{ "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": 0.0, "completions/max_length": 1786.0, "completions/max_terminated_length": 1786.0, "completions/mean_length": 541.859375, "completions/mean_terminated_length": 541.859375, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "epoch": 0.002232142857142857, "frac_reward_zero_std": 0.0, "grad_norm": 12.971773147583008, "learning_rate": 0.0, "loss": -0.0, "num_tokens": 112399.0, "reward": 0.25312498211860657, "reward_std": 0.35192999243736267, "rewards/format_reward/mean": 0.34375, "rewards/format_reward/std": 0.25, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 1 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1913.0, "completions/max_terminated_length": 1913.0, "completions/mean_length": 460.625, "completions/mean_terminated_length": 460.625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.004464285714285714, "frac_reward_zero_std": 0.0, "grad_norm": 9.931671142578125, "learning_rate": 2e-07, "loss": -0.0, "num_tokens": 239831.0, "reward": 0.33671873807907104, "reward_std": 0.35391804575920105, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.20275264978408813, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 1663.0, "completions/max_terminated_length": 1663.0, "completions/mean_length": 578.5625, "completions/mean_terminated_length": 578.5625, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "epoch": 0.006696428571428571, "frac_reward_zero_std": 0.0, "grad_norm": 2.283196210861206, "learning_rate": 4e-07, "loss": 0.0, "num_tokens": 357427.0, "reward": 0.28125, "reward_std": 0.3595561385154724, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.279951810836792, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 1725.0, "completions/mean_length": 508.953125, "completions/mean_terminated_length": 484.5238342285156, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.008928571428571428, "frac_reward_zero_std": 0.0, "grad_norm": 14.620399475097656, "learning_rate": 6e-07, "loss": 0.0, "num_tokens": 498696.0, "reward": 0.38203126192092896, "reward_std": 0.4592018127441406, "rewards/format_reward/mean": 0.3828125, "rewards/format_reward/std": 0.2634054720401764, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 4 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1176.0, "completions/max_terminated_length": 1176.0, "completions/mean_length": 443.328125, "completions/mean_terminated_length": 443.328125, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.011160714285714286, "frac_reward_zero_std": 0.125, "grad_norm": 23.866165161132812, "learning_rate": 8e-07, "loss": -0.0, "num_tokens": 606045.0, "reward": 0.23671872913837433, "reward_std": 0.23968853056430817, "rewards/format_reward/mean": 0.3359375, "rewards/format_reward/std": 0.2824873626232147, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1733.0, "completions/mean_length": 569.5, "completions/mean_terminated_length": 521.8064575195312, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.013392857142857142, "frac_reward_zero_std": 0.25, "grad_norm": 4.394739627838135, "learning_rate": 1e-06, "loss": -0.0, "num_tokens": 764053.0, "reward": 0.3164062201976776, "reward_std": 0.24395309388637543, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.2302463799715042, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1963.0, "completions/mean_length": 531.71875, "completions/mean_terminated_length": 507.65081787109375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.015625, "frac_reward_zero_std": 0.0, "grad_norm": 9.567261695861816, "learning_rate": 9.99726628670463e-07, "loss": -0.0, "num_tokens": 895587.0, "reward": 0.2867187261581421, "reward_std": 0.34152650833129883, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.22257846593856812, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 7 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1868.0, "completions/max_terminated_length": 1868.0, "completions/mean_length": 575.71875, "completions/mean_terminated_length": 575.71875, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "epoch": 0.017857142857142856, "frac_reward_zero_std": 0.0, "grad_norm": 5.7189435958862305, "learning_rate": 9.989068136093872e-07, "loss": 0.0, "num_tokens": 1015801.0, "reward": 0.26093748211860657, "reward_std": 0.2545996308326721, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.20351573824882507, "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": 1845.0, "completions/mean_length": 584.140625, "completions/mean_terminated_length": 560.90478515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.020089285714285716, "frac_reward_zero_std": 0.125, "grad_norm": 6.5300421714782715, "learning_rate": 9.975414512725056e-07, "loss": -0.0, "num_tokens": 1153538.0, "reward": 0.3515625, "reward_std": 0.40456390380859375, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1695.0, "completions/mean_length": 586.046875, "completions/mean_terminated_length": 562.84130859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.022321428571428572, "frac_reward_zero_std": 0.25, "grad_norm": 4.844789028167725, "learning_rate": 9.956320346634875e-07, "loss": -0.0, "num_tokens": 1279109.0, "reward": 0.31874996423721313, "reward_std": 0.30643922090530396, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 2034.0, "completions/mean_length": 403.140625, "completions/mean_terminated_length": 350.08062744140625, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.024553571428571428, "frac_reward_zero_std": 0.25, "grad_norm": 25.20620346069336, "learning_rate": 9.931806517013612e-07, "loss": -0.0, "num_tokens": 1407374.0, "reward": 0.35468748211860657, "reward_std": 0.24326452612876892, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.2221602201461792, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1665.0, "completions/mean_length": 540.59375, "completions/mean_terminated_length": 491.96771240234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.026785714285714284, "frac_reward_zero_std": 0.0, "grad_norm": 10.811991691589355, "learning_rate": 9.901899829374047e-07, "loss": -0.0, "num_tokens": 1536228.0, "reward": 0.2890624701976776, "reward_std": 0.38433414697647095, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1231.0, "completions/mean_length": 437.3125, "completions/mean_terminated_length": 385.3548278808594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.029017857142857144, "frac_reward_zero_std": 0.125, "grad_norm": 9.110626220703125, "learning_rate": 9.866632986240029e-07, "loss": -0.0, "num_tokens": 1661912.0, "reward": 0.38593748211860657, "reward_std": 0.3386583626270294, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.18298126757144928, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 13 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1989.0, "completions/max_terminated_length": 1989.0, "completions/mean_length": 396.1875, "completions/mean_terminated_length": 396.1875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03125, "frac_reward_zero_std": 0.0, "grad_norm": 22.122051239013672, "learning_rate": 9.826044551386742e-07, "loss": 0.0, "num_tokens": 1785052.0, "reward": 0.5171874761581421, "reward_std": 0.37966397404670715, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.29504841566085815, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 1319.0, "completions/mean_length": 398.1875, "completions/mean_terminated_length": 344.9677429199219, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.033482142857142856, "frac_reward_zero_std": 0.25, "grad_norm": 8.21301555633545, "learning_rate": 9.780178907671788e-07, "loss": -0.0, "num_tokens": 1906464.0, "reward": 0.23906248807907104, "reward_std": 0.214445561170578, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.1985812783241272, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 0.078125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1598.0, "completions/mean_length": 539.109375, "completions/mean_terminated_length": 411.2372741699219, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "epoch": 0.03571428571428571, "frac_reward_zero_std": 0.0, "grad_norm": 17.956470489501953, "learning_rate": 9.729086208503173e-07, "loss": -0.0, "num_tokens": 2039351.0, "reward": 0.4976562261581421, "reward_std": 0.4534168541431427, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.26899561285972595, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 1142.0, "completions/mean_length": 333.8125, "completions/mean_terminated_length": 306.6031799316406, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03794642857142857, "frac_reward_zero_std": 0.125, "grad_norm": 19.72380256652832, "learning_rate": 9.672822322997304e-07, "loss": -0.0, "num_tokens": 2139947.0, "reward": 0.3046875, "reward_std": 0.3565414547920227, "rewards/format_reward/mean": 0.546875, "rewards/format_reward/std": 0.2916666865348816, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 1142.0, "completions/mean_length": 248.0625, "completions/mean_terminated_length": 219.49208068847656, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.04017857142857143, "frac_reward_zero_std": 0.375, "grad_norm": 27.903902053833008, "learning_rate": 9.611448774886923e-07, "loss": 0.0, "num_tokens": 2281071.0, "reward": 0.18984374403953552, "reward_std": 0.21500104665756226, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.18881812691688538, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "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": 1478.0, "completions/mean_length": 237.8125, "completions/mean_terminated_length": 209.07937622070312, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04241071428571429, "frac_reward_zero_std": 0.125, "grad_norm": 23.812612533569336, "learning_rate": 9.545032675245813e-07, "loss": -0.0, "num_tokens": 2410507.0, "reward": 0.3968749940395355, "reward_std": 0.37050747871398926, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.21593283116817474, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 832.0, "completions/max_terminated_length": 832.0, "completions/mean_length": 102.78125, "completions/mean_terminated_length": 102.78125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.044642857142857144, "frac_reward_zero_std": 0.125, "grad_norm": 17.30659294128418, "learning_rate": 9.473646649103817e-07, "loss": -0.0, "num_tokens": 2492773.0, "reward": 0.5757812261581421, "reward_std": 0.4401569366455078, "rewards/format_reward/mean": 0.6015625, "rewards/format_reward/std": 0.28423789143562317, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 685.0, "completions/mean_length": 71.796875, "completions/mean_terminated_length": 40.42857360839844, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.046875, "frac_reward_zero_std": 0.625, "grad_norm": 10.079203605651855, "learning_rate": 9.397368756032444e-07, "loss": 0.0, "num_tokens": 2624568.0, "reward": 0.10312500596046448, "reward_std": 0.10811922699213028, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.22712838649749756, "rewards/mcq_exact_match_reward/mean": 0.046875, "rewards/mcq_exact_match_reward/std": 0.21304203569889069, "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": 1416.0, "completions/max_terminated_length": 1416.0, "completions/mean_length": 105.59375, "completions/mean_terminated_length": 105.59375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.049107142857142856, "frac_reward_zero_std": 0.375, "grad_norm": 15.665273666381836, "learning_rate": 9.316282404787869e-07, "loss": -0.0, "num_tokens": 2746590.0, "reward": 0.3062500059604645, "reward_std": 0.26257947087287903, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.18898223340511322, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1835.0, "completions/mean_length": 111.84375, "completions/mean_terminated_length": 81.11111450195312, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05133928571428571, "frac_reward_zero_std": 0.375, "grad_norm": 17.90364646911621, "learning_rate": 9.230476262104676e-07, "loss": -0.0, "num_tokens": 2852316.0, "reward": 0.26093751192092896, "reward_std": 0.11917313188314438, "rewards/format_reward/mean": 0.578125, "rewards/format_reward/std": 0.2221602201461792, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 721.0, "completions/max_terminated_length": 721.0, "completions/mean_length": 59.4375, "completions/mean_terminated_length": 59.4375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05357142857142857, "frac_reward_zero_std": 0.125, "grad_norm": 23.127439498901367, "learning_rate": 9.1400441557401e-07, "loss": 0.0, "num_tokens": 2948880.0, "reward": 0.4039062261581421, "reward_std": 0.3720070719718933, "rewards/format_reward/mean": 0.6015625, "rewards/format_reward/std": 0.28423789143562317, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 473.0, "completions/max_terminated_length": 473.0, "completions/mean_length": 38.484375, "completions/mean_terminated_length": 38.484375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05580357142857143, "frac_reward_zero_std": 0.125, "grad_norm": 23.131183624267578, "learning_rate": 9.045084971874737e-07, "loss": -0.0, "num_tokens": 3042983.0, "reward": 0.49609375, "reward_std": 0.39612793922424316, "rewards/format_reward/mean": 0.5859375, "rewards/format_reward/std": 0.244862899184227, "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": 396.0, "completions/max_terminated_length": 396.0, "completions/mean_length": 14.578125, "completions/mean_terminated_length": 14.578125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05803571428571429, "frac_reward_zero_std": 0.375, "grad_norm": 21.188722610473633, "learning_rate": 8.945702546981968e-07, "loss": -0.0, "num_tokens": 3153444.0, "reward": 0.26874998211860657, "reward_std": 0.29173365235328674, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.08908708393573761, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 19.0, "completions/max_terminated_length": 19.0, "completions/mean_length": 8.234375, "completions/mean_terminated_length": 8.234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.060267857142857144, "frac_reward_zero_std": 0.0, "grad_norm": 31.691007614135742, "learning_rate": 8.842005554284295e-07, "loss": 0.0, "num_tokens": 3245059.0, "reward": 0.3148437738418579, "reward_std": 0.4270421266555786, "rewards/format_reward/mean": 0.6484375, "rewards/format_reward/std": 0.2302463799715042, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 910.0, "completions/max_terminated_length": 910.0, "completions/mean_length": 23.265625, "completions/mean_terminated_length": 23.265625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0625, "frac_reward_zero_std": 0.375, "grad_norm": 24.359390258789062, "learning_rate": 8.734107384920769e-07, "loss": 0.0, "num_tokens": 3361140.0, "reward": 0.40156251192092896, "reward_std": 0.28653684258461, "rewards/format_reward/mean": 0.578125, "rewards/format_reward/std": 0.18298126757144928, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 237.0, "completions/max_terminated_length": 237.0, "completions/mean_length": 9.609375, "completions/mean_terminated_length": 9.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06473214285714286, "frac_reward_zero_std": 0.625, "grad_norm": 17.99944305419922, "learning_rate": 8.622126023955445e-07, "loss": 0.0, "num_tokens": 3490331.0, "reward": 0.20546874403953552, "reward_std": 0.17656923830509186, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 20.0, "completions/max_terminated_length": 20.0, "completions/mean_length": 6.625, "completions/mean_terminated_length": 6.625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06696428571428571, "frac_reward_zero_std": 0.75, "grad_norm": 20.938568115234375, "learning_rate": 8.506183921362442e-07, "loss": 0.0, "num_tokens": 3605843.0, "reward": 0.2874999940395355, "reward_std": 0.11888101696968079, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 33.0, "completions/max_terminated_length": 33.0, "completions/mean_length": 6.421875, "completions/mean_terminated_length": 6.421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06919642857142858, "frac_reward_zero_std": 0.5, "grad_norm": 20.33233070373535, "learning_rate": 8.386407858128706e-07, "loss": -0.0, "num_tokens": 3710350.0, "reward": 0.44062498211860657, "reward_std": 0.210945725440979, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 6.109375, "completions/mean_terminated_length": 6.109375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07142857142857142, "frac_reward_zero_std": 0.5, "grad_norm": 26.781448364257812, "learning_rate": 8.262928807620843e-07, "loss": 0.0, "num_tokens": 3807701.0, "reward": 0.2984374761581421, "reward_std": 0.16964475810527802, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 6.84375, "completions/mean_terminated_length": 6.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07366071428571429, "frac_reward_zero_std": 0.5, "grad_norm": 22.09980583190918, "learning_rate": 8.135881792367685e-07, "loss": 0.0, "num_tokens": 3908627.0, "reward": 0.39921873807907104, "reward_std": 0.1689612716436386, "rewards/format_reward/mean": 0.5546875, "rewards/format_reward/std": 0.180765300989151, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 568.0, "completions/max_terminated_length": 568.0, "completions/mean_length": 19.828125, "completions/mean_terminated_length": 19.828125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07589285714285714, "frac_reward_zero_std": 0.25, "grad_norm": 24.214155197143555, "learning_rate": 8.005405736415125e-07, "loss": 0.0, "num_tokens": 3997160.0, "reward": 0.3937499523162842, "reward_std": 0.3061639666557312, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 476.0, "completions/max_terminated_length": 476.0, "completions/mean_length": 16.640625, "completions/mean_terminated_length": 16.640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.078125, "frac_reward_zero_std": 0.25, "grad_norm": 19.65945053100586, "learning_rate": 7.871643313414718e-07, "loss": -0.0, "num_tokens": 4102257.0, "reward": 0.33124998211860657, "reward_std": 0.3424546718597412, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 158.0, "completions/max_terminated_length": 158.0, "completions/mean_length": 9.828125, "completions/mean_terminated_length": 9.828125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08035714285714286, "frac_reward_zero_std": 0.5, "grad_norm": 16.9875545501709, "learning_rate": 7.734740790612136e-07, "loss": 0.0, "num_tokens": 4218214.0, "reward": 0.40625, "reward_std": 0.22461532056331635, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 148.0, "completions/max_terminated_length": 148.0, "completions/mean_length": 16.703125, "completions/mean_terminated_length": 16.703125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08258928571428571, "frac_reward_zero_std": 0.625, "grad_norm": 15.114908218383789, "learning_rate": 7.594847868906076e-07, "loss": -0.0, "num_tokens": 4304195.0, "reward": 0.5093749761581421, "reward_std": 0.1893727034330368, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 6.8125, "completions/mean_terminated_length": 6.8125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08482142857142858, "frac_reward_zero_std": 0.375, "grad_norm": 16.950824737548828, "learning_rate": 7.452117519152541e-07, "loss": -0.0, "num_tokens": 4430631.0, "reward": 0.5249999761581421, "reward_std": 0.29143062233924866, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 221.0, "completions/max_terminated_length": 221.0, "completions/mean_length": 10.125, "completions/mean_terminated_length": 10.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08705357142857142, "frac_reward_zero_std": 0.5, "grad_norm": 25.844154357910156, "learning_rate": 7.306705814893439e-07, "loss": -0.0, "num_tokens": 4522799.0, "reward": 0.4312499761581421, "reward_std": 0.2177756428718567, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "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": 0.0, "completions/max_length": 60.0, "completions/max_terminated_length": 60.0, "completions/mean_length": 8.203125, "completions/mean_terminated_length": 8.203125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08928571428571429, "frac_reward_zero_std": 0.625, "grad_norm": 31.361331939697266, "learning_rate": 7.158771761692464e-07, "loss": 0.0, "num_tokens": 4608428.0, "reward": 0.6703124642372131, "reward_std": 0.049638569355010986, "rewards/format_reward/mean": 0.609375, "rewards/format_reward/std": 0.2083333432674408, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 8.90625, "completions/mean_terminated_length": 8.90625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09151785714285714, "frac_reward_zero_std": 0.375, "grad_norm": 17.690139770507812, "learning_rate": 7.008477123264847e-07, "loss": 0.0, "num_tokens": 4691318.0, "reward": 0.5726562738418579, "reward_std": 0.2708982527256012, "rewards/format_reward/mean": 0.7265625, "rewards/format_reward/std": 0.250866562128067, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 6.09375, "completions/mean_terminated_length": 6.09375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09375, "frac_reward_zero_std": 0.5, "grad_norm": 18.50206756591797, "learning_rate": 6.855986244591103e-07, "loss": -0.0, "num_tokens": 4789988.0, "reward": 0.4257812201976776, "reward_std": 0.2243541181087494, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 359.0, "completions/max_terminated_length": 359.0, "completions/mean_length": 16.03125, "completions/mean_terminated_length": 16.03125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09598214285714286, "frac_reward_zero_std": 0.25, "grad_norm": 54.797157287597656, "learning_rate": 6.701465872208216e-07, "loss": -0.0, "num_tokens": 4877334.0, "reward": 0.8117187023162842, "reward_std": 0.18093490600585938, "rewards/format_reward/mean": 0.6171875, "rewards/format_reward/std": 0.21347814798355103, "rewards/mcq_exact_match_reward/mean": 0.75, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 7.515625, "completions/mean_terminated_length": 7.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09821428571428571, "frac_reward_zero_std": 0.625, "grad_norm": 12.843533515930176, "learning_rate": 6.545084971874736e-07, "loss": 0.0, "num_tokens": 4983247.0, "reward": 0.59375, "reward_std": 0.16675157845020294, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 8.375, "completions/mean_terminated_length": 8.375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10044642857142858, "frac_reward_zero_std": 0.375, "grad_norm": 21.998193740844727, "learning_rate": 6.387014543809223e-07, "loss": -0.0, "num_tokens": 5059511.0, "reward": 0.49140626192092896, "reward_std": 0.20160752534866333, "rewards/format_reward/mean": 0.6953125, "rewards/format_reward/std": 0.24587368965148926, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 7.546875, "completions/mean_terminated_length": 7.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10267857142857142, "frac_reward_zero_std": 0.375, "grad_norm": 28.897733688354492, "learning_rate": 6.227427435703995e-07, "loss": -0.0, "num_tokens": 5150106.0, "reward": 0.28125, "reward_std": 0.28247907757759094, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 6.109375, "completions/mean_terminated_length": 6.109375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10491071428571429, "frac_reward_zero_std": 0.5, "grad_norm": 23.489076614379883, "learning_rate": 6.066498153718734e-07, "loss": -0.0, "num_tokens": 5243569.0, "reward": 0.5351561903953552, "reward_std": 0.18003448843955994, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 16.0, "completions/max_terminated_length": 16.0, "completions/mean_length": 9.90625, "completions/mean_terminated_length": 9.90625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10714285714285714, "frac_reward_zero_std": 0.75, "grad_norm": 7.710181713104248, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 5309691.0, "reward": 0.17500001192092896, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.09375, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "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": 43.0, "completions/max_terminated_length": 43.0, "completions/mean_length": 8.8125, "completions/mean_terminated_length": 8.8125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.109375, "frac_reward_zero_std": 0.25, "grad_norm": 21.57369613647461, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 5411943.0, "reward": 0.5812499523162842, "reward_std": 0.2802783250808716, "rewards/format_reward/mean": 0.65625, "rewards/format_reward/std": 0.25, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 572.0, "completions/max_terminated_length": 572.0, "completions/mean_length": 20.78125, "completions/mean_terminated_length": 20.78125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.11160714285714286, "frac_reward_zero_std": 0.125, "grad_norm": 31.959693908691406, "learning_rate": 5.577423184847931e-07, "loss": 0.0, "num_tokens": 5512089.0, "reward": 0.40937498211860657, "reward_std": 0.20373183488845825, "rewards/format_reward/mean": 0.65625, "rewards/format_reward/std": 0.233588308095932, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 50 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 5512089, "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 }