{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.25, "eval_steps": 500, "global_step": 100, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1368.0, "completions/max_terminated_length": 1368.0, "completions/mean_length": 466.640625, "completions/mean_terminated_length": 466.640625, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.0025, "frac_reward_zero_std": 0.0, "grad_norm": 3.6276707649230957, "learning_rate": 0.0, "loss": 0.0, "num_tokens": 99089.0, "reward": 0.25468748807907104, "reward_std": 0.3456140458583832, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.32694777846336365, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1289.0, "completions/mean_length": 510.84375, "completions/mean_terminated_length": 486.4444580078125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.005, "frac_reward_zero_std": 0.0, "grad_norm": 5.58949613571167, "learning_rate": 2e-07, "loss": 0.0, "num_tokens": 214919.0, "reward": 0.15781250596046448, "reward_std": 0.24574220180511475, "rewards/format_reward/mean": 0.328125, "rewards/format_reward/std": 0.31140682101249695, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "step": 2 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1549.0, "completions/mean_length": 451.578125, "completions/mean_terminated_length": 426.2381286621094, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.0075, "frac_reward_zero_std": 0.0, "grad_norm": 16.275543212890625, "learning_rate": 4e-07, "loss": -0.0, "num_tokens": 330628.0, "reward": 0.16874998807907104, "reward_std": 0.23435896635055542, "rewards/format_reward/mean": 0.28125, "rewards/format_reward/std": 0.279951810836792, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "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": 1385.0, "completions/mean_length": 414.9375, "completions/mean_terminated_length": 389.0158996582031, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.01, "frac_reward_zero_std": 0.125, "grad_norm": 10.092217445373535, "learning_rate": 6e-07, "loss": -0.0, "num_tokens": 445192.0, "reward": 0.4273437261581421, "reward_std": 0.30322587490081787, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.23974503576755524, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 1389.0, "completions/mean_length": 500.90625, "completions/mean_terminated_length": 476.3492431640625, "completions/min_length": 124.0, "completions/min_terminated_length": 124.0, "epoch": 0.0125, "frac_reward_zero_std": 0.0, "grad_norm": 1.7393639087677002, "learning_rate": 8e-07, "loss": 0.0, "num_tokens": 550650.0, "reward": 0.2562499940395355, "reward_std": 0.2739989161491394, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.26726123690605164, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 5 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1952.0, "completions/max_terminated_length": 1952.0, "completions/mean_length": 513.84375, "completions/mean_terminated_length": 513.84375, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "epoch": 0.015, "frac_reward_zero_std": 0.125, "grad_norm": 3.056152105331421, "learning_rate": 1e-06, "loss": 0.0, "num_tokens": 658600.0, "reward": 0.13515624403953552, "reward_std": 0.21850848197937012, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.27537402510643005, "rewards/mcq_exact_match_reward/mean": 0.09375, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "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": 1545.0, "completions/mean_length": 496.0, "completions/mean_terminated_length": 471.3651123046875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0175, "frac_reward_zero_std": 0.0, "grad_norm": 7.945972442626953, "learning_rate": 9.99726628670463e-07, "loss": -0.0, "num_tokens": 801456.0, "reward": 0.3140624761581421, "reward_std": 0.4141543507575989, "rewards/format_reward/mean": 0.328125, "rewards/format_reward/std": 0.23935678601264954, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1745.0, "completions/max_terminated_length": 1745.0, "completions/mean_length": 493.015625, "completions/mean_terminated_length": 493.015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.02, "frac_reward_zero_std": 0.0, "grad_norm": 2.0856635570526123, "learning_rate": 9.989068136093872e-07, "loss": 0.0, "num_tokens": 909249.0, "reward": 0.3374999761581421, "reward_std": 0.4492857754230499, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.279951810836792, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 8 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1234.0, "completions/max_terminated_length": 1234.0, "completions/mean_length": 340.203125, "completions/mean_terminated_length": 340.203125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0225, "frac_reward_zero_std": 0.0, "grad_norm": 7.63404655456543, "learning_rate": 9.975414512725056e-07, "loss": -0.0, "num_tokens": 1008406.0, "reward": 0.41484373807907104, "reward_std": 0.43470299243927, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.32342031598091125, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 1352.0, "completions/max_terminated_length": 1352.0, "completions/mean_length": 456.28125, "completions/mean_terminated_length": 456.28125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.025, "frac_reward_zero_std": 0.0, "grad_norm": 7.498995780944824, "learning_rate": 9.956320346634875e-07, "loss": -0.0, "num_tokens": 1125568.0, "reward": 0.16171874105930328, "reward_std": 0.2626776397228241, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.23974503576755524, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 983.0, "completions/mean_length": 485.421875, "completions/mean_terminated_length": 460.61907958984375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0275, "frac_reward_zero_std": 0.0, "grad_norm": 5.1299896240234375, "learning_rate": 9.931806517013612e-07, "loss": 0.0, "num_tokens": 1227979.0, "reward": 0.3476562798023224, "reward_std": 0.2908340096473694, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.3020063638687134, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 1975.0, "completions/max_terminated_length": 1975.0, "completions/mean_length": 468.90625, "completions/mean_terminated_length": 468.90625, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.03, "frac_reward_zero_std": 0.0, "grad_norm": 9.794697761535645, "learning_rate": 9.901899829374047e-07, "loss": 0.0, "num_tokens": 1351461.0, "reward": 0.15078124403953552, "reward_std": 0.2603379487991333, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.30282649397850037, "rewards/mcq_exact_match_reward/mean": 0.109375, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "step": 12 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1106.0, "completions/mean_length": 483.25, "completions/mean_terminated_length": 458.4127197265625, "completions/min_length": 77.0, "completions/min_terminated_length": 77.0, "epoch": 0.0325, "frac_reward_zero_std": 0.0, "grad_norm": 1.724840760231018, "learning_rate": 9.866632986240029e-07, "loss": 0.0, "num_tokens": 1444797.0, "reward": 0.2984375059604645, "reward_std": 0.400360643863678, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.356000155210495, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 1661.0, "completions/mean_length": 643.390625, "completions/mean_terminated_length": 621.0952758789062, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.035, "frac_reward_zero_std": 0.125, "grad_norm": 1.3986207246780396, "learning_rate": 9.826044551386742e-07, "loss": 0.0, "num_tokens": 1562494.0, "reward": 0.12421873956918716, "reward_std": 0.16812779009342194, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.28597769141197205, "rewards/mcq_exact_match_reward/mean": 0.078125, "rewards/mcq_exact_match_reward/std": 0.27048972249031067, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1757.0, "completions/mean_length": 706.015625, "completions/mean_terminated_length": 640.016357421875, "completions/min_length": 267.0, "completions/min_terminated_length": 267.0, "epoch": 0.0375, "frac_reward_zero_std": 0.125, "grad_norm": 0.8747150301933289, "learning_rate": 9.780178907671788e-07, "loss": 0.0, "num_tokens": 1701055.0, "reward": 0.20546874403953552, "reward_std": 0.30746352672576904, "rewards/format_reward/mean": 0.3359375, "rewards/format_reward/std": 0.25283610820770264, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1073.0, "completions/mean_length": 565.84375, "completions/mean_terminated_length": 518.0322265625, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.04, "frac_reward_zero_std": 0.125, "grad_norm": 1.3505909442901611, "learning_rate": 9.729086208503173e-07, "loss": 0.0, "num_tokens": 1804197.0, "reward": 0.203125, "reward_std": 0.3033024072647095, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.25, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 1656.0, "completions/max_terminated_length": 1656.0, "completions/mean_length": 520.71875, "completions/mean_terminated_length": 520.71875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0425, "frac_reward_zero_std": 0.0, "grad_norm": 6.046233177185059, "learning_rate": 9.672822322997304e-07, "loss": -0.0, "num_tokens": 1925067.0, "reward": 0.37031248211860657, "reward_std": 0.36943358182907104, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.2221602201461792, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 1373.0, "completions/mean_length": 644.5625, "completions/mean_terminated_length": 622.2857666015625, "completions/min_length": 189.0, "completions/min_terminated_length": 189.0, "epoch": 0.045, "frac_reward_zero_std": 0.0, "grad_norm": 1.2121214866638184, "learning_rate": 9.611448774886923e-07, "loss": -0.0, "num_tokens": 2070607.0, "reward": 0.36562496423721313, "reward_std": 0.4727749824523926, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.2357022762298584, "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": 1937.0, "completions/max_terminated_length": 1937.0, "completions/mean_length": 517.84375, "completions/mean_terminated_length": 517.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0475, "frac_reward_zero_std": 0.0, "grad_norm": 3.286956310272217, "learning_rate": 9.545032675245813e-07, "loss": -0.0, "num_tokens": 2198317.0, "reward": 0.29531246423721313, "reward_std": 0.4393918812274933, "rewards/format_reward/mean": 0.453125, "rewards/format_reward/std": 0.30496877431869507, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 1840.0, "completions/mean_length": 614.125, "completions/mean_terminated_length": 567.8709716796875, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.05, "frac_reward_zero_std": 0.125, "grad_norm": 1.2146230936050415, "learning_rate": 9.473646649103817e-07, "loss": 0.0, "num_tokens": 2330621.0, "reward": 0.3695312738418579, "reward_std": 0.27393239736557007, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.3195630609989166, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 1226.0, "completions/max_terminated_length": 1226.0, "completions/mean_length": 433.203125, "completions/mean_terminated_length": 433.203125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0525, "frac_reward_zero_std": 0.125, "grad_norm": 15.635724067687988, "learning_rate": 9.397368756032444e-07, "loss": -0.0, "num_tokens": 2444394.0, "reward": 0.4898436963558197, "reward_std": 0.4091060161590576, "rewards/format_reward/mean": 0.5234375, "rewards/format_reward/std": 0.20758795738220215, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1891.0, "completions/mean_length": 642.328125, "completions/mean_terminated_length": 596.9838256835938, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.055, "frac_reward_zero_std": 0.0, "grad_norm": 10.34483528137207, "learning_rate": 9.316282404787869e-07, "loss": -0.0, "num_tokens": 2562967.0, "reward": 0.2789062559604645, "reward_std": 0.3288741409778595, "rewards/format_reward/mean": 0.6015625, "rewards/format_reward/std": 0.26991602778434753, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 1104.0, "completions/max_terminated_length": 1104.0, "completions/mean_length": 493.53125, "completions/mean_terminated_length": 493.53125, "completions/min_length": 184.0, "completions/min_terminated_length": 184.0, "epoch": 0.0575, "frac_reward_zero_std": 0.0, "grad_norm": 1.5289628505706787, "learning_rate": 9.230476262104676e-07, "loss": 0.0, "num_tokens": 2652025.0, "reward": 0.14921875298023224, "reward_std": 0.24584797024726868, "rewards/format_reward/mean": 0.5546875, "rewards/format_reward/std": 0.29703813791275024, "rewards/mcq_exact_match_reward/mean": 0.09375, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1872.0, "completions/mean_length": 672.640625, "completions/mean_terminated_length": 650.8095703125, "completions/min_length": 151.0, "completions/min_terminated_length": 151.0, "epoch": 0.06, "frac_reward_zero_std": 0.0, "grad_norm": 0.882468581199646, "learning_rate": 9.1400441557401e-07, "loss": -0.0, "num_tokens": 2776778.0, "reward": 0.36406248807907104, "reward_std": 0.3527684509754181, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.2514837086200714, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 1541.0, "completions/mean_length": 577.0, "completions/mean_terminated_length": 553.6508178710938, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.0625, "frac_reward_zero_std": 0.0, "grad_norm": 1.0591400861740112, "learning_rate": 9.045084971874737e-07, "loss": 0.0, "num_tokens": 2904682.0, "reward": 0.5609375238418579, "reward_std": 0.42555922269821167, "rewards/format_reward/mean": 0.609375, "rewards/format_reward/std": 0.2592533528804779, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 1526.0, "completions/mean_length": 521.203125, "completions/mean_terminated_length": 496.9682922363281, "completions/min_length": 74.0, "completions/min_terminated_length": 74.0, "epoch": 0.065, "frac_reward_zero_std": 0.0, "grad_norm": 1.9715129137039185, "learning_rate": 8.945702546981968e-07, "loss": -0.0, "num_tokens": 3026623.0, "reward": 0.3656249940395355, "reward_std": 0.4152831733226776, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.3021090030670166, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 1511.0, "completions/max_terminated_length": 1511.0, "completions/mean_length": 501.578125, "completions/mean_terminated_length": 501.578125, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.0675, "frac_reward_zero_std": 0.0, "grad_norm": 1.4899768829345703, "learning_rate": 8.842005554284295e-07, "loss": 0.0, "num_tokens": 3146908.0, "reward": 0.3734374940395355, "reward_std": 0.4189872741699219, "rewards/format_reward/mean": 0.609375, "rewards/format_reward/std": 0.3503824472427368, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1947.0, "completions/mean_length": 509.5, "completions/mean_terminated_length": 459.8709411621094, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.07, "frac_reward_zero_std": 0.0, "grad_norm": 1.6197670698165894, "learning_rate": 8.734107384920769e-07, "loss": 0.0, "num_tokens": 3265052.0, "reward": 0.3656249940395355, "reward_std": 0.41435983777046204, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.40335890650749207, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 28 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 837.0, "completions/max_terminated_length": 837.0, "completions/mean_length": 351.984375, "completions/mean_terminated_length": 351.984375, "completions/min_length": 43.0, "completions/min_terminated_length": 43.0, "epoch": 0.0725, "frac_reward_zero_std": 0.125, "grad_norm": 1.2482775449752808, "learning_rate": 8.622126023955445e-07, "loss": 0.0, "num_tokens": 3350019.0, "reward": 0.46406251192092896, "reward_std": 0.38648083806037903, "rewards/format_reward/mean": 0.890625, "rewards/format_reward/std": 0.2592533528804779, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 1624.0, "completions/max_terminated_length": 1624.0, "completions/mean_length": 547.109375, "completions/mean_terminated_length": 547.109375, "completions/min_length": 111.0, "completions/min_terminated_length": 111.0, "epoch": 0.075, "frac_reward_zero_std": 0.0, "grad_norm": 1.4233943223953247, "learning_rate": 8.506183921362442e-07, "loss": -0.0, "num_tokens": 3475026.0, "reward": 0.4007812738418579, "reward_std": 0.4338732361793518, "rewards/format_reward/mean": 0.7265625, "rewards/format_reward/std": 0.3665550649166107, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1638.0, "completions/mean_length": 533.09375, "completions/mean_terminated_length": 484.2257995605469, "completions/min_length": 90.0, "completions/min_terminated_length": 90.0, "epoch": 0.0775, "frac_reward_zero_std": 0.0, "grad_norm": 1.2113726139068604, "learning_rate": 8.386407858128706e-07, "loss": 0.0, "num_tokens": 3607664.0, "reward": 0.29609373211860657, "reward_std": 0.31607964634895325, "rewards/format_reward/mean": 0.7734375, "rewards/format_reward/std": 0.3555644750595093, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 2034.0, "completions/mean_length": 706.234375, "completions/mean_terminated_length": 662.9515991210938, "completions/min_length": 246.0, "completions/min_terminated_length": 246.0, "epoch": 0.08, "frac_reward_zero_std": 0.125, "grad_norm": 0.7207805514335632, "learning_rate": 8.262928807620843e-07, "loss": 0.0, "num_tokens": 3751759.0, "reward": 0.1875, "reward_std": 0.22905191779136658, "rewards/format_reward/mean": 0.78125, "rewards/format_reward/std": 0.3766498267650604, "rewards/mcq_exact_match_reward/mean": 0.109375, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "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": 907.0, "completions/max_terminated_length": 907.0, "completions/mean_length": 378.46875, "completions/mean_terminated_length": 378.46875, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.0825, "frac_reward_zero_std": 0.0, "grad_norm": 1.2787797451019287, "learning_rate": 8.135881792367685e-07, "loss": 0.0, "num_tokens": 3881013.0, "reward": 0.3851562738418579, "reward_std": 0.3336564302444458, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.2780628502368927, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1309.0, "completions/mean_length": 543.671875, "completions/mean_terminated_length": 519.793701171875, "completions/min_length": 81.0, "completions/min_terminated_length": 81.0, "epoch": 0.085, "frac_reward_zero_std": 0.125, "grad_norm": 1.1020925045013428, "learning_rate": 8.005405736415125e-07, "loss": 0.0, "num_tokens": 4002408.0, "reward": 0.39765626192092896, "reward_std": 0.25910478830337524, "rewards/format_reward/mean": 0.8515625, "rewards/format_reward/std": 0.3294980227947235, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 1645.0, "completions/mean_length": 505.0625, "completions/mean_terminated_length": 480.5714416503906, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.0875, "frac_reward_zero_std": 0.375, "grad_norm": 0.9048623442649841, "learning_rate": 7.871643313414718e-07, "loss": -0.0, "num_tokens": 4111804.0, "reward": 0.49140623211860657, "reward_std": 0.2548467516899109, "rewards/format_reward/mean": 0.8515625, "rewards/format_reward/std": 0.3294980227947235, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1843.0, "completions/mean_length": 661.359375, "completions/mean_terminated_length": 593.1638793945312, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.09, "frac_reward_zero_std": 0.125, "grad_norm": 1.104202151298523, "learning_rate": 7.734740790612136e-07, "loss": -0.0, "num_tokens": 4241067.0, "reward": 0.2640625238418579, "reward_std": 0.29564547538757324, "rewards/format_reward/mean": 0.765625, "rewards/format_reward/std": 0.4079444408416748, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1395.0, "completions/mean_length": 492.40625, "completions/mean_terminated_length": 467.7143249511719, "completions/min_length": 175.0, "completions/min_terminated_length": 175.0, "epoch": 0.0925, "frac_reward_zero_std": 0.0, "grad_norm": 1.0682345628738403, "learning_rate": 7.594847868906076e-07, "loss": 0.0, "num_tokens": 4343157.0, "reward": 0.32109376788139343, "reward_std": 0.40080320835113525, "rewards/format_reward/mean": 0.8671875, "rewards/format_reward/std": 0.28510910272598267, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 939.0, "completions/mean_length": 399.375, "completions/mean_terminated_length": 373.20635986328125, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.095, "frac_reward_zero_std": 0.25, "grad_norm": 1.5079172849655151, "learning_rate": 7.452117519152541e-07, "loss": -0.0, "num_tokens": 4434813.0, "reward": 0.43828123807907104, "reward_std": 0.34535354375839233, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.2203386276960373, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1218.0, "completions/mean_length": 519.40625, "completions/mean_terminated_length": 470.0967712402344, "completions/min_length": 128.0, "completions/min_terminated_length": 128.0, "epoch": 0.0975, "frac_reward_zero_std": 0.0, "grad_norm": 1.2309036254882812, "learning_rate": 7.306705814893439e-07, "loss": 0.0, "num_tokens": 4561871.0, "reward": 0.45468753576278687, "reward_std": 0.38227903842926025, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.19352105259895325, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 863.0, "completions/max_terminated_length": 863.0, "completions/mean_length": 366.28125, "completions/mean_terminated_length": 366.28125, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.1, "frac_reward_zero_std": 0.0, "grad_norm": 1.4971320629119873, "learning_rate": 7.158771761692464e-07, "loss": 0.0, "num_tokens": 4653321.0, "reward": 0.6742187738418579, "reward_std": 0.46801555156707764, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 1418.0, "completions/max_terminated_length": 1418.0, "completions/mean_length": 409.921875, "completions/mean_terminated_length": 409.921875, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.1025, "frac_reward_zero_std": 0.125, "grad_norm": 1.4559797048568726, "learning_rate": 7.008477123264847e-07, "loss": 0.0, "num_tokens": 4777852.0, "reward": 0.22968751192092896, "reward_std": 0.2808600664138794, "rewards/format_reward/mean": 0.890625, "rewards/format_reward/std": 0.28824523091316223, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1941.0, "completions/mean_length": 483.9375, "completions/mean_terminated_length": 459.11114501953125, "completions/min_length": 98.0, "completions/min_terminated_length": 98.0, "epoch": 0.105, "frac_reward_zero_std": 0.25, "grad_norm": 0.9216733574867249, "learning_rate": 6.855986244591103e-07, "loss": 0.0, "num_tokens": 4895888.0, "reward": 0.2914062738418579, "reward_std": 0.29309266805648804, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.2780628502368927, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 793.0, "completions/mean_length": 399.8125, "completions/mean_terminated_length": 373.65081787109375, "completions/min_length": 99.0, "completions/min_terminated_length": 99.0, "epoch": 0.1075, "frac_reward_zero_std": 0.125, "grad_norm": 1.3962833881378174, "learning_rate": 6.701465872208216e-07, "loss": 0.0, "num_tokens": 5013452.0, "reward": 0.3382812738418579, "reward_std": 0.35107070207595825, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.31800705194473267, "rewards/mcq_exact_match_reward/mean": 0.25, "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": 1108.0, "completions/max_terminated_length": 1108.0, "completions/mean_length": 370.140625, "completions/mean_terminated_length": 370.140625, "completions/min_length": 101.0, "completions/min_terminated_length": 101.0, "epoch": 0.11, "frac_reward_zero_std": 0.375, "grad_norm": 1.1737865209579468, "learning_rate": 6.545084971874736e-07, "loss": 0.0, "num_tokens": 5107461.0, "reward": 0.41093751788139343, "reward_std": 0.3137226104736328, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 982.0, "completions/mean_length": 446.234375, "completions/mean_terminated_length": 394.56451416015625, "completions/min_length": 67.0, "completions/min_terminated_length": 67.0, "epoch": 0.1125, "frac_reward_zero_std": 0.375, "grad_norm": 1.3070462942123413, "learning_rate": 6.387014543809223e-07, "loss": -0.0, "num_tokens": 5219996.0, "reward": 0.41796875, "reward_std": 0.24860012531280518, "rewards/format_reward/mean": 0.8984375, "rewards/format_reward/std": 0.2978719174861908, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 923.0, "completions/mean_length": 458.046875, "completions/mean_terminated_length": 432.8095397949219, "completions/min_length": 57.0, "completions/min_terminated_length": 57.0, "epoch": 0.115, "frac_reward_zero_std": 0.125, "grad_norm": 1.233520269393921, "learning_rate": 6.227427435703995e-07, "loss": 0.0, "num_tokens": 5319583.0, "reward": 0.632031261920929, "reward_std": 0.3984364867210388, "rewards/format_reward/mean": 0.8515625, "rewards/format_reward/std": 0.3527633249759674, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 769.0, "completions/mean_length": 397.125, "completions/mean_terminated_length": 343.8709716796875, "completions/min_length": 107.0, "completions/min_terminated_length": 107.0, "epoch": 0.1175, "frac_reward_zero_std": 0.125, "grad_norm": 1.7266496419906616, "learning_rate": 6.066498153718734e-07, "loss": 0.0, "num_tokens": 5400327.0, "reward": 0.4945312738418579, "reward_std": 0.39649826288223267, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.3052735924720764, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1686.0, "completions/mean_length": 578.578125, "completions/mean_terminated_length": 555.2540283203125, "completions/min_length": 190.0, "completions/min_terminated_length": 190.0, "epoch": 0.12, "frac_reward_zero_std": 0.0, "grad_norm": 1.098490595817566, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 5519596.0, "reward": 0.4632812738418579, "reward_std": 0.4184955060482025, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.31800705194473267, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 1035.0, "completions/max_terminated_length": 1035.0, "completions/mean_length": 403.234375, "completions/mean_terminated_length": 403.234375, "completions/min_length": 110.0, "completions/min_terminated_length": 110.0, "epoch": 0.1225, "frac_reward_zero_std": 0.125, "grad_norm": 1.0547609329223633, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 5616747.0, "reward": 0.4046874940395355, "reward_std": 0.35222142934799194, "rewards/format_reward/mean": 0.921875, "rewards/format_reward/std": 0.27048972249031067, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1334.0, "completions/mean_length": 430.5625, "completions/mean_terminated_length": 404.888916015625, "completions/min_length": 115.0, "completions/min_terminated_length": 115.0, "epoch": 0.125, "frac_reward_zero_std": 0.125, "grad_norm": 1.1792281866073608, "learning_rate": 5.577423184847931e-07, "loss": 0.0, "num_tokens": 5709775.0, "reward": 0.35859376192092896, "reward_std": 0.33752134442329407, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.2498759627342224, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 50 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1015.0, "completions/mean_length": 409.4375, "completions/mean_terminated_length": 383.4285888671875, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.1275, "frac_reward_zero_std": 0.0, "grad_norm": 1.5146547555923462, "learning_rate": 5.412896727361662e-07, "loss": 0.0, "num_tokens": 5814739.0, "reward": 0.4320312738418579, "reward_std": 0.47502028942108154, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.3052735924720764, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 51 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1956.0, "completions/mean_length": 530.21875, "completions/mean_terminated_length": 506.12701416015625, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.13, "frac_reward_zero_std": 0.25, "grad_norm": 1.0168224573135376, "learning_rate": 5.247918773366111e-07, "loss": -0.0, "num_tokens": 5941265.0, "reward": 0.34375, "reward_std": 0.3308578133583069, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.208927720785141, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 52 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1120.0, "completions/mean_length": 482.140625, "completions/mean_terminated_length": 431.6290283203125, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.1325, "frac_reward_zero_std": 0.125, "grad_norm": 1.017253041267395, "learning_rate": 5.082669723831793e-07, "loss": 0.0, "num_tokens": 6096258.0, "reward": 0.3335937559604645, "reward_std": 0.24125297367572784, "rewards/format_reward/mean": 0.8359375, "rewards/format_reward/std": 0.3456611931324005, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 53 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 872.0, "completions/mean_length": 443.46875, "completions/mean_terminated_length": 391.70965576171875, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.135, "frac_reward_zero_std": 0.0, "grad_norm": 1.3861647844314575, "learning_rate": 4.917330276168208e-07, "loss": 0.0, "num_tokens": 6208176.0, "reward": 0.3984375, "reward_std": 0.37967199087142944, "rewards/format_reward/mean": 0.859375, "rewards/format_reward/std": 0.3503824472427368, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 54 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1076.0, "completions/max_terminated_length": 1076.0, "completions/mean_length": 406.546875, "completions/mean_terminated_length": 406.546875, "completions/min_length": 76.0, "completions/min_terminated_length": 76.0, "epoch": 0.1375, "frac_reward_zero_std": 0.125, "grad_norm": 1.5673335790634155, "learning_rate": 4.752081226633888e-07, "loss": 0.0, "num_tokens": 6305963.0, "reward": 0.37812501192092896, "reward_std": 0.40637484192848206, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.1510545015335083, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 55 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 908.0, "completions/max_terminated_length": 908.0, "completions/mean_length": 389.671875, "completions/mean_terminated_length": 389.671875, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.14, "frac_reward_zero_std": 0.375, "grad_norm": 1.057645559310913, "learning_rate": 4.5871032726383385e-07, "loss": 0.0, "num_tokens": 6389150.0, "reward": 0.23281249403953552, "reward_std": 0.27970460057258606, "rewards/format_reward/mean": 0.921875, "rewards/format_reward/std": 0.27048972249031067, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "step": 56 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1126.0, "completions/mean_length": 476.109375, "completions/mean_terminated_length": 451.15875244140625, "completions/min_length": 94.0, "completions/min_terminated_length": 94.0, "epoch": 0.1425, "frac_reward_zero_std": 0.0, "grad_norm": 1.2653309106826782, "learning_rate": 4.4225768151520694e-07, "loss": 0.0, "num_tokens": 6514733.0, "reward": 0.4164062738418579, "reward_std": 0.4222263693809509, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.31800705194473267, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 57 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1165.0, "completions/mean_length": 407.34375, "completions/mean_terminated_length": 354.4193420410156, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.145, "frac_reward_zero_std": 0.25, "grad_norm": 1.414528489112854, "learning_rate": 4.258681761440789e-07, "loss": -0.0, "num_tokens": 6610027.0, "reward": 0.4554687738418579, "reward_std": 0.3235210180282593, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 58 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1211.0, "completions/mean_length": 519.8125, "completions/mean_terminated_length": 444.65570068359375, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.1475, "frac_reward_zero_std": 0.25, "grad_norm": 1.1935831308364868, "learning_rate": 4.095597328339452e-07, "loss": 0.0, "num_tokens": 6727839.0, "reward": 0.22812500596046448, "reward_std": 0.24529126286506653, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.32120805978775024, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "step": 59 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1338.0, "completions/max_terminated_length": 1338.0, "completions/mean_length": 374.453125, "completions/mean_terminated_length": 374.453125, "completions/min_length": 102.0, "completions/min_terminated_length": 102.0, "epoch": 0.15, "frac_reward_zero_std": 0.25, "grad_norm": 1.3652595281600952, "learning_rate": 3.9335018462812664e-07, "loss": 0.0, "num_tokens": 6837228.0, "reward": 0.3671875, "reward_std": 0.35627028346061707, "rewards/format_reward/mean": 0.859375, "rewards/format_reward/std": 0.3503824472427368, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 60 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 777.0, "completions/mean_length": 382.96875, "completions/mean_terminated_length": 356.5397033691406, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.1525, "frac_reward_zero_std": 0.0, "grad_norm": 1.9001901149749756, "learning_rate": 3.772572564296004e-07, "loss": 0.0, "num_tokens": 6932578.0, "reward": 0.4164062738418579, "reward_std": 0.46945908665657043, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.31800705194473267, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 61 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.078125, "completions/max_length": 2048.0, "completions/max_terminated_length": 890.0, "completions/mean_length": 551.984375, "completions/mean_terminated_length": 425.2033996582031, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.155, "frac_reward_zero_std": 0.125, "grad_norm": 1.1120458841323853, "learning_rate": 3.612985456190778e-07, "loss": 0.0, "num_tokens": 7073729.0, "reward": 0.40234375, "reward_std": 0.3692801594734192, "rewards/format_reward/mean": 0.8984375, "rewards/format_reward/std": 0.2978719174861908, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 62 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1396.0, "completions/max_terminated_length": 1396.0, "completions/mean_length": 403.828125, "completions/mean_terminated_length": 403.828125, "completions/min_length": 131.0, "completions/min_terminated_length": 131.0, "epoch": 0.1575, "frac_reward_zero_std": 0.0, "grad_norm": 1.1444545984268188, "learning_rate": 3.454915028125263e-07, "loss": 0.0, "num_tokens": 7171030.0, "reward": 0.3718750476837158, "reward_std": 0.35097694396972656, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.29378482699394226, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 63 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1412.0, "completions/mean_length": 450.859375, "completions/mean_terminated_length": 399.33868408203125, "completions/min_length": 129.0, "completions/min_terminated_length": 129.0, "epoch": 0.16, "frac_reward_zero_std": 0.125, "grad_norm": 1.259133219718933, "learning_rate": 3.2985341277917846e-07, "loss": 0.0, "num_tokens": 7279477.0, "reward": 0.49921876192092896, "reward_std": 0.28595492243766785, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.2498759627342224, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 64 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1031.0, "completions/mean_length": 521.015625, "completions/mean_terminated_length": 496.7778015136719, "completions/min_length": 161.0, "completions/min_terminated_length": 161.0, "epoch": 0.1625, "frac_reward_zero_std": 0.125, "grad_norm": 1.2433366775512695, "learning_rate": 3.1440137554088953e-07, "loss": 0.0, "num_tokens": 7420022.0, "reward": 0.37812501192092896, "reward_std": 0.36857056617736816, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.35073620080947876, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 65 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1639.0, "completions/max_terminated_length": 1639.0, "completions/mean_length": 393.953125, "completions/mean_terminated_length": 393.953125, "completions/min_length": 120.0, "completions/min_terminated_length": 120.0, "epoch": 0.165, "frac_reward_zero_std": 0.125, "grad_norm": 1.4713146686553955, "learning_rate": 2.9915228767351535e-07, "loss": 0.0, "num_tokens": 7518627.0, "reward": 0.40703126788139343, "reward_std": 0.29805976152420044, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.180765300989151, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 66 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1069.0, "completions/mean_length": 388.609375, "completions/mean_terminated_length": 362.2698669433594, "completions/min_length": 60.0, "completions/min_terminated_length": 60.0, "epoch": 0.1675, "frac_reward_zero_std": 0.125, "grad_norm": 1.5749222040176392, "learning_rate": 2.841228238307536e-07, "loss": 0.0, "num_tokens": 7619010.0, "reward": 0.39140626788139343, "reward_std": 0.38446736335754395, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.2203386276960373, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 67 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1429.0, "completions/mean_length": 406.375, "completions/mean_terminated_length": 380.3174743652344, "completions/min_length": 108.0, "completions/min_terminated_length": 108.0, "epoch": 0.17, "frac_reward_zero_std": 0.25, "grad_norm": 1.0842872858047485, "learning_rate": 2.6932941851065615e-07, "loss": 0.0, "num_tokens": 7709650.0, "reward": 0.26250001788139343, "reward_std": 0.2868100106716156, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.279951810836792, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 68 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1046.0, "completions/max_terminated_length": 1046.0, "completions/mean_length": 336.171875, "completions/mean_terminated_length": 336.171875, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.1725, "frac_reward_zero_std": 0.125, "grad_norm": 1.5573029518127441, "learning_rate": 2.547882480847461e-07, "loss": 0.0, "num_tokens": 7810701.0, "reward": 0.47265625, "reward_std": 0.2891082167625427, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 69 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 968.0, "completions/max_terminated_length": 968.0, "completions/mean_length": 366.96875, "completions/mean_terminated_length": 366.96875, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.175, "frac_reward_zero_std": 0.375, "grad_norm": 1.144339680671692, "learning_rate": 2.4051521310939254e-07, "loss": 0.0, "num_tokens": 7911211.0, "reward": 0.49921876192092896, "reward_std": 0.29788392782211304, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.23345555365085602, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 70 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1005.0, "completions/mean_length": 458.28125, "completions/mean_terminated_length": 407.0, "completions/min_length": 91.0, "completions/min_terminated_length": 91.0, "epoch": 0.1775, "frac_reward_zero_std": 0.0, "grad_norm": 1.3294062614440918, "learning_rate": 2.2652592093878665e-07, "loss": 0.0, "num_tokens": 8019733.0, "reward": 0.28046876192092896, "reward_std": 0.3795745074748993, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.2498759627342224, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 71 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1555.0, "completions/mean_length": 532.609375, "completions/mean_terminated_length": 458.0819396972656, "completions/min_length": 88.0, "completions/min_terminated_length": 88.0, "epoch": 0.18, "frac_reward_zero_std": 0.125, "grad_norm": 1.2976422309875488, "learning_rate": 2.128356686585282e-07, "loss": 0.0, "num_tokens": 8128588.0, "reward": 0.581250011920929, "reward_std": 0.3822355568408966, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.3726780116558075, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 72 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1071.0, "completions/max_terminated_length": 1071.0, "completions/mean_length": 397.140625, "completions/mean_terminated_length": 397.140625, "completions/min_length": 116.0, "completions/min_terminated_length": 116.0, "epoch": 0.1825, "frac_reward_zero_std": 0.25, "grad_norm": 1.5232187509536743, "learning_rate": 1.9945942635848745e-07, "loss": 0.0, "num_tokens": 8236637.0, "reward": 0.6156250238418579, "reward_std": 0.35612428188323975, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 73 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1494.0, "completions/mean_length": 457.828125, "completions/mean_terminated_length": 432.58734130859375, "completions/min_length": 73.0, "completions/min_terminated_length": 73.0, "epoch": 0.185, "frac_reward_zero_std": 0.0, "grad_norm": 1.1997640132904053, "learning_rate": 1.8641182076323148e-07, "loss": 0.0, "num_tokens": 8353210.0, "reward": 0.36796876788139343, "reward_std": 0.3068220019340515, "rewards/format_reward/mean": 0.8671875, "rewards/format_reward/std": 0.28510910272598267, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 74 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1080.0, "completions/mean_length": 390.953125, "completions/mean_terminated_length": 337.5, "completions/min_length": 110.0, "completions/min_terminated_length": 110.0, "epoch": 0.1875, "frac_reward_zero_std": 0.125, "grad_norm": 1.7119704484939575, "learning_rate": 1.7370711923791564e-07, "loss": 0.0, "num_tokens": 8456735.0, "reward": 0.503125011920929, "reward_std": 0.37247955799102783, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 75 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.078125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1152.0, "completions/mean_length": 513.03125, "completions/mean_terminated_length": 382.94915771484375, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.19, "frac_reward_zero_std": 0.25, "grad_norm": 1.3417837619781494, "learning_rate": 1.6135921418712955e-07, "loss": 0.0, "num_tokens": 8562321.0, "reward": 0.3421875238418579, "reward_std": 0.33041971921920776, "rewards/format_reward/mean": 0.765625, "rewards/format_reward/std": 0.42695629596710205, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 76 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 938.0, "completions/mean_length": 394.890625, "completions/mean_terminated_length": 368.65081787109375, "completions/min_length": 86.0, "completions/min_terminated_length": 86.0, "epoch": 0.1925, "frac_reward_zero_std": 0.25, "grad_norm": 1.6523009538650513, "learning_rate": 1.493816078637557e-07, "loss": 0.0, "num_tokens": 8678778.0, "reward": 0.30937501788139343, "reward_std": 0.3358280658721924, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.29378482699394226, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 77 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1095.0, "completions/max_terminated_length": 1095.0, "completions/mean_length": 352.34375, "completions/mean_terminated_length": 352.34375, "completions/min_length": 103.0, "completions/min_terminated_length": 103.0, "epoch": 0.195, "frac_reward_zero_std": 0.125, "grad_norm": 1.6738253831863403, "learning_rate": 1.3778739760445552e-07, "loss": 0.0, "num_tokens": 8779408.0, "reward": 0.4554687738418579, "reward_std": 0.27686768770217896, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 78 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1308.0, "completions/mean_length": 468.171875, "completions/mean_terminated_length": 362.8500061035156, "completions/min_length": 105.0, "completions/min_terminated_length": 105.0, "epoch": 0.1975, "frac_reward_zero_std": 0.125, "grad_norm": 1.4916421175003052, "learning_rate": 1.2658926150792322e-07, "loss": 0.0, "num_tokens": 8876827.0, "reward": 0.37031251192092896, "reward_std": 0.33942168951034546, "rewards/format_reward/mean": 0.890625, "rewards/format_reward/std": 0.3145764470100403, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 79 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 658.0, "completions/max_terminated_length": 658.0, "completions/mean_length": 309.53125, "completions/mean_terminated_length": 309.53125, "completions/min_length": 58.0, "completions/min_terminated_length": 58.0, "epoch": 0.2, "frac_reward_zero_std": 0.125, "grad_norm": 1.4920448064804077, "learning_rate": 1.1579944457157059e-07, "loss": 0.0, "num_tokens": 8959341.0, "reward": 0.596875011920929, "reward_std": 0.39014071226119995, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 80 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1603.0, "completions/mean_length": 459.03125, "completions/mean_terminated_length": 433.8095397949219, "completions/min_length": 85.0, "completions/min_terminated_length": 85.0, "epoch": 0.2025, "frac_reward_zero_std": 0.0, "grad_norm": 1.7325104475021362, "learning_rate": 1.0542974530180327e-07, "loss": 0.0, "num_tokens": 9088631.0, "reward": 0.4375000298023224, "reward_std": 0.4429008364677429, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 81 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1075.0, "completions/mean_length": 401.578125, "completions/mean_terminated_length": 375.4444580078125, "completions/min_length": 84.0, "completions/min_terminated_length": 84.0, "epoch": 0.205, "frac_reward_zero_std": 0.25, "grad_norm": 1.112608551979065, "learning_rate": 9.549150281252632e-08, "loss": 0.0, "num_tokens": 9194820.0, "reward": 0.25859373807907104, "reward_std": 0.17519733309745789, "rewards/format_reward/mean": 0.8671875, "rewards/format_reward/std": 0.3362041711807251, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 82 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1350.0, "completions/mean_length": 444.1875, "completions/mean_terminated_length": 365.31146240234375, "completions/min_length": 61.0, "completions/min_terminated_length": 61.0, "epoch": 0.2075, "frac_reward_zero_std": 0.0, "grad_norm": 1.3191931247711182, "learning_rate": 8.599558442598998e-08, "loss": 0.0, "num_tokens": 9312160.0, "reward": 0.4703125059604645, "reward_std": 0.34370893239974976, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.21304203569889069, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 83 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1043.0, "completions/mean_length": 447.96875, "completions/mean_terminated_length": 422.5714416503906, "completions/min_length": 113.0, "completions/min_terminated_length": 113.0, "epoch": 0.21, "frac_reward_zero_std": 0.125, "grad_norm": 0.9768643379211426, "learning_rate": 7.695237378953224e-08, "loss": 0.0, "num_tokens": 9423734.0, "reward": 0.2789062559604645, "reward_std": 0.3008596897125244, "rewards/format_reward/mean": 0.9140625, "rewards/format_reward/std": 0.27537402510643005, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 84 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 898.0, "completions/mean_length": 410.171875, "completions/mean_terminated_length": 357.33868408203125, "completions/min_length": 72.0, "completions/min_terminated_length": 72.0, "epoch": 0.2125, "frac_reward_zero_std": 0.125, "grad_norm": 1.3862519264221191, "learning_rate": 6.837175952121304e-08, "loss": 0.0, "num_tokens": 9522825.0, "reward": 0.503125011920929, "reward_std": 0.4141109585762024, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 85 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1155.0, "completions/mean_length": 425.390625, "completions/mean_terminated_length": 399.63494873046875, "completions/min_length": 80.0, "completions/min_terminated_length": 80.0, "epoch": 0.215, "frac_reward_zero_std": 0.125, "grad_norm": 1.3413077592849731, "learning_rate": 6.026312439675551e-08, "loss": -0.0, "num_tokens": 9631802.0, "reward": 0.42500001192092896, "reward_std": 0.3962019681930542, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.1510545015335083, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 86 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1212.0, "completions/mean_length": 380.421875, "completions/mean_terminated_length": 326.6290283203125, "completions/min_length": 54.0, "completions/min_terminated_length": 54.0, "epoch": 0.2175, "frac_reward_zero_std": 0.25, "grad_norm": 1.3858424425125122, "learning_rate": 5.263533508961826e-08, "loss": 0.0, "num_tokens": 9738549.0, "reward": 0.5765625238418579, "reward_std": 0.3437184989452362, "rewards/format_reward/mean": 0.921875, "rewards/format_reward/std": 0.27048972249031067, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 87 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1557.0, "completions/mean_length": 497.21875, "completions/mean_terminated_length": 472.60321044921875, "completions/min_length": 92.0, "completions/min_terminated_length": 92.0, "epoch": 0.22, "frac_reward_zero_std": 0.0, "grad_norm": 1.4040886163711548, "learning_rate": 4.549673247541874e-08, "loss": -0.0, "num_tokens": 9857763.0, "reward": 0.6234375238418579, "reward_std": 0.4146481156349182, "rewards/format_reward/mean": 0.921875, "rewards/format_reward/std": 0.25539806485176086, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 88 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 920.0, "completions/mean_length": 504.5, "completions/mean_terminated_length": 428.59014892578125, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.2225, "frac_reward_zero_std": 0.125, "grad_norm": 1.140403151512146, "learning_rate": 3.8855122511307626e-08, "loss": 0.0, "num_tokens": 9988363.0, "reward": 0.35078126192092896, "reward_std": 0.2528582215309143, "rewards/format_reward/mean": 0.8515625, "rewards/format_reward/std": 0.3527633249759674, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 89 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 841.0, "completions/max_terminated_length": 841.0, "completions/mean_length": 344.65625, "completions/mean_terminated_length": 344.65625, "completions/min_length": 89.0, "completions/min_terminated_length": 89.0, "epoch": 0.225, "frac_reward_zero_std": 0.25, "grad_norm": 1.1781779527664185, "learning_rate": 3.271776770026963e-08, "loss": 0.0, "num_tokens": 10073501.0, "reward": 0.5984375476837158, "reward_std": 0.3432989716529846, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 90 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 865.0, "completions/mean_length": 394.453125, "completions/mean_terminated_length": 368.20635986328125, "completions/min_length": 95.0, "completions/min_terminated_length": 95.0, "epoch": 0.2275, "frac_reward_zero_std": 0.125, "grad_norm": 1.1496305465698242, "learning_rate": 2.7091379149682682e-08, "loss": 0.0, "num_tokens": 10182890.0, "reward": 0.44843751192092896, "reward_std": 0.27532410621643066, "rewards/format_reward/mean": 0.734375, "rewards/format_reward/std": 0.42695629596710205, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 91 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 914.0, "completions/mean_length": 315.421875, "completions/mean_terminated_length": 287.920654296875, "completions/min_length": 87.0, "completions/min_terminated_length": 87.0, "epoch": 0.23, "frac_reward_zero_std": 0.125, "grad_norm": 1.5781230926513672, "learning_rate": 2.1982109232821176e-08, "loss": -0.0, "num_tokens": 10276637.0, "reward": 0.532031238079071, "reward_std": 0.3431118428707123, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.2203386276960373, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "step": 92 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1765.0, "completions/max_terminated_length": 1765.0, "completions/mean_length": 362.609375, "completions/mean_terminated_length": 362.609375, "completions/min_length": 127.0, "completions/min_terminated_length": 127.0, "epoch": 0.2325, "frac_reward_zero_std": 0.375, "grad_norm": 0.9790539145469666, "learning_rate": 1.7395544861325718e-08, "loss": 0.0, "num_tokens": 10387332.0, "reward": 0.53125, "reward_std": 0.2589051425457001, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "step": 93 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1120.0, "completions/max_terminated_length": 1120.0, "completions/mean_length": 357.734375, "completions/mean_terminated_length": 357.734375, "completions/min_length": 22.0, "completions/min_terminated_length": 22.0, "epoch": 0.235, "frac_reward_zero_std": 0.5, "grad_norm": 0.9088587760925293, "learning_rate": 1.3336701375997127e-08, "loss": 0.0, "num_tokens": 10484267.0, "reward": 0.5515625476837158, "reward_std": 0.21040581166744232, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 94 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1099.0, "completions/mean_length": 418.96875, "completions/mean_terminated_length": 366.4193420410156, "completions/min_length": 79.0, "completions/min_terminated_length": 79.0, "epoch": 0.2375, "frac_reward_zero_std": 0.0, "grad_norm": 2.3513271808624268, "learning_rate": 9.810017062595321e-09, "loss": 0.0, "num_tokens": 10602945.0, "reward": 0.3890625238418579, "reward_std": 0.4030550718307495, "rewards/format_reward/mean": 0.921875, "rewards/format_reward/std": 0.25539806485176086, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 95 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1008.0, "completions/mean_length": 523.703125, "completions/mean_terminated_length": 474.5322570800781, "completions/min_length": 101.0, "completions/min_terminated_length": 101.0, "epoch": 0.24, "frac_reward_zero_std": 0.0, "grad_norm": 1.3562307357788086, "learning_rate": 6.819348298638839e-09, "loss": 0.0, "num_tokens": 10752654.0, "reward": 0.43281251192092896, "reward_std": 0.383518785238266, "rewards/format_reward/mean": 0.890625, "rewards/format_reward/std": 0.28824523091316223, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 96 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1464.0, "completions/mean_length": 340.453125, "completions/mean_terminated_length": 313.3492126464844, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.2425, "frac_reward_zero_std": 0.25, "grad_norm": 1.4542995691299438, "learning_rate": 4.367965336512403e-09, "loss": 0.0, "num_tokens": 10840315.0, "reward": 0.45468753576278687, "reward_std": 0.33074572682380676, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.21304203569889069, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 97 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1283.0, "completions/mean_length": 433.484375, "completions/mean_terminated_length": 381.4032287597656, "completions/min_length": 112.0, "completions/min_terminated_length": 112.0, "epoch": 0.245, "frac_reward_zero_std": 0.0, "grad_norm": 1.5124303102493286, "learning_rate": 2.458548727494292e-09, "loss": 0.0, "num_tokens": 10940346.0, "reward": 0.4710937738418579, "reward_std": 0.44797325134277344, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 98 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 824.0, "completions/mean_length": 438.203125, "completions/mean_terminated_length": 386.274169921875, "completions/min_length": 117.0, "completions/min_terminated_length": 117.0, "epoch": 0.2475, "frac_reward_zero_std": 0.0, "grad_norm": 1.436461091041565, "learning_rate": 1.0931863906127325e-09, "loss": -0.0, "num_tokens": 11051959.0, "reward": 0.42890626192092896, "reward_std": 0.46917158365249634, "rewards/format_reward/mean": 0.8515625, "rewards/format_reward/std": 0.3527633249759674, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 99 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1421.0, "completions/max_terminated_length": 1421.0, "completions/mean_length": 417.8125, "completions/mean_terminated_length": 417.8125, "completions/min_length": 83.0, "completions/min_terminated_length": 83.0, "epoch": 0.25, "frac_reward_zero_std": 0.125, "grad_norm": 1.1657195091247559, "learning_rate": 2.733713295369755e-10, "loss": 0.0, "num_tokens": 11174979.0, "reward": 0.37812501192092896, "reward_std": 0.3555063009262085, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 100 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 11174979, "num_train_epochs": 1, "save_steps": 50, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }