{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.11160714285714286, "eval_steps": 500, "global_step": 50, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1702.0, "completions/mean_length": 599.390625, "completions/mean_terminated_length": 576.3968505859375, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "epoch": 0.002232142857142857, "frac_reward_zero_std": 0.125, "grad_norm": 3.039191484451294, "learning_rate": 0.0, "loss": -0.0, "num_tokens": 126017.0, "reward": 0.3007812201976776, "reward_std": 0.3582488000392914, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.2302463799715042, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 1963.0, "completions/mean_length": 522.0, "completions/mean_terminated_length": 497.7778015136719, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.004464285714285714, "frac_reward_zero_std": 0.0, "grad_norm": 6.132624626159668, "learning_rate": 2e-07, "loss": -0.0, "num_tokens": 270345.0, "reward": 0.27968746423721313, "reward_std": 0.3124620020389557, "rewards/format_reward/mean": 0.296875, "rewards/format_reward/std": 0.24750742316246033, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 2 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1562.0, "completions/max_terminated_length": 1562.0, "completions/mean_length": 521.453125, "completions/mean_terminated_length": 521.453125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.006696428571428571, "frac_reward_zero_std": 0.0, "grad_norm": 5.321042537689209, "learning_rate": 4e-07, "loss": 0.0, "num_tokens": 384158.0, "reward": 0.33828121423721313, "reward_std": 0.33478155732154846, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.244862899184227, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1824.0, "completions/mean_length": 566.625, "completions/mean_terminated_length": 543.1111450195312, "completions/min_length": 19.0, "completions/min_terminated_length": 19.0, "epoch": 0.008928571428571428, "frac_reward_zero_std": 0.125, "grad_norm": 4.300935745239258, "learning_rate": 6e-07, "loss": 0.0, "num_tokens": 491766.0, "reward": 0.28437501192092896, "reward_std": 0.4092683792114258, "rewards/format_reward/mean": 0.34375, "rewards/format_reward/std": 0.265398770570755, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1882.0, "completions/mean_length": 502.921875, "completions/mean_terminated_length": 478.3968505859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.011160714285714286, "frac_reward_zero_std": 0.125, "grad_norm": 5.058187961578369, "learning_rate": 8e-07, "loss": -0.0, "num_tokens": 624329.0, "reward": 0.32109373807907104, "reward_std": 0.33916139602661133, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.25479042530059814, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1884.0, "completions/mean_length": 472.765625, "completions/mean_terminated_length": 447.7619323730469, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.013392857142857142, "frac_reward_zero_std": 0.0, "grad_norm": 7.012692451477051, "learning_rate": 1e-06, "loss": -0.0, "num_tokens": 752274.0, "reward": 0.31171873211860657, "reward_std": 0.4537365436553955, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.2567298710346222, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 1048.0, "completions/max_terminated_length": 1048.0, "completions/mean_length": 457.828125, "completions/mean_terminated_length": 457.828125, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.015625, "frac_reward_zero_std": 0.0, "grad_norm": 1.8473551273345947, "learning_rate": 9.99726628670463e-07, "loss": -0.0, "num_tokens": 851575.0, "reward": 0.3999999761581421, "reward_std": 0.4424627423286438, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.33184191584587097, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1625.0, "completions/mean_length": 573.46875, "completions/mean_terminated_length": 525.9031982421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.017857142857142856, "frac_reward_zero_std": 0.0, "grad_norm": 6.523875713348389, "learning_rate": 9.989068136093872e-07, "loss": -0.0, "num_tokens": 1012597.0, "reward": 0.25703123211860657, "reward_std": 0.3503851294517517, "rewards/format_reward/mean": 0.3828125, "rewards/format_reward/std": 0.2780628502368927, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1183.0, "completions/mean_length": 479.78125, "completions/mean_terminated_length": 454.888916015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.020089285714285716, "frac_reward_zero_std": 0.0, "grad_norm": 3.1459288597106934, "learning_rate": 9.975414512725056e-07, "loss": -0.0, "num_tokens": 1129599.0, "reward": 0.4296874701976776, "reward_std": 0.4543060064315796, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.28824523091316223, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 2027.0, "completions/max_terminated_length": 2027.0, "completions/mean_length": 420.078125, "completions/mean_terminated_length": 420.078125, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.022321428571428572, "frac_reward_zero_std": 0.0, "grad_norm": 28.552722930908203, "learning_rate": 9.956320346634875e-07, "loss": -0.0, "num_tokens": 1254508.0, "reward": 0.23515623807907104, "reward_std": 0.4202980697154999, "rewards/format_reward/mean": 0.3203125, "rewards/format_reward/std": 0.30035942792892456, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1344.0, "completions/mean_length": 436.671875, "completions/mean_terminated_length": 411.09527587890625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.024553571428571428, "frac_reward_zero_std": 0.0, "grad_norm": 9.229541778564453, "learning_rate": 9.931806517013612e-07, "loss": 0.0, "num_tokens": 1377511.0, "reward": 0.31562501192092896, "reward_std": 0.3412793278694153, "rewards/format_reward/mean": 0.34375, "rewards/format_reward/std": 0.279951810836792, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1980.0, "completions/max_terminated_length": 1980.0, "completions/mean_length": 527.609375, "completions/mean_terminated_length": 527.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.026785714285714284, "frac_reward_zero_std": 0.0, "grad_norm": 8.898627281188965, "learning_rate": 9.901899829374047e-07, "loss": -0.0, "num_tokens": 1532278.0, "reward": 0.38749998807907104, "reward_std": 0.35279834270477295, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.22712838649749756, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1841.0, "completions/mean_length": 615.421875, "completions/mean_terminated_length": 592.6825561523438, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "epoch": 0.029017857142857144, "frac_reward_zero_std": 0.0, "grad_norm": 1.8077502250671387, "learning_rate": 9.866632986240029e-07, "loss": 0.0, "num_tokens": 1686001.0, "reward": 0.19218748807907104, "reward_std": 0.2915004789829254, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.2592533528804779, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1679.0, "completions/mean_length": 498.109375, "completions/mean_terminated_length": 448.1128845214844, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03125, "frac_reward_zero_std": 0.0, "grad_norm": 3.8591432571411133, "learning_rate": 9.826044551386742e-07, "loss": 0.0, "num_tokens": 1832504.0, "reward": 0.16171874105930328, "reward_std": 0.3050600588321686, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.2708333432674408, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1878.0, "completions/mean_length": 368.40625, "completions/mean_terminated_length": 314.2257995605469, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.033482142857142856, "frac_reward_zero_std": 0.125, "grad_norm": 16.3862247467041, "learning_rate": 9.780178907671788e-07, "loss": -0.0, "num_tokens": 1932050.0, "reward": 0.5445312857627869, "reward_std": 0.43138980865478516, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.3463779389858246, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 1045.0, "completions/max_terminated_length": 1045.0, "completions/mean_length": 236.40625, "completions/mean_terminated_length": 236.40625, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.03571428571428571, "frac_reward_zero_std": 0.0, "grad_norm": 14.02306079864502, "learning_rate": 9.729086208503173e-07, "loss": 0.0, "num_tokens": 2008804.0, "reward": 0.4281250238418579, "reward_std": 0.28502750396728516, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.3673154413700104, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 16 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1199.0, "completions/max_terminated_length": 1199.0, "completions/mean_length": 361.015625, "completions/mean_terminated_length": 361.015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03794642857142857, "frac_reward_zero_std": 0.0, "grad_norm": 9.910021781921387, "learning_rate": 9.672822322997304e-07, "loss": -0.0, "num_tokens": 2122037.0, "reward": 0.4976562261581421, "reward_std": 0.3884924352169037, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.2538151443004608, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 0.0, "completions/max_length": 1548.0, "completions/max_terminated_length": 1548.0, "completions/mean_length": 428.578125, "completions/mean_terminated_length": 428.578125, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "epoch": 0.04017857142857143, "frac_reward_zero_std": 0.0, "grad_norm": 6.756157875061035, "learning_rate": 9.611448774886923e-07, "loss": -0.0, "num_tokens": 2240658.0, "reward": 0.34843748807907104, "reward_std": 0.3557349443435669, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.301698237657547, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1136.0, "completions/mean_length": 438.0625, "completions/mean_terminated_length": 412.5079650878906, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04241071428571429, "frac_reward_zero_std": 0.0, "grad_norm": 13.136040687561035, "learning_rate": 9.545032675245813e-07, "loss": 0.0, "num_tokens": 2355630.0, "reward": 0.3148437738418579, "reward_std": 0.3863198757171631, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.32723212242126465, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1413.0, "completions/mean_length": 484.421875, "completions/mean_terminated_length": 433.9838562011719, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.044642857142857144, "frac_reward_zero_std": 0.0, "grad_norm": 7.762112617492676, "learning_rate": 9.473646649103817e-07, "loss": 0.0, "num_tokens": 2466873.0, "reward": 0.45625001192092896, "reward_std": 0.367803692817688, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.28171807527542114, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 1627.0, "completions/max_terminated_length": 1627.0, "completions/mean_length": 357.65625, "completions/mean_terminated_length": 357.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.046875, "frac_reward_zero_std": 0.125, "grad_norm": 14.70483112335205, "learning_rate": 9.397368756032444e-07, "loss": -0.0, "num_tokens": 2587139.0, "reward": 0.35234373807907104, "reward_std": 0.3356337249279022, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.22146137058734894, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1873.0, "completions/mean_length": 405.578125, "completions/mean_terminated_length": 324.8032531738281, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.049107142857142856, "frac_reward_zero_std": 0.125, "grad_norm": 16.061939239501953, "learning_rate": 9.316282404787869e-07, "loss": -0.0, "num_tokens": 2714952.0, "reward": 0.6874999403953552, "reward_std": 0.2616451382637024, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.25, "rewards/mcq_exact_match_reward/mean": 0.640625, "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": 1735.0, "completions/max_terminated_length": 1735.0, "completions/mean_length": 296.234375, "completions/mean_terminated_length": 296.234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05133928571428571, "frac_reward_zero_std": 0.0, "grad_norm": 10.560757637023926, "learning_rate": 9.230476262104676e-07, "loss": -0.0, "num_tokens": 2820519.0, "reward": 0.48749998211860657, "reward_std": 0.4836316406726837, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 775.0, "completions/max_terminated_length": 775.0, "completions/mean_length": 219.328125, "completions/mean_terminated_length": 219.328125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05357142857142857, "frac_reward_zero_std": 0.0, "grad_norm": 18.072036743164062, "learning_rate": 9.1400441557401e-07, "loss": -0.0, "num_tokens": 2913068.0, "reward": 0.4664062261581421, "reward_std": 0.4330206513404846, "rewards/format_reward/mean": 0.6015625, "rewards/format_reward/std": 0.28423789143562317, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 871.0, "completions/mean_length": 230.734375, "completions/mean_terminated_length": 201.88890075683594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05580357142857143, "frac_reward_zero_std": 0.25, "grad_norm": 16.38471031188965, "learning_rate": 9.045084971874737e-07, "loss": 0.0, "num_tokens": 3022987.0, "reward": 0.23124998807907104, "reward_std": 0.2328900247812271, "rewards/format_reward/mean": 0.59375, "rewards/format_reward/std": 0.279951810836792, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1607.0, "completions/mean_length": 264.234375, "completions/mean_terminated_length": 235.920654296875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05803571428571429, "frac_reward_zero_std": 0.125, "grad_norm": 12.785871505737305, "learning_rate": 8.945702546981968e-07, "loss": 0.0, "num_tokens": 3131082.0, "reward": 0.20468750596046448, "reward_std": 0.2825583517551422, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.2514837086200714, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1263.0, "completions/mean_length": 243.1875, "completions/mean_terminated_length": 214.53968811035156, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.060267857142857144, "frac_reward_zero_std": 0.25, "grad_norm": 20.926250457763672, "learning_rate": 8.842005554284295e-07, "loss": 0.0, "num_tokens": 3233622.0, "reward": 0.2906250059604645, "reward_std": 0.32549476623535156, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.28867512941360474, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 678.0, "completions/max_terminated_length": 678.0, "completions/mean_length": 219.015625, "completions/mean_terminated_length": 219.015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0625, "frac_reward_zero_std": 0.25, "grad_norm": 19.252267837524414, "learning_rate": 8.734107384920769e-07, "loss": -0.0, "num_tokens": 3335495.0, "reward": 0.45625001192092896, "reward_std": 0.14555320143699646, "rewards/format_reward/mean": 0.65625, "rewards/format_reward/std": 0.265398770570755, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 938.0, "completions/max_terminated_length": 938.0, "completions/mean_length": 137.8125, "completions/mean_terminated_length": 137.8125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06473214285714286, "frac_reward_zero_std": 0.125, "grad_norm": 25.66993522644043, "learning_rate": 8.622126023955445e-07, "loss": 0.0, "num_tokens": 3438643.0, "reward": 0.37109375, "reward_std": 0.24377912282943726, "rewards/format_reward/mean": 0.5859375, "rewards/format_reward/std": 0.28942593932151794, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.078125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1295.0, "completions/mean_length": 489.796875, "completions/mean_terminated_length": 357.7457580566406, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06696428571428571, "frac_reward_zero_std": 0.0, "grad_norm": 11.776336669921875, "learning_rate": 8.506183921362442e-07, "loss": 0.0, "num_tokens": 3570398.0, "reward": 0.38359373807907104, "reward_std": 0.2550874352455139, "rewards/format_reward/mean": 0.5546875, "rewards/format_reward/std": 0.3463779389858246, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1191.0, "completions/mean_length": 245.3125, "completions/mean_terminated_length": 216.69842529296875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06919642857142858, "frac_reward_zero_std": 0.25, "grad_norm": 14.228128433227539, "learning_rate": 8.386407858128706e-07, "loss": -0.0, "num_tokens": 3680066.0, "reward": 0.42421871423721313, "reward_std": 0.3229408264160156, "rewards/format_reward/mean": 0.6484375, "rewards/format_reward/std": 0.29113471508026123, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1292.0, "completions/mean_length": 381.890625, "completions/mean_terminated_length": 299.9508056640625, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "epoch": 0.07142857142857142, "frac_reward_zero_std": 0.0, "grad_norm": 10.392132759094238, "learning_rate": 8.262928807620843e-07, "loss": 0.0, "num_tokens": 3796651.0, "reward": 0.3695312440395355, "reward_std": 0.3850117325782776, "rewards/format_reward/mean": 0.7265625, "rewards/format_reward/std": 0.3555644750595093, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 793.0, "completions/mean_length": 157.203125, "completions/mean_terminated_length": 127.19048309326172, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07366071428571429, "frac_reward_zero_std": 0.0, "grad_norm": 13.654447555541992, "learning_rate": 8.135881792367685e-07, "loss": -0.0, "num_tokens": 3884096.0, "reward": 0.3695312440395355, "reward_std": 0.3136134743690491, "rewards/format_reward/mean": 0.7265625, "rewards/format_reward/std": 0.2807259261608124, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 593.0, "completions/mean_length": 115.921875, "completions/mean_terminated_length": 53.596771240234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07589285714285714, "frac_reward_zero_std": 0.25, "grad_norm": 14.979105949401855, "learning_rate": 8.005405736415125e-07, "loss": 0.0, "num_tokens": 3993571.0, "reward": 0.5625, "reward_std": 0.2877725660800934, "rewards/format_reward/mean": 0.78125, "rewards/format_reward/std": 0.33184191584587097, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 616.0, "completions/mean_length": 142.671875, "completions/mean_terminated_length": 112.42858123779297, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.078125, "frac_reward_zero_std": 0.25, "grad_norm": 7.067958831787109, "learning_rate": 7.871643313414718e-07, "loss": 0.0, "num_tokens": 4079990.0, "reward": 0.45625001192092896, "reward_std": 0.30316486954689026, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.2745848298072815, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 914.0, "completions/mean_length": 136.546875, "completions/mean_terminated_length": 74.88709259033203, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08035714285714286, "frac_reward_zero_std": 0.125, "grad_norm": 17.58712387084961, "learning_rate": 7.734740790612136e-07, "loss": -0.0, "num_tokens": 4179425.0, "reward": 0.5210937261581421, "reward_std": 0.34353575110435486, "rewards/format_reward/mean": 0.6796875, "rewards/format_reward/std": 0.30035942792892456, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 494.0, "completions/mean_length": 147.15625, "completions/mean_terminated_length": 85.83870697021484, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08258928571428571, "frac_reward_zero_std": 0.125, "grad_norm": 24.559791564941406, "learning_rate": 7.594847868906076e-07, "loss": 0.0, "num_tokens": 4278715.0, "reward": 0.40625, "reward_std": 0.24086612462997437, "rewards/format_reward/mean": 0.78125, "rewards/format_reward/std": 0.33184191584587097, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 536.0, "completions/mean_length": 164.734375, "completions/mean_terminated_length": 72.11475372314453, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "epoch": 0.08482142857142858, "frac_reward_zero_std": 0.375, "grad_norm": 11.3303861618042, "learning_rate": 7.452117519152541e-07, "loss": -0.0, "num_tokens": 4362602.0, "reward": 0.5164062976837158, "reward_std": 0.14126159250736237, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.2203386276960373, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1278.0, "completions/mean_length": 149.515625, "completions/mean_terminated_length": 119.3809585571289, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08705357142857142, "frac_reward_zero_std": 0.375, "grad_norm": 4.936997413635254, "learning_rate": 7.306705814893439e-07, "loss": -0.0, "num_tokens": 4488955.0, "reward": 0.4609375, "reward_std": 0.27706843614578247, "rewards/format_reward/mean": 0.859375, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 0.0625, "completions/max_length": 2048.0, "completions/max_terminated_length": 321.0, "completions/mean_length": 171.0625, "completions/mean_terminated_length": 45.93333435058594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08928571428571429, "frac_reward_zero_std": 0.25, "grad_norm": 6.486630916595459, "learning_rate": 7.158771761692464e-07, "loss": 0.0, "num_tokens": 4575783.0, "reward": 0.3515625, "reward_std": 0.20217931270599365, "rewards/format_reward/mean": 0.859375, "rewards/format_reward/std": 0.301698237657547, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 761.0, "completions/mean_length": 150.234375, "completions/mean_terminated_length": 120.11111450195312, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09151785714285714, "frac_reward_zero_std": 0.25, "grad_norm": 8.273109436035156, "learning_rate": 7.008477123264847e-07, "loss": 0.0, "num_tokens": 4673086.0, "reward": 0.4156250059604645, "reward_std": 0.3660367727279663, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.26726123690605164, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 482.0, "completions/mean_length": 105.90625, "completions/mean_terminated_length": 43.25806427001953, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09375, "frac_reward_zero_std": 0.375, "grad_norm": 13.193151473999023, "learning_rate": 6.855986244591103e-07, "loss": -0.0, "num_tokens": 4787264.0, "reward": 0.725781261920929, "reward_std": 0.26460421085357666, "rewards/format_reward/mean": 0.8515625, "rewards/format_reward/std": 0.27716949582099915, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 1239.0, "completions/max_terminated_length": 1239.0, "completions/mean_length": 82.359375, "completions/mean_terminated_length": 82.359375, "completions/min_length": 18.0, "completions/min_terminated_length": 18.0, "epoch": 0.09598214285714286, "frac_reward_zero_std": 0.5, "grad_norm": 9.40908145904541, "learning_rate": 6.701465872208216e-07, "loss": 0.0, "num_tokens": 4890263.0, "reward": 0.45781251788139343, "reward_std": 0.2623252868652344, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 663.0, "completions/mean_length": 120.53125, "completions/mean_terminated_length": 58.354835510253906, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09821428571428571, "frac_reward_zero_std": 0.5, "grad_norm": 4.324347019195557, "learning_rate": 6.545084971874736e-07, "loss": 0.0, "num_tokens": 4983257.0, "reward": 0.13593751192092896, "reward_std": 0.11201904714107513, "rewards/format_reward/mean": 0.890625, "rewards/format_reward/std": 0.2592533528804779, "rewards/mcq_exact_match_reward/mean": 0.046875, "rewards/mcq_exact_match_reward/std": 0.21304203569889069, "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": 840.0, "completions/max_terminated_length": 840.0, "completions/mean_length": 66.765625, "completions/mean_terminated_length": 66.765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10044642857142858, "frac_reward_zero_std": 0.25, "grad_norm": 12.35912799835205, "learning_rate": 6.387014543809223e-07, "loss": -0.0, "num_tokens": 5077778.0, "reward": 0.717968761920929, "reward_std": 0.35076451301574707, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.1751912236213684, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 1023.0, "completions/max_terminated_length": 1023.0, "completions/mean_length": 51.0625, "completions/mean_terminated_length": 51.0625, "completions/min_length": 17.0, "completions/min_terminated_length": 17.0, "epoch": 0.10267857142857142, "frac_reward_zero_std": 0.75, "grad_norm": 5.579497337341309, "learning_rate": 6.227427435703995e-07, "loss": 0.0, "num_tokens": 5161278.0, "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": 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": 82.0, "completions/max_terminated_length": 82.0, "completions/mean_length": 20.46875, "completions/mean_terminated_length": 20.46875, "completions/min_length": 17.0, "completions/min_terminated_length": 17.0, "epoch": 0.10491071428571429, "frac_reward_zero_std": 0.5, "grad_norm": 15.84175968170166, "learning_rate": 6.066498153718734e-07, "loss": 0.0, "num_tokens": 5240732.0, "reward": 0.8343750238418579, "reward_std": 0.2382849156856537, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 144.0, "completions/max_terminated_length": 144.0, "completions/mean_length": 19.25, "completions/mean_terminated_length": 19.25, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10714285714285714, "frac_reward_zero_std": 0.5, "grad_norm": 9.524042129516602, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 5339300.0, "reward": 0.4375, "reward_std": 0.22461533546447754, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 423.0, "completions/mean_length": 57.03125, "completions/mean_terminated_length": 25.428573608398438, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.109375, "frac_reward_zero_std": 0.5, "grad_norm": 10.890397071838379, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 5432182.0, "reward": 0.48515626788139343, "reward_std": 0.15907271206378937, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.180765300989151, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 488.0, "completions/max_terminated_length": 488.0, "completions/mean_length": 28.03125, "completions/mean_terminated_length": 28.03125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.11160714285714286, "frac_reward_zero_std": 0.375, "grad_norm": 6.575137615203857, "learning_rate": 5.577423184847931e-07, "loss": 0.0, "num_tokens": 5546896.0, "reward": 0.8343750238418579, "reward_std": 0.22097086906433105, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 50 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 5546896, "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": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }