{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.25, "eval_steps": 500, "global_step": 100, "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": 1775.0, "completions/max_terminated_length": 1775.0, "completions/mean_length": 373.203125, "completions/mean_terminated_length": 373.203125, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.0025, "frac_reward_zero_std": 0.0, "grad_norm": 61.19800567626953, "learning_rate": 0.0, "loss": -0.0, "num_tokens": 103973.0, "reward": 0.4570312201976776, "reward_std": 0.420915812253952, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.2302463799715042, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 1 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1893.0, "completions/mean_length": 500.46875, "completions/mean_terminated_length": 475.90478515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.005, "frac_reward_zero_std": 0.125, "grad_norm": 8.731510162353516, "learning_rate": 2e-07, "loss": -0.0, "num_tokens": 253003.0, "reward": 0.20546874403953552, "reward_std": 0.30662786960601807, "rewards/format_reward/mean": 0.3359375, "rewards/format_reward/std": 0.2366211861371994, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 2 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1358.0, "completions/mean_length": 451.625, "completions/mean_terminated_length": 426.2857360839844, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.0075, "frac_reward_zero_std": 0.0, "grad_norm": 7.176198959350586, "learning_rate": 4e-07, "loss": -0.0, "num_tokens": 350011.0, "reward": 0.36406248807907104, "reward_std": 0.38881734013557434, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.3503824472427368, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 3 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1419.0, "completions/max_terminated_length": 1419.0, "completions/mean_length": 549.84375, "completions/mean_terminated_length": 549.84375, "completions/min_length": 7.0, "completions/min_terminated_length": 7.0, "epoch": 0.01, "frac_reward_zero_std": 0.0, "grad_norm": 3.659719467163086, "learning_rate": 6e-07, "loss": 0.0, "num_tokens": 472809.0, "reward": 0.31640625, "reward_std": 0.26256585121154785, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.24688033759593964, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 4 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1599.0, "completions/mean_length": 554.0, "completions/mean_terminated_length": 505.8064270019531, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0125, "frac_reward_zero_std": 0.0, "grad_norm": 5.477923393249512, "learning_rate": 8e-07, "loss": -0.0, "num_tokens": 602145.0, "reward": 0.21796873211860657, "reward_std": 0.28978729248046875, "rewards/format_reward/mean": 0.3046875, "rewards/format_reward/std": 0.2762732207775116, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 5 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1515.0, "completions/max_terminated_length": 1515.0, "completions/mean_length": 516.9375, "completions/mean_terminated_length": 516.9375, "completions/min_length": 82.0, "completions/min_terminated_length": 82.0, "epoch": 0.015, "frac_reward_zero_std": 0.125, "grad_norm": 0.9887422919273376, "learning_rate": 1e-06, "loss": -0.0, "num_tokens": 718333.0, "reward": 0.28125, "reward_std": 0.4292016327381134, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.2745848298072815, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 6 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1518.0, "completions/max_terminated_length": 1518.0, "completions/mean_length": 413.5, "completions/mean_terminated_length": 413.5, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0175, "frac_reward_zero_std": 0.0, "grad_norm": 8.222546577453613, "learning_rate": 9.99726628670463e-07, "loss": -0.0, "num_tokens": 817261.0, "reward": 0.23749999701976776, "reward_std": 0.3429509401321411, "rewards/format_reward/mean": 0.34375, "rewards/format_reward/std": 0.279951810836792, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 7 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1751.0, "completions/mean_length": 601.84375, "completions/mean_terminated_length": 578.888916015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.02, "frac_reward_zero_std": 0.0, "grad_norm": 2.261916399002075, "learning_rate": 9.989068136093872e-07, "loss": 0.0, "num_tokens": 936907.0, "reward": 0.12890625, "reward_std": 0.24654410779476166, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.26246222853660583, "rewards/mcq_exact_match_reward/mean": 0.09375, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "step": 8 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1878.0, "completions/max_terminated_length": 1878.0, "completions/mean_length": 553.078125, "completions/mean_terminated_length": 553.078125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0225, "frac_reward_zero_std": 0.0, "grad_norm": 9.411349296569824, "learning_rate": 9.975414512725056e-07, "loss": -0.0, "num_tokens": 1048784.0, "reward": 0.3070312440395355, "reward_std": 0.37265777587890625, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.30282649397850037, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 9 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1987.0, "completions/max_terminated_length": 1987.0, "completions/mean_length": 617.15625, "completions/mean_terminated_length": 617.15625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.025, "frac_reward_zero_std": 0.0, "grad_norm": 15.806661605834961, "learning_rate": 9.956320346634875e-07, "loss": 0.0, "num_tokens": 1186242.0, "reward": 0.40156251192092896, "reward_std": 0.35580530762672424, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.25539806485176086, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 10 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1087.0, "completions/max_terminated_length": 1087.0, "completions/mean_length": 387.078125, "completions/mean_terminated_length": 387.078125, "completions/min_length": 11.0, "completions/min_terminated_length": 11.0, "epoch": 0.0275, "frac_reward_zero_std": 0.0, "grad_norm": 2.2821664810180664, "learning_rate": 9.931806517013612e-07, "loss": -0.0, "num_tokens": 1269063.0, "reward": 0.24374999105930328, "reward_std": 0.3170398771762848, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.3196600377559662, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 11 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1673.0, "completions/max_terminated_length": 1673.0, "completions/mean_length": 456.359375, "completions/mean_terminated_length": 456.359375, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.03, "frac_reward_zero_std": 0.0, "grad_norm": 25.601293563842773, "learning_rate": 9.901899829374047e-07, "loss": 0.0, "num_tokens": 1379534.0, "reward": 0.14999999105930328, "reward_std": 0.23115794360637665, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.3196600377559662, "rewards/mcq_exact_match_reward/mean": 0.109375, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "step": 12 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1235.0, "completions/mean_length": 554.734375, "completions/mean_terminated_length": 481.2950439453125, "completions/min_length": 93.0, "completions/min_terminated_length": 93.0, "epoch": 0.0325, "frac_reward_zero_std": 0.0, "grad_norm": 1.4019166231155396, "learning_rate": 9.866632986240029e-07, "loss": -0.0, "num_tokens": 1485709.0, "reward": 0.25312498211860657, "reward_std": 0.38949641585350037, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.34503278136253357, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 13 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1669.0, "completions/mean_length": 482.546875, "completions/mean_terminated_length": 457.69842529296875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.035, "frac_reward_zero_std": 0.0, "grad_norm": 9.08604907989502, "learning_rate": 9.826044551386742e-07, "loss": 0.0, "num_tokens": 1593384.0, "reward": 0.24296873807907104, "reward_std": 0.3087196350097656, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.31090864539146423, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 14 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1038.0, "completions/mean_length": 536.59375, "completions/mean_terminated_length": 435.8333435058594, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.0375, "frac_reward_zero_std": 0.0, "grad_norm": 2.18623948097229, "learning_rate": 9.780178907671788e-07, "loss": 0.0, "num_tokens": 1700126.0, "reward": 0.19062499701976776, "reward_std": 0.26142001152038574, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.3333333432674408, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "step": 15 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1552.0, "completions/max_terminated_length": 1552.0, "completions/mean_length": 322.5625, "completions/mean_terminated_length": 322.5625, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.04, "frac_reward_zero_std": 0.125, "grad_norm": 36.68941879272461, "learning_rate": 9.729086208503173e-07, "loss": 0.0, "num_tokens": 1817850.0, "reward": 0.5250000357627869, "reward_std": 0.2990384101867676, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.28867512941360474, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 16 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1580.0, "completions/mean_length": 480.46875, "completions/mean_terminated_length": 455.58734130859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0425, "frac_reward_zero_std": 0.0, "grad_norm": 11.76828384399414, "learning_rate": 9.672822322997304e-07, "loss": 0.0, "num_tokens": 1930832.0, "reward": 0.40390628576278687, "reward_std": 0.3442089259624481, "rewards/format_reward/mean": 0.6015625, "rewards/format_reward/std": 0.3354656398296356, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 17 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 807.0, "completions/mean_length": 272.90625, "completions/mean_terminated_length": 244.7301788330078, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.045, "frac_reward_zero_std": 0.0, "grad_norm": 13.187530517578125, "learning_rate": 9.611448774886923e-07, "loss": -0.0, "num_tokens": 2018682.0, "reward": 0.23749999701976776, "reward_std": 0.27528345584869385, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.26726123690605164, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 18 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1438.0, "completions/mean_length": 379.375, "completions/mean_terminated_length": 325.5483703613281, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0475, "frac_reward_zero_std": 0.25, "grad_norm": 8.468963623046875, "learning_rate": 9.545032675245813e-07, "loss": 0.0, "num_tokens": 2134634.0, "reward": 0.3609375059604645, "reward_std": 0.24825170636177063, "rewards/format_reward/mean": 0.640625, "rewards/format_reward/std": 0.28824523091316223, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 19 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 984.0, "completions/max_terminated_length": 984.0, "completions/mean_length": 173.625, "completions/mean_terminated_length": 173.625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05, "frac_reward_zero_std": 0.375, "grad_norm": 5.215916156768799, "learning_rate": 9.473646649103817e-07, "loss": 0.0, "num_tokens": 2234330.0, "reward": 0.09296874701976776, "reward_std": 0.09501710534095764, "rewards/format_reward/mean": 0.6171875, "rewards/format_reward/std": 0.2478829026222229, "rewards/mcq_exact_match_reward/mean": 0.03125, "rewards/mcq_exact_match_reward/std": 0.17536810040473938, "step": 20 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 689.0, "completions/max_terminated_length": 689.0, "completions/mean_length": 153.75, "completions/mean_terminated_length": 153.75, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0525, "frac_reward_zero_std": 0.125, "grad_norm": 16.54804039001465, "learning_rate": 9.397368756032444e-07, "loss": -0.0, "num_tokens": 2322818.0, "reward": 0.21328124403953552, "reward_std": 0.20918546617031097, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.19654127955436707, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "step": 21 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 629.0, "completions/max_terminated_length": 629.0, "completions/mean_length": 139.796875, "completions/mean_terminated_length": 139.796875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.055, "frac_reward_zero_std": 0.0, "grad_norm": 10.184212684631348, "learning_rate": 9.316282404787869e-07, "loss": 0.0, "num_tokens": 2425645.0, "reward": 0.4273437261581421, "reward_std": 0.3809056282043457, "rewards/format_reward/mean": 0.6796875, "rewards/format_reward/std": 0.27265870571136475, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 22 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 867.0, "completions/max_terminated_length": 867.0, "completions/mean_length": 100.53125, "completions/mean_terminated_length": 100.53125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0575, "frac_reward_zero_std": 0.125, "grad_norm": 15.394311904907227, "learning_rate": 9.230476262104676e-07, "loss": -0.0, "num_tokens": 2531863.0, "reward": 0.5539062023162842, "reward_std": 0.3521394729614258, "rewards/format_reward/mean": 0.6953125, "rewards/format_reward/std": 0.2615155577659607, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 23 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1364.0, "completions/max_terminated_length": 1364.0, "completions/mean_length": 119.421875, "completions/mean_terminated_length": 119.421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06, "frac_reward_zero_std": 0.25, "grad_norm": 12.308192253112793, "learning_rate": 9.1400441557401e-07, "loss": 0.0, "num_tokens": 2639666.0, "reward": 0.28046876192092896, "reward_std": 0.3111737072467804, "rewards/format_reward/mean": 0.6171875, "rewards/format_reward/std": 0.2634054720401764, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 24 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1153.0, "completions/max_terminated_length": 1153.0, "completions/mean_length": 149.46875, "completions/mean_terminated_length": 149.46875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0625, "frac_reward_zero_std": 0.375, "grad_norm": 24.342159271240234, "learning_rate": 9.045084971874737e-07, "loss": 0.0, "num_tokens": 2713240.0, "reward": 0.524218738079071, "reward_std": 0.18854781985282898, "rewards/format_reward/mean": 0.8671875, "rewards/format_reward/std": 0.23974503576755524, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "step": 25 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1686.0, "completions/max_terminated_length": 1686.0, "completions/mean_length": 81.515625, "completions/mean_terminated_length": 81.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.065, "frac_reward_zero_std": 0.375, "grad_norm": 15.462127685546875, "learning_rate": 8.945702546981968e-07, "loss": 0.0, "num_tokens": 2804001.0, "reward": 0.35234373807907104, "reward_std": 0.22906258702278137, "rewards/format_reward/mean": 0.7109375, "rewards/format_reward/std": 0.24888142943382263, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 26 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 522.0, "completions/max_terminated_length": 522.0, "completions/mean_length": 64.90625, "completions/mean_terminated_length": 64.90625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0675, "frac_reward_zero_std": 0.125, "grad_norm": 25.694332122802734, "learning_rate": 8.842005554284295e-07, "loss": 0.0, "num_tokens": 2891123.0, "reward": 0.3046875, "reward_std": 0.3016444146633148, "rewards/format_reward/mean": 0.859375, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 27 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 215.0, "completions/max_terminated_length": 215.0, "completions/mean_length": 25.703125, "completions/mean_terminated_length": 25.703125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07, "frac_reward_zero_std": 0.125, "grad_norm": 19.93283462524414, "learning_rate": 8.734107384920769e-07, "loss": 0.0, "num_tokens": 2974664.0, "reward": 0.38749998807907104, "reward_std": 0.29532599449157715, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.21593283116817474, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 28 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 545.0, "completions/max_terminated_length": 545.0, "completions/mean_length": 48.109375, "completions/mean_terminated_length": 48.109375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0725, "frac_reward_zero_std": 0.5, "grad_norm": 12.472508430480957, "learning_rate": 8.622126023955445e-07, "loss": 0.0, "num_tokens": 3047815.0, "reward": 0.546875, "reward_std": 0.2207425981760025, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 29 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 906.0, "completions/max_terminated_length": 906.0, "completions/mean_length": 68.953125, "completions/mean_terminated_length": 68.953125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.075, "frac_reward_zero_std": 0.5, "grad_norm": 11.43026065826416, "learning_rate": 8.506183921362442e-07, "loss": 0.0, "num_tokens": 3132036.0, "reward": 0.23828125, "reward_std": 0.1782316118478775, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "step": 30 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 84.0, "completions/max_terminated_length": 84.0, "completions/mean_length": 13.234375, "completions/mean_terminated_length": 13.234375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0775, "frac_reward_zero_std": 0.375, "grad_norm": 16.86567497253418, "learning_rate": 8.386407858128706e-07, "loss": 0.0, "num_tokens": 3201923.0, "reward": 0.8031250238418579, "reward_std": 0.2824692726135254, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.703125, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 31 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 390.0, "completions/max_terminated_length": 390.0, "completions/mean_length": 41.515625, "completions/mean_terminated_length": 41.515625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.08, "frac_reward_zero_std": 0.625, "grad_norm": 10.917762756347656, "learning_rate": 8.262928807620843e-07, "loss": 0.0, "num_tokens": 3273684.0, "reward": 0.3343750238418579, "reward_std": 0.1804211586713791, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "step": 32 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 908.0, "completions/max_terminated_length": 908.0, "completions/mean_length": 47.546875, "completions/mean_terminated_length": 47.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0825, "frac_reward_zero_std": 0.25, "grad_norm": 22.92045021057129, "learning_rate": 8.135881792367685e-07, "loss": -0.0, "num_tokens": 3381207.0, "reward": 0.33906251192092896, "reward_std": 0.2934828996658325, "rewards/format_reward/mean": 0.890625, "rewards/format_reward/std": 0.2083333432674408, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 33 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 279.0, "completions/max_terminated_length": 279.0, "completions/mean_length": 18.4375, "completions/mean_terminated_length": 18.4375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.085, "frac_reward_zero_std": 0.5, "grad_norm": 43.271331787109375, "learning_rate": 8.005405736415125e-07, "loss": 0.0, "num_tokens": 3464435.0, "reward": 0.3296875059604645, "reward_std": 0.10829188674688339, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.14689241349697113, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "step": 34 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 192.0, "completions/max_terminated_length": 192.0, "completions/mean_length": 16.140625, "completions/mean_terminated_length": 16.140625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0875, "frac_reward_zero_std": 0.75, "grad_norm": 4.92822265625, "learning_rate": 7.871643313414718e-07, "loss": 0.0, "num_tokens": 3551180.0, "reward": 0.6312500238418579, "reward_std": 0.1246790662407875, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 35 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 142.0, "completions/max_terminated_length": 142.0, "completions/mean_length": 14.28125, "completions/mean_terminated_length": 14.28125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.09, "frac_reward_zero_std": 0.625, "grad_norm": 8.274945259094238, "learning_rate": 7.734740790612136e-07, "loss": 0.0, "num_tokens": 3630430.0, "reward": 0.9429687857627869, "reward_std": 0.1955905258655548, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.84375, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "step": 36 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 304.0, "completions/max_terminated_length": 304.0, "completions/mean_length": 27.703125, "completions/mean_terminated_length": 27.703125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.0925, "frac_reward_zero_std": 0.75, "grad_norm": 8.869138717651367, "learning_rate": 7.594847868906076e-07, "loss": -0.0, "num_tokens": 3699035.0, "reward": 0.4750000238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 37 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 139.0, "completions/max_terminated_length": 139.0, "completions/mean_length": 14.0, "completions/mean_terminated_length": 14.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.095, "frac_reward_zero_std": 0.75, "grad_norm": 5.243677616119385, "learning_rate": 7.452117519152541e-07, "loss": 0.0, "num_tokens": 3777547.0, "reward": 0.42656251788139343, "reward_std": 0.06911306083202362, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 38 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 17.0, "completions/max_terminated_length": 17.0, "completions/mean_length": 12.03125, "completions/mean_terminated_length": 12.03125, "completions/min_length": 8.0, "completions/min_terminated_length": 8.0, "epoch": 0.0975, "frac_reward_zero_std": 0.625, "grad_norm": 9.26572036743164, "learning_rate": 7.306705814893439e-07, "loss": 0.0, "num_tokens": 3853357.0, "reward": 0.5984375476837158, "reward_std": 0.14709636569023132, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 39 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 21.0, "completions/max_terminated_length": 21.0, "completions/mean_length": 12.140625, "completions/mean_terminated_length": 12.140625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1, "frac_reward_zero_std": 0.625, "grad_norm": 9.793388366699219, "learning_rate": 7.158771761692464e-07, "loss": 0.0, "num_tokens": 3941350.0, "reward": 0.44218751788139343, "reward_std": 0.11330723762512207, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 40 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 39.0, "completions/max_terminated_length": 39.0, "completions/mean_length": 12.703125, "completions/mean_terminated_length": 12.703125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1025, "frac_reward_zero_std": 0.75, "grad_norm": 7.937704563140869, "learning_rate": 7.008477123264847e-07, "loss": -0.0, "num_tokens": 4002171.0, "reward": 0.4750000238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 41 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 147.0, "completions/max_terminated_length": 147.0, "completions/mean_length": 16.484375, "completions/mean_terminated_length": 16.484375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.105, "frac_reward_zero_std": 0.625, "grad_norm": 16.10399627685547, "learning_rate": 6.855986244591103e-07, "loss": 0.0, "num_tokens": 4089746.0, "reward": 0.4281250238418579, "reward_std": 0.15992169082164764, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 42 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 17.0, "completions/max_terminated_length": 17.0, "completions/mean_length": 12.109375, "completions/mean_terminated_length": 12.109375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1075, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 6.701465872208216e-07, "loss": 0.0, "num_tokens": 4181217.0, "reward": 0.6000000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 43 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.11, "frac_reward_zero_std": 0.75, "grad_norm": 7.800466537475586, "learning_rate": 6.545084971874736e-07, "loss": 0.0, "num_tokens": 4269889.0, "reward": 0.2250000238418579, "reward_std": 0.1157275140285492, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "step": 44 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1125, "frac_reward_zero_std": 0.75, "grad_norm": 8.696141242980957, "learning_rate": 6.387014543809223e-07, "loss": 0.0, "num_tokens": 4347441.0, "reward": 0.8187500238418579, "reward_std": 0.1293872892856598, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.71875, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 45 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.015625, "completions/mean_terminated_length": 12.015625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.115, "frac_reward_zero_std": 0.875, "grad_norm": 11.9544095993042, "learning_rate": 6.227427435703995e-07, "loss": 0.0, "num_tokens": 4416098.0, "reward": 0.5218750238418579, "reward_std": 0.0646936446428299, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 46 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1175, "frac_reward_zero_std": 0.5, "grad_norm": 13.539289474487305, "learning_rate": 6.066498153718734e-07, "loss": 0.0, "num_tokens": 4496778.0, "reward": 0.6781250238418579, "reward_std": 0.2109457552433014, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 47 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 684.0, "completions/max_terminated_length": 684.0, "completions/mean_length": 22.5, "completions/mean_terminated_length": 22.5, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.12, "frac_reward_zero_std": 0.5, "grad_norm": 7.529565334320068, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 4585818.0, "reward": 0.6156250238418579, "reward_std": 0.19044627249240875, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 48 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.03125, "completions/mean_terminated_length": 12.03125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1225, "frac_reward_zero_std": 0.375, "grad_norm": 22.16350555419922, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 4659644.0, "reward": 0.5375000238418579, "reward_std": 0.2845909595489502, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "step": 49 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.125, "frac_reward_zero_std": 0.5, "grad_norm": 39.590476989746094, "learning_rate": 5.577423184847931e-07, "loss": 0.0, "num_tokens": 4735756.0, "reward": 0.5062500238418579, "reward_std": 0.2130674123764038, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 50 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1275, "frac_reward_zero_std": 0.875, "grad_norm": 10.22485637664795, "learning_rate": 5.412896727361662e-07, "loss": 0.0, "num_tokens": 4804972.0, "reward": 0.6312500238418579, "reward_std": 0.0578637570142746, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 51 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 18.0, "completions/max_terminated_length": 18.0, "completions/mean_length": 12.421875, "completions/mean_terminated_length": 12.421875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.13, "frac_reward_zero_std": 0.875, "grad_norm": 8.306846618652344, "learning_rate": 5.247918773366111e-07, "loss": -0.0, "num_tokens": 4875847.0, "reward": 0.34062501788139343, "reward_std": 0.005786375608295202, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.29378482699394226, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 52 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1325, "frac_reward_zero_std": 0.875, "grad_norm": 8.738691329956055, "learning_rate": 5.082669723831793e-07, "loss": 0.0, "num_tokens": 4953159.0, "reward": 0.8812500238418579, "reward_std": 0.0578637570142746, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.78125, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 53 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 21.0, "completions/max_terminated_length": 21.0, "completions/mean_length": 12.140625, "completions/mean_terminated_length": 12.140625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.135, "frac_reward_zero_std": 0.875, "grad_norm": 4.081331729888916, "learning_rate": 4.917330276168208e-07, "loss": 0.0, "num_tokens": 5061584.0, "reward": 0.7093750238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 54 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1375, "frac_reward_zero_std": 0.75, "grad_norm": 10.815126419067383, "learning_rate": 4.752081226633888e-07, "loss": -0.0, "num_tokens": 5149504.0, "reward": 0.5375000238418579, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "step": 55 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.14, "frac_reward_zero_std": 0.75, "grad_norm": 9.388416290283203, "learning_rate": 4.5871032726383385e-07, "loss": 0.0, "num_tokens": 5237832.0, "reward": 0.6000000238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 56 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1425, "frac_reward_zero_std": 0.625, "grad_norm": 23.38650131225586, "learning_rate": 4.4225768151520694e-07, "loss": 0.0, "num_tokens": 5324544.0, "reward": 0.6468750238418579, "reward_std": 0.1893727034330368, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 57 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.0625, "completions/mean_terminated_length": 12.0625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.145, "frac_reward_zero_std": 0.625, "grad_norm": 23.5119571685791, "learning_rate": 4.258681761440789e-07, "loss": 0.0, "num_tokens": 5404412.0, "reward": 0.4437500238418579, "reward_std": 0.19149437546730042, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 58 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.0625, "completions/mean_terminated_length": 12.0625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1475, "frac_reward_zero_std": 0.625, "grad_norm": 14.529144287109375, "learning_rate": 4.095597328339452e-07, "loss": -0.0, "num_tokens": 5495104.0, "reward": 0.7250000238418579, "reward_std": 0.1462521106004715, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 59 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.15, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 3.9335018462812664e-07, "loss": 0.0, "num_tokens": 5589024.0, "reward": 0.6000000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 60 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.1875, "completions/mean_terminated_length": 12.1875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1525, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 3.772572564296004e-07, "loss": 0.0, "num_tokens": 5651684.0, "reward": 0.4750000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 61 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.155, "frac_reward_zero_std": 0.625, "grad_norm": 10.171913146972656, "learning_rate": 3.612985456190778e-07, "loss": 0.0, "num_tokens": 5739452.0, "reward": 0.6625000238418579, "reward_std": 0.1462521106004715, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5625, "rewards/mcq_exact_match_reward/std": 0.5, "step": 62 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 74.0, "completions/max_terminated_length": 74.0, "completions/mean_length": 13.0, "completions/mean_terminated_length": 13.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1575, "frac_reward_zero_std": 0.75, "grad_norm": 14.351091384887695, "learning_rate": 3.454915028125263e-07, "loss": -0.0, "num_tokens": 5815612.0, "reward": 0.48984378576278687, "reward_std": 0.046403881162405014, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 63 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.0625, "completions/mean_terminated_length": 12.0625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.16, "frac_reward_zero_std": 0.875, "grad_norm": 12.969809532165527, "learning_rate": 3.2985341277917846e-07, "loss": 0.0, "num_tokens": 5885248.0, "reward": 0.5218750238418579, "reward_std": 0.0646936446428299, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 64 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.1875, "completions/mean_terminated_length": 12.1875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1625, "frac_reward_zero_std": 0.75, "grad_norm": 19.828441619873047, "learning_rate": 3.1440137554088953e-07, "loss": 0.0, "num_tokens": 5979916.0, "reward": 0.8187500238418579, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.71875, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 65 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.165, "frac_reward_zero_std": 0.75, "grad_norm": 13.414032936096191, "learning_rate": 2.9915228767351535e-07, "loss": 0.0, "num_tokens": 6072084.0, "reward": 0.6000000238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 66 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 17.0, "completions/max_terminated_length": 17.0, "completions/mean_length": 12.34375, "completions/mean_terminated_length": 12.34375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1675, "frac_reward_zero_std": 0.75, "grad_norm": 16.610271453857422, "learning_rate": 2.841228238307536e-07, "loss": -0.0, "num_tokens": 6139618.0, "reward": 0.7867187857627869, "reward_std": 0.06902501732110977, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.6875, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 67 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.17, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 2.6932941851065615e-07, "loss": 0.0, "num_tokens": 6234266.0, "reward": 0.7250000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 68 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1725, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 2.547882480847461e-07, "loss": 0.0, "num_tokens": 6349570.0, "reward": 0.6000000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 69 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.175, "frac_reward_zero_std": 0.875, "grad_norm": 23.53253746032715, "learning_rate": 2.4051521310939254e-07, "loss": 0.0, "num_tokens": 6428258.0, "reward": 0.3968750238418579, "reward_std": 0.0646936446428299, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 70 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 108.0, "completions/max_terminated_length": 108.0, "completions/mean_length": 13.984375, "completions/mean_terminated_length": 13.984375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1775, "frac_reward_zero_std": 0.625, "grad_norm": 35.17768478393555, "learning_rate": 2.2652592093878665e-07, "loss": 0.0, "num_tokens": 6519457.0, "reward": 0.5062500238418579, "reward_std": 0.1962026059627533, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 71 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 12.015625, "completions/mean_terminated_length": 12.015625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.18, "frac_reward_zero_std": 0.625, "grad_norm": 23.059221267700195, "learning_rate": 2.128356686585282e-07, "loss": 0.0, "num_tokens": 6592186.0, "reward": 0.909375011920929, "reward_std": 0.11467419564723969, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.8125, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 72 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1825, "frac_reward_zero_std": 0.75, "grad_norm": 33.47691345214844, "learning_rate": 1.9945942635848745e-07, "loss": 0.0, "num_tokens": 6668098.0, "reward": 0.4906250238418579, "reward_std": 0.12255740165710449, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 73 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1068.0, "completions/max_terminated_length": 1068.0, "completions/mean_length": 43.53125, "completions/mean_terminated_length": 43.53125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.185, "frac_reward_zero_std": 0.75, "grad_norm": 18.336223602294922, "learning_rate": 1.8641182076323148e-07, "loss": 0.0, "num_tokens": 6765828.0, "reward": 0.4125000238418579, "reward_std": 0.1157275140285492, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 74 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 337.0, "completions/mean_length": 54.375, "completions/mean_terminated_length": 22.730159759521484, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1875, "frac_reward_zero_std": 0.625, "grad_norm": 18.733304977416992, "learning_rate": 1.7370711923791564e-07, "loss": 0.0, "num_tokens": 6831220.0, "reward": 0.59375, "reward_std": 0.15365058183670044, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 75 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.5625, "completions/mean_terminated_length": 12.5625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.19, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 1.6135921418712955e-07, "loss": 0.0, "num_tokens": 6909624.0, "reward": 0.8500000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.75, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 76 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.125, "completions/mean_terminated_length": 12.125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1925, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 1.493816078637557e-07, "loss": 0.0, "num_tokens": 7007592.0, "reward": 0.8500000238418579, "reward_std": 0.0, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.75, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 77 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.195, "frac_reward_zero_std": 0.75, "grad_norm": 9.707902908325195, "learning_rate": 1.3778739760445552e-07, "loss": -0.0, "num_tokens": 7090952.0, "reward": 0.5062500238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 78 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1975, "frac_reward_zero_std": 0.875, "grad_norm": 12.181106567382812, "learning_rate": 1.2658926150792322e-07, "loss": 0.0, "num_tokens": 7175512.0, "reward": 0.5843750238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 79 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 190.0, "completions/max_terminated_length": 190.0, "completions/mean_length": 15.5, "completions/mean_terminated_length": 15.5, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2, "frac_reward_zero_std": 0.625, "grad_norm": 25.612462997436523, "learning_rate": 1.1579944457157059e-07, "loss": 0.0, "num_tokens": 7238288.0, "reward": 0.7671874761581421, "reward_std": 0.16192206740379333, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.21304203569889069, "rewards/mcq_exact_match_reward/mean": 0.671875, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 80 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2025, "frac_reward_zero_std": 0.75, "grad_norm": 9.881487846374512, "learning_rate": 1.0542974530180327e-07, "loss": 0.0, "num_tokens": 7315232.0, "reward": 0.6937500238418579, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 81 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.205, "frac_reward_zero_std": 0.75, "grad_norm": 14.778446197509766, "learning_rate": 9.549150281252632e-08, "loss": -0.0, "num_tokens": 7387104.0, "reward": 0.5062500238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 82 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2075, "frac_reward_zero_std": 0.375, "grad_norm": 27.24396324157715, "learning_rate": 8.599558442598998e-08, "loss": 0.0, "num_tokens": 7459904.0, "reward": 0.7093750238418579, "reward_std": 0.24831004440784454, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 83 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 23.0, "completions/max_terminated_length": 23.0, "completions/mean_length": 12.796875, "completions/mean_terminated_length": 12.796875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.21, "frac_reward_zero_std": 0.875, "grad_norm": 6.658245086669922, "learning_rate": 7.695237378953224e-08, "loss": 0.0, "num_tokens": 7533923.0, "reward": 0.4843750298023224, "reward_std": 0.04712729901075363, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 84 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.515625, "completions/mean_terminated_length": 12.515625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2125, "frac_reward_zero_std": 0.875, "grad_norm": 11.308636665344238, "learning_rate": 6.837175952121304e-08, "loss": -0.0, "num_tokens": 7620268.0, "reward": 0.4906250238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 85 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 21.0, "completions/max_terminated_length": 21.0, "completions/mean_length": 12.546875, "completions/mean_terminated_length": 12.546875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.215, "frac_reward_zero_std": 0.875, "grad_norm": 22.906007766723633, "learning_rate": 6.026312439675551e-08, "loss": 0.0, "num_tokens": 7700767.0, "reward": 0.6109375357627869, "reward_std": 0.046501487493515015, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.21304203569889069, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 86 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2175, "frac_reward_zero_std": 0.625, "grad_norm": 16.732364654541016, "learning_rate": 5.263533508961826e-08, "loss": 0.0, "num_tokens": 7806775.0, "reward": 0.6312500238418579, "reward_std": 0.1462521106004715, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 87 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.22, "frac_reward_zero_std": 0.75, "grad_norm": 17.71401596069336, "learning_rate": 4.549673247541874e-08, "loss": 0.0, "num_tokens": 7894863.0, "reward": 0.5687500238418579, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 88 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.0625, "completions/mean_terminated_length": 12.0625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2225, "frac_reward_zero_std": 0.75, "grad_norm": 14.154738426208496, "learning_rate": 3.8855122511307626e-08, "loss": 0.0, "num_tokens": 7989947.0, "reward": 0.8031250238418579, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.703125, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 89 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.225, "frac_reward_zero_std": 0.75, "grad_norm": 13.400158882141113, "learning_rate": 3.271776770026963e-08, "loss": 0.0, "num_tokens": 8054307.0, "reward": 0.8968750238418579, "reward_std": 0.12255740165710449, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.796875, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 90 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.0625, "completions/mean_terminated_length": 12.0625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2275, "frac_reward_zero_std": 0.75, "grad_norm": 13.739428520202637, "learning_rate": 2.7091379149682682e-08, "loss": 0.0, "num_tokens": 8120047.0, "reward": 0.4750000238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 91 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.23, "frac_reward_zero_std": 0.875, "grad_norm": 9.631512641906738, "learning_rate": 2.1982109232821176e-08, "loss": 0.0, "num_tokens": 8195223.0, "reward": 0.8187500238418579, "reward_std": 0.0578637570142746, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.71875, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 92 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2325, "frac_reward_zero_std": 0.875, "grad_norm": 6.432887077331543, "learning_rate": 1.7395544861325718e-08, "loss": 0.0, "num_tokens": 8275647.0, "reward": 0.5843750238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 93 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.5, "completions/mean_terminated_length": 12.5, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.235, "frac_reward_zero_std": 0.625, "grad_norm": 24.301124572753906, "learning_rate": 1.3336701375997127e-08, "loss": 0.0, "num_tokens": 8335151.0, "reward": 0.5218750238418579, "reward_std": 0.1530819982290268, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 94 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2375, "frac_reward_zero_std": 0.75, "grad_norm": 15.701408386230469, "learning_rate": 9.810017062595321e-09, "loss": 0.0, "num_tokens": 8406895.0, "reward": 1.037500023841858, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.9375, "rewards/mcq_exact_match_reward/std": 0.24397502839565277, "step": 95 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.24, "frac_reward_zero_std": 0.75, "grad_norm": 11.823748588562012, "learning_rate": 6.819348298638839e-09, "loss": -0.0, "num_tokens": 8512639.0, "reward": 0.6312500238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 96 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 12.0, "completions/mean_terminated_length": 12.0, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2425, "frac_reward_zero_std": 0.625, "grad_norm": 12.021856307983398, "learning_rate": 4.367965336512403e-09, "loss": 0.0, "num_tokens": 8589575.0, "reward": 0.4906250238418579, "reward_std": 0.1893727034330368, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 97 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.1875, "completions/mean_terminated_length": 12.1875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.245, "frac_reward_zero_std": 0.625, "grad_norm": 14.939786911010742, "learning_rate": 2.458548727494292e-09, "loss": 0.0, "num_tokens": 8694291.0, "reward": 0.7093750238418579, "reward_std": 0.17358146607875824, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 98 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 302.0, "completions/max_terminated_length": 302.0, "completions/mean_length": 16.53125, "completions/mean_terminated_length": 16.53125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2475, "frac_reward_zero_std": 0.875, "grad_norm": 9.607973098754883, "learning_rate": 1.0931863906127325e-09, "loss": -0.0, "num_tokens": 8777405.0, "reward": 0.6156250238418579, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 99 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 12.25, "completions/mean_terminated_length": 12.25, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.25, "frac_reward_zero_std": 0.625, "grad_norm": 13.788451194763184, "learning_rate": 2.733713295369755e-10, "loss": 0.0, "num_tokens": 8868341.0, "reward": 0.8968750238418579, "reward_std": 0.1530819833278656, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.796875, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 100 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 8868341, "num_train_epochs": 1, "save_steps": 50, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }