{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.125, "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": 1210.0, "completions/max_terminated_length": 1210.0, "completions/mean_length": 456.578125, "completions/mean_terminated_length": 456.578125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0025, "frac_reward_zero_std": 0.0, "grad_norm": 6.1835432052612305, "learning_rate": 0.0, "loss": 0.0, "num_tokens": 98709.0, "reward": 0.3023437261581421, "reward_std": 0.3030892014503479, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.255761981010437, "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": 0.0, "completions/max_length": 1409.0, "completions/max_terminated_length": 1409.0, "completions/mean_length": 428.609375, "completions/mean_terminated_length": 428.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.005, "frac_reward_zero_std": 0.0, "grad_norm": 6.306211948394775, "learning_rate": 2e-07, "loss": 0.0, "num_tokens": 190116.0, "reward": 0.3867187201976776, "reward_std": 0.42795610427856445, "rewards/format_reward/mean": 0.4296875, "rewards/format_reward/std": 0.2798410654067993, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1647.0, "completions/max_terminated_length": 1647.0, "completions/mean_length": 541.171875, "completions/mean_terminated_length": 541.171875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0075, "frac_reward_zero_std": 0.0, "grad_norm": 5.829588413238525, "learning_rate": 4e-07, "loss": 0.0, "num_tokens": 315431.0, "reward": 0.3203125, "reward_std": 0.3534185290336609, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.2083333432674408, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1794.0, "completions/max_terminated_length": 1794.0, "completions/mean_length": 439.640625, "completions/mean_terminated_length": 439.640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.01, "frac_reward_zero_std": 0.125, "grad_norm": 7.357433795928955, "learning_rate": 6e-07, "loss": 0.0, "num_tokens": 437328.0, "reward": 0.46406248211860657, "reward_std": 0.386005163192749, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.27048972249031067, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 1588.0, "completions/max_terminated_length": 1588.0, "completions/mean_length": 517.078125, "completions/mean_terminated_length": 517.078125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0125, "frac_reward_zero_std": 0.0, "grad_norm": 4.498136520385742, "learning_rate": 8e-07, "loss": -0.0, "num_tokens": 561029.0, "reward": 0.3125, "reward_std": 0.33156174421310425, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.24397502839565277, "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": 1960.0, "completions/mean_length": 553.609375, "completions/mean_terminated_length": 529.888916015625, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.015, "frac_reward_zero_std": 0.0, "grad_norm": 10.413219451904297, "learning_rate": 1e-06, "loss": -0.0, "num_tokens": 697244.0, "reward": 0.20546874403953552, "reward_std": 0.2879093885421753, "rewards/format_reward/mean": 0.3359375, "rewards/format_reward/std": 0.30930912494659424, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 2019.0, "completions/mean_length": 612.140625, "completions/mean_terminated_length": 589.3492431640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0175, "frac_reward_zero_std": 0.125, "grad_norm": 8.344427108764648, "learning_rate": 9.99726628670463e-07, "loss": 0.0, "num_tokens": 821517.0, "reward": 0.26093751192092896, "reward_std": 0.2377234697341919, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.23935678601264954, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 1054.0, "completions/mean_length": 413.171875, "completions/mean_terminated_length": 387.2222595214844, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.02, "frac_reward_zero_std": 0.125, "grad_norm": 4.564246654510498, "learning_rate": 9.989068136093872e-07, "loss": -0.0, "num_tokens": 941120.0, "reward": 0.20078124105930328, "reward_std": 0.2517307698726654, "rewards/format_reward/mean": 0.2890625, "rewards/format_reward/std": 0.2928335666656494, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 1830.0, "completions/mean_length": 458.5625, "completions/mean_terminated_length": 433.3333435058594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0225, "frac_reward_zero_std": 0.125, "grad_norm": 5.534543037414551, "learning_rate": 9.975414512725056e-07, "loss": -0.0, "num_tokens": 1071980.0, "reward": 0.26171875, "reward_std": 0.2811623215675354, "rewards/format_reward/mean": 0.4296875, "rewards/format_reward/std": 0.2798410654067993, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 1176.0, "completions/max_terminated_length": 1176.0, "completions/mean_length": 347.671875, "completions/mean_terminated_length": 347.671875, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.025, "frac_reward_zero_std": 0.0, "grad_norm": 20.916383743286133, "learning_rate": 9.956320346634875e-07, "loss": -0.0, "num_tokens": 1191127.0, "reward": 0.2515624761581421, "reward_std": 0.3212316036224365, "rewards/format_reward/mean": 0.328125, "rewards/format_reward/std": 0.2847827076911926, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 1692.0, "completions/max_terminated_length": 1692.0, "completions/mean_length": 482.453125, "completions/mean_terminated_length": 482.453125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0275, "frac_reward_zero_std": 0.25, "grad_norm": 14.693345069885254, "learning_rate": 9.931806517013612e-07, "loss": 0.0, "num_tokens": 1321468.0, "reward": 0.16484373807907104, "reward_std": 0.2602120637893677, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.2387082874774933, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 1155.0, "completions/max_terminated_length": 1155.0, "completions/mean_length": 227.53125, "completions/mean_terminated_length": 227.53125, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.03, "frac_reward_zero_std": 0.0, "grad_norm": 96.4611587524414, "learning_rate": 9.901899829374047e-07, "loss": -0.0, "num_tokens": 1416582.0, "reward": 0.24453124403953552, "reward_std": 0.25452080368995667, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.27537402510643005, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 1446.0, "completions/mean_length": 234.0625, "completions/mean_terminated_length": 205.2698516845703, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.0325, "frac_reward_zero_std": 0.125, "grad_norm": 54.38446807861328, "learning_rate": 9.866632986240029e-07, "loss": 0.0, "num_tokens": 1524946.0, "reward": 0.171875, "reward_std": 0.2965255379676819, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.21593283116817474, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 1835.0, "completions/mean_length": 214.53125, "completions/mean_terminated_length": 185.4285888671875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.035, "frac_reward_zero_std": 0.5, "grad_norm": 9.216438293457031, "learning_rate": 9.826044551386742e-07, "loss": -0.0, "num_tokens": 1661188.0, "reward": 0.7367186546325684, "reward_std": 0.23405678570270538, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.1406387835741043, "rewards/mcq_exact_match_reward/mean": 0.6875, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 14 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1820.0, "completions/max_terminated_length": 1820.0, "completions/mean_length": 86.6875, "completions/mean_terminated_length": 86.6875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0375, "frac_reward_zero_std": 0.625, "grad_norm": 22.44569206237793, "learning_rate": 9.780178907671788e-07, "loss": -0.0, "num_tokens": 1762560.0, "reward": 0.34062498807907104, "reward_std": 0.09199240058660507, "rewards/format_reward/mean": 0.59375, "rewards/format_reward/std": 0.19669894874095917, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1781.0, "completions/max_terminated_length": 1781.0, "completions/mean_length": 35.84375, "completions/mean_terminated_length": 35.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04, "frac_reward_zero_std": 0.625, "grad_norm": 21.775293350219727, "learning_rate": 9.729086208503173e-07, "loss": 0.0, "num_tokens": 1870910.0, "reward": 0.3492187261581421, "reward_std": 0.10529261082410812, "rewards/format_reward/mean": 0.5234375, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 912.0, "completions/max_terminated_length": 912.0, "completions/mean_length": 49.9375, "completions/mean_terminated_length": 49.9375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0425, "frac_reward_zero_std": 0.25, "grad_norm": 29.980073928833008, "learning_rate": 9.672822322997304e-07, "loss": 0.0, "num_tokens": 1953066.0, "reward": 0.43203121423721313, "reward_std": 0.25999781489372253, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.23345555365085602, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 641.0, "completions/max_terminated_length": 641.0, "completions/mean_length": 46.40625, "completions/mean_terminated_length": 46.40625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.045, "frac_reward_zero_std": 0.625, "grad_norm": 16.981128692626953, "learning_rate": 9.611448774886923e-07, "loss": -0.0, "num_tokens": 2047812.0, "reward": 0.38203123211860657, "reward_std": 0.17039035260677338, "rewards/format_reward/mean": 0.5390625, "rewards/format_reward/std": 0.13524486124515533, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 18 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 6.0, "completions/max_terminated_length": 6.0, "completions/mean_length": 6.0, "completions/mean_terminated_length": 6.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0475, "frac_reward_zero_std": 0.375, "grad_norm": 34.2733039855957, "learning_rate": 9.545032675245813e-07, "loss": 0.0, "num_tokens": 2138780.0, "reward": 0.6124999523162842, "reward_std": 0.25513988733291626, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5625, "rewards/mcq_exact_match_reward/std": 0.5, "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": 130.0, "completions/max_terminated_length": 130.0, "completions/mean_length": 9.328125, "completions/mean_terminated_length": 9.328125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05, "frac_reward_zero_std": 0.375, "grad_norm": 27.232397079467773, "learning_rate": 9.473646649103817e-07, "loss": 0.0, "num_tokens": 2222433.0, "reward": 0.45546871423721313, "reward_std": 0.285457044839859, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 624.0, "completions/mean_length": 152.875, "completions/mean_terminated_length": 122.79365539550781, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0525, "frac_reward_zero_std": 0.375, "grad_norm": 15.760602951049805, "learning_rate": 9.397368756032444e-07, "loss": -0.0, "num_tokens": 2306673.0, "reward": 0.5406249761581421, "reward_std": 0.21701714396476746, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.208927720785141, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 610.0, "completions/mean_length": 77.96875, "completions/mean_terminated_length": 46.69841384887695, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.055, "frac_reward_zero_std": 0.25, "grad_norm": 14.47245979309082, "learning_rate": 9.316282404787869e-07, "loss": -0.0, "num_tokens": 2404767.0, "reward": 0.551562488079071, "reward_std": 0.31678467988967896, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 435.0, "completions/max_terminated_length": 435.0, "completions/mean_length": 47.609375, "completions/mean_terminated_length": 47.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0575, "frac_reward_zero_std": 0.25, "grad_norm": 17.22133445739746, "learning_rate": 9.230476262104676e-07, "loss": 0.0, "num_tokens": 2482174.0, "reward": 0.37187498807907104, "reward_std": 0.2584024667739868, "rewards/format_reward/mean": 0.59375, "rewards/format_reward/std": 0.21593283116817474, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 474.0, "completions/max_terminated_length": 474.0, "completions/mean_length": 34.65625, "completions/mean_terminated_length": 34.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06, "frac_reward_zero_std": 0.25, "grad_norm": 24.49734878540039, "learning_rate": 9.1400441557401e-07, "loss": -0.0, "num_tokens": 2566152.0, "reward": 0.19140625, "reward_std": 0.26238274574279785, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.18881812691688538, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "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": 640.0, "completions/max_terminated_length": 640.0, "completions/mean_length": 22.15625, "completions/mean_terminated_length": 22.15625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0625, "frac_reward_zero_std": 0.25, "grad_norm": 19.766305923461914, "learning_rate": 9.045084971874737e-07, "loss": -0.0, "num_tokens": 2656122.0, "reward": 0.48359373211860657, "reward_std": 0.3281433582305908, "rewards/format_reward/mean": 0.6171875, "rewards/format_reward/std": 0.21347814798355103, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 15.0, "completions/max_terminated_length": 15.0, "completions/mean_length": 7.75, "completions/mean_terminated_length": 7.75, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.065, "frac_reward_zero_std": 0.75, "grad_norm": 14.147409439086914, "learning_rate": 8.945702546981968e-07, "loss": 0.0, "num_tokens": 2752602.0, "reward": 0.20234376192092896, "reward_std": 0.13070547580718994, "rewards/format_reward/mean": 0.6171875, "rewards/format_reward/std": 0.21347814798355103, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "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": 296.0, "completions/max_terminated_length": 296.0, "completions/mean_length": 13.265625, "completions/mean_terminated_length": 13.265625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0675, "frac_reward_zero_std": 0.625, "grad_norm": 13.20996379852295, "learning_rate": 8.842005554284295e-07, "loss": -0.0, "num_tokens": 2824067.0, "reward": 0.4367187023162842, "reward_std": 0.09059804677963257, "rewards/format_reward/mean": 0.6171875, "rewards/format_reward/std": 0.21347814798355103, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 111.0, "completions/max_terminated_length": 111.0, "completions/mean_length": 8.296875, "completions/mean_terminated_length": 8.296875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07, "frac_reward_zero_std": 0.875, "grad_norm": 27.578950881958008, "learning_rate": 8.734107384920769e-07, "loss": -0.0, "num_tokens": 2920830.0, "reward": 0.8054686784744263, "reward_std": 0.002209708094596863, "rewards/format_reward/mean": 0.5546875, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.75, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 17.0, "completions/max_terminated_length": 17.0, "completions/mean_length": 6.640625, "completions/mean_terminated_length": 6.640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0725, "frac_reward_zero_std": 0.5, "grad_norm": 19.41982650756836, "learning_rate": 8.622126023955445e-07, "loss": -0.0, "num_tokens": 3014095.0, "reward": 0.3046875, "reward_std": 0.2165345400571823, "rewards/format_reward/mean": 0.546875, "rewards/format_reward/std": 0.14689241349697113, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 6.78125, "completions/mean_terminated_length": 6.78125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.075, "frac_reward_zero_std": 0.625, "grad_norm": 12.432705879211426, "learning_rate": 8.506183921362442e-07, "loss": 0.0, "num_tokens": 3090281.0, "reward": 0.5406249761581421, "reward_std": 0.1530819833278656, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 6.765625, "completions/mean_terminated_length": 6.765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0775, "frac_reward_zero_std": 0.5, "grad_norm": 20.327106475830078, "learning_rate": 8.386407858128706e-07, "loss": -0.0, "num_tokens": 3175322.0, "reward": 0.3843749761581421, "reward_std": 0.2561880052089691, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 356.0, "completions/max_terminated_length": 356.0, "completions/mean_length": 12.140625, "completions/mean_terminated_length": 12.140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08, "frac_reward_zero_std": 0.375, "grad_norm": 16.507497787475586, "learning_rate": 8.262928807620843e-07, "loss": 0.0, "num_tokens": 3271651.0, "reward": 0.30546873807907104, "reward_std": 0.23685014247894287, "rewards/format_reward/mean": 0.5546875, "rewards/format_reward/std": 0.15728822350502014, "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": 180.0, "completions/max_terminated_length": 180.0, "completions/mean_length": 10.96875, "completions/mean_terminated_length": 10.96875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0825, "frac_reward_zero_std": 0.5, "grad_norm": 12.270334243774414, "learning_rate": 8.135881792367685e-07, "loss": -0.0, "num_tokens": 3364625.0, "reward": 0.3343749940395355, "reward_std": 0.22673699259757996, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 7.5625, "completions/mean_terminated_length": 7.5625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.085, "frac_reward_zero_std": 0.375, "grad_norm": 26.52332305908203, "learning_rate": 8.005405736415125e-07, "loss": -0.0, "num_tokens": 3448837.0, "reward": 0.4531249701976776, "reward_std": 0.24831001460552216, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 93.0, "completions/max_terminated_length": 93.0, "completions/mean_length": 8.765625, "completions/mean_terminated_length": 8.765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0875, "frac_reward_zero_std": 0.625, "grad_norm": 20.24908447265625, "learning_rate": 7.871643313414718e-07, "loss": -0.0, "num_tokens": 3529142.0, "reward": 0.515625, "reward_std": 0.1530819833278656, "rewards/format_reward/mean": 0.625, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 12.0, "completions/max_terminated_length": 12.0, "completions/mean_length": 6.75, "completions/mean_terminated_length": 6.75, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09, "frac_reward_zero_std": 0.5, "grad_norm": 24.714027404785156, "learning_rate": 7.734740790612136e-07, "loss": 0.0, "num_tokens": 3642918.0, "reward": 0.4000000059604645, "reward_std": 0.22461532056331635, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 6.78125, "completions/mean_terminated_length": 6.78125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0925, "frac_reward_zero_std": 0.5, "grad_norm": 15.52238941192627, "learning_rate": 7.594847868906076e-07, "loss": -0.0, "num_tokens": 3741496.0, "reward": 0.5562499761581421, "reward_std": 0.2041158378124237, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 9.0, "completions/max_terminated_length": 9.0, "completions/mean_length": 6.125, "completions/mean_terminated_length": 6.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.095, "frac_reward_zero_std": 0.625, "grad_norm": 20.026525497436523, "learning_rate": 7.452117519152541e-07, "loss": 0.0, "num_tokens": 3833632.0, "reward": 0.5015624761581421, "reward_std": 0.15526625514030457, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 8.390625, "completions/mean_terminated_length": 8.390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0975, "frac_reward_zero_std": 0.625, "grad_norm": 19.0867977142334, "learning_rate": 7.306705814893439e-07, "loss": -0.0, "num_tokens": 3933001.0, "reward": 0.4781249761581421, "reward_std": 0.1530819833278656, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.3021090030670166, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 8.265625, "completions/mean_terminated_length": 8.265625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1, "frac_reward_zero_std": 0.625, "grad_norm": 16.062389373779297, "learning_rate": 7.158771761692464e-07, "loss": -0.0, "num_tokens": 4016090.0, "reward": 0.8031250238418579, "reward_std": 0.1804211586713791, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.734375, "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": 0.0, "completions/max_length": 128.0, "completions/max_terminated_length": 128.0, "completions/mean_length": 10.421875, "completions/mean_terminated_length": 10.421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1025, "frac_reward_zero_std": 0.75, "grad_norm": 19.808956146240234, "learning_rate": 7.008477123264847e-07, "loss": -0.0, "num_tokens": 4098005.0, "reward": 0.41718751192092896, "reward_std": 0.04708736017346382, "rewards/format_reward/mean": 0.578125, "rewards/format_reward/std": 0.18298126757144928, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 6.375, "completions/mean_terminated_length": 6.375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.105, "frac_reward_zero_std": 0.5, "grad_norm": 42.87741470336914, "learning_rate": 6.855986244591103e-07, "loss": 0.0, "num_tokens": 4216309.0, "reward": 0.49062496423721313, "reward_std": 0.17009234428405762, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 13.0, "completions/max_terminated_length": 13.0, "completions/mean_length": 8.078125, "completions/mean_terminated_length": 8.078125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1075, "frac_reward_zero_std": 0.375, "grad_norm": 49.60019302368164, "learning_rate": 6.701465872208216e-07, "loss": 0.0, "num_tokens": 4291514.0, "reward": 0.45781248807907104, "reward_std": 0.22963017225265503, "rewards/format_reward/mean": 0.671875, "rewards/format_reward/std": 0.23935678601264954, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 139.0, "completions/max_terminated_length": 139.0, "completions/mean_length": 12.328125, "completions/mean_terminated_length": 12.328125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.11, "frac_reward_zero_std": 0.5, "grad_norm": 27.64950180053711, "learning_rate": 6.545084971874736e-07, "loss": 0.0, "num_tokens": 4392287.0, "reward": 0.66015625, "reward_std": 0.09210129082202911, "rewards/format_reward/mean": 0.6640625, "rewards/format_reward/std": 0.25283610820770264, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 6.75, "completions/mean_terminated_length": 6.75, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1125, "frac_reward_zero_std": 0.5, "grad_norm": 33.55113220214844, "learning_rate": 6.387014543809223e-07, "loss": -0.0, "num_tokens": 4503087.0, "reward": 0.6812499761581421, "reward_std": 0.19727617502212524, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.1666666716337204, "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": 190.0, "completions/max_terminated_length": 190.0, "completions/mean_length": 13.796875, "completions/mean_terminated_length": 13.796875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.115, "frac_reward_zero_std": 0.625, "grad_norm": 24.258800506591797, "learning_rate": 6.227427435703995e-07, "loss": 0.0, "num_tokens": 4575274.0, "reward": 0.5093749761581421, "reward_std": 0.07260168343782425, "rewards/format_reward/mean": 0.71875, "rewards/format_reward/std": 0.25, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 254.0, "completions/max_terminated_length": 254.0, "completions/mean_length": 14.71875, "completions/mean_terminated_length": 14.71875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1175, "frac_reward_zero_std": 0.375, "grad_norm": 59.530765533447266, "learning_rate": 6.066498153718734e-07, "loss": 0.0, "num_tokens": 4678936.0, "reward": 0.7367187738418579, "reward_std": 0.22119058668613434, "rewards/format_reward/mean": 0.8046875, "rewards/format_reward/std": 0.24587368965148926, "rewards/mcq_exact_match_reward/mean": 0.65625, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 130.0, "completions/max_terminated_length": 130.0, "completions/mean_length": 12.5, "completions/mean_terminated_length": 12.5, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.12, "frac_reward_zero_std": 0.5, "grad_norm": 38.72102355957031, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 4778856.0, "reward": 0.6085937023162842, "reward_std": 0.14830872416496277, "rewards/format_reward/mean": 0.7734375, "rewards/format_reward/std": 0.250866562128067, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 95.0, "completions/max_terminated_length": 95.0, "completions/mean_length": 12.640625, "completions/mean_terminated_length": 12.640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1225, "frac_reward_zero_std": 0.875, "grad_norm": 8.399174690246582, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 4870049.0, "reward": 0.4781249761581421, "reward_std": 0.04419416934251785, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.2182178944349289, "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": 80.0, "completions/max_terminated_length": 80.0, "completions/mean_length": 11.96875, "completions/mean_terminated_length": 11.96875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.125, "frac_reward_zero_std": 0.625, "grad_norm": 16.016551971435547, "learning_rate": 5.577423184847931e-07, "loss": 0.0, "num_tokens": 4961535.0, "reward": 0.8093750476837158, "reward_std": 0.16769355535507202, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.19669894874095917, "rewards/mcq_exact_match_reward/mean": 0.71875, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 50 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 4961535, "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 }