": 0.09375, "rewards/accuracy_reward_func/std": 0.42608407139778137, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 871, "step_time": 49.688143651001155 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1795.0, "completions/max_terminated_length": 1795.0, "completions/mean_length": 705.75, "completions/mean_terminated_length": 705.75, "completions/min_length": 232.0, "completions/min_terminated_length": 232.0, "entropy": 0.6099121104925871, "epoch": 0.4943310657596372, "frac_reward_zero_std": 0.5, "grad_norm": 0.462890625, "learning_rate": 5.062358276643991e-07, "loss": -0.0043, "num_tokens": 51221348.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 872, "step_time": 61.60922578629106 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 680.453125, "completions/mean_terminated_length": 680.453125, "completions/min_length": 376.0, "completions/min_terminated_length": 376.0, "entropy": 1.0266446508467197, "epoch": 0.49489795918367346, "frac_reward_zero_std": 0.5, "grad_norm": 0.60546875, "learning_rate": 5.056689342403628e-07, "loss": -0.0002, "num_tokens": 51275617.0, "reward": 0.28125, "reward_std": 0.6291528940200806, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 873, "step_time": 34.654701344668865 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 823.9375, "completions/mean_terminated_length": 823.9375, "completions/min_length": 439.0, "completions/min_terminated_length": 439.0, "entropy": 0.8492737971246243, "epoch": 0.4954648526077097, "frac_reward_zero_std": 0.625, "grad_norm": 0.431640625, "learning_rate": 5.051020408163264e-07, "loss": 0.0176, "num_tokens": 51341557.0, "reward": 0.1875, "reward_std": 0.5, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 874, "step_time": 53.710381127893925 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 645.59375, "completions/mean_terminated_length": 645.59375, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "entropy": 1.10162472166121, "epoch": 0.49603174603174605, "frac_reward_zero_std": 0.5, "grad_norm": 0.76171875, "learning_rate": 5.045351473922902e-07, "loss": 0.0093, "num_tokens": 51401931.0, "reward": 0.2578125, "reward_std": 0.629891574382782, "rewards/accuracy_reward_func/mean": 0.21875, "rewards/accuracy_reward_func/std": 0.6291528940200806, "rewards/format_reward_func/mean": 0.0390625, "rewards/format_reward_func/std": 0.13524486124515533, "step": 875, "step_time": 42.514109442941844 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1422.0, "completions/max_terminated_length": 1422.0, "completions/mean_length": 661.0625, "completions/mean_terminated_length": 661.0625, "completions/min_length": 318.0, "completions/min_terminated_length": 318.0, "entropy": 0.8126551173627377, "epoch": 0.4965986394557823, "frac_reward_zero_std": 0.625, "grad_norm": 0.484375, "learning_rate": 5.039682539682539e-07, "loss": 0.0154, "num_tokens": 51454799.0, "reward": 0.375, "reward_std": 0.7867957949638367, "rewards/accuracy_reward_func/mean": 0.3125, "rewards/accuracy_reward_func/std": 0.7319250702857971, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 876, "step_time": 49.50599535834044 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 736.921875, "completions/mean_terminated_length": 736.921875, "completions/min_length": 261.0, "completions/min_terminated_length": 261.0, "entropy": 0.8450348358601332, "epoch": 0.4971655328798186, "frac_reward_zero_std": 0.75, "grad_norm": 0.35546875, "learning_rate": 5.034013605442177e-07, "loss": 0.0187, "num_tokens": 51514826.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 877, "step_time": 45.9421418486163 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/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": 681.6875, "completions/mean_terminated_length": 681.6875, "completions/min_length": 242.0, "completions/min_terminated_length": 242.0, "entropy": 0.9226914681494236, "epoch": 0.4977324263038549, "frac_reward_zero_std": 0.625, "grad_norm": 0.53125, "learning_rate": 5.028344671201814e-07, "loss": 0.0093, "num_tokens": 51568558.0, "reward": 0.2109375, "reward_std": 0.5477564930915833, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0546875, "rewards/format_reward_func/std": 0.15728822350502014, "step": 878, "step_time": 47.00045523699373 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1278.0, "completions/max_terminated_length": 1278.0, "completions/mean_length": 725.0625, "completions/mean_terminated_length": 725.0625, "completions/min_length": 299.0, "completions/min_terminated_length": 299.0, "entropy": 0.8534124288707972, "epoch": 0.49829931972789115, "frac_reward_zero_std": 0.5, "grad_norm": 0.625, "learning_rate": 5.022675736961452e-07, "loss": -0.0423, "num_tokens": 51626098.0, "reward": 0.21875, "reward_std": 0.5482656359672546, "rewards/accuracy_reward_func/mean": 0.15625, "rewards/accuracy_reward_func/std": 0.5409794449806213, "rewards/format_reward_func/mean": 0.0625, "rewards/format_reward_func/std": 0.1666666716337204, "step": 879, "step_time": 44.45098415762186 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1383.0, "completions/max_terminated_length": 1383.0, "completions/mean_length": 698.984375, "completions/mean_terminated_length": 698.984375, "completions/min_length": 137.0, "completions/min_terminated_length": 137.0, "entropy": 0.6996864043176174, "epoch": 0.4988662131519274, "frac_reward_zero_std": 0.25, "grad_norm": 0.81640625, "learning_rate": 5.017006802721087e-07, "loss": -0.0009, "num_tokens": 51680737.0, "reward": 0.390625, "reward_std": 0.7531349658966064, "rewards/accuracy_reward_func/mean": 0.34375, "rewards/accuracy_reward_func/std": 0.7605084180831909, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 880, "step_time": 47.530423319898546 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1462.0, "completions/max_terminated_length": 1462.0, "completions/mean_length": 766.0, "completions/mean_terminated_length": 766.0, "completions/min_length": 296.0, "completions/min_terminated_length": 296.0, "entropy": 0.8448827806860209, "epoch": 0.49943310657596374, "frac_reward_zero_std": 0.625, "grad_norm": 0.470703125, "learning_rate": 5.011337868480725e-07, "loss": 0.0227, "num_tokens": 51739321.0, "reward": 0.171875, "reward_std": 0.49776285886764526, "rewards/accuracy_reward_func/mean": 0.125, "rewards/accuracy_reward_func/std": 0.48795005679130554, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 881, "step_time": 50.53285096306354 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1428.0, "completions/max_terminated_length": 1428.0, "completions/mean_length": 702.734375, "completions/mean_terminated_length": 702.734375, "completions/min_length": 226.0, "completions/min_terminated_length": 226.0, "entropy": 0.733205184340477, "epoch": 0.5, "frac_reward_zero_std": 0.625, "grad_norm": 0.484375, "learning_rate": 5.005668934240362e-07, "loss": -0.0118, "num_tokens": 51794136.0, "reward": 0.234375, "reward_std": 0.5907269716262817, "rewards/accuracy_reward_func/mean": 0.1875, "rewards/accuracy_reward_func/std": 0.5875696539878845, "rewards/format_reward_func/mean": 0.046875, "rewards/format_reward_func/std": 0.14689241349697113, "step": 882, "step_time": 49.36098977178335 } ], "logging_steps": 1, "max_steps": 1764, "num_input_tokens_seen": 51794136, "num_train_epochs": 1, "save_steps": 500, "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 }