{ "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1533.0, "completions/mean_length": 530.640625, "completions/mean_terminated_length": 456.016357421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0025, "frac_reward_zero_std": 0.0, "grad_norm": 3.378190755844116, "learning_rate": 0.0, "loss": 0.0, "num_tokens": 148889.0, "reward": 0.24296872317790985, "reward_std": 0.3012738525867462, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.22146137058734894, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 1344.0, "completions/max_terminated_length": 1344.0, "completions/mean_length": 366.0, "completions/mean_terminated_length": 366.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.005, "frac_reward_zero_std": 0.0, "grad_norm": 13.094253540039062, "learning_rate": 2e-07, "loss": -0.0, "num_tokens": 251593.0, "reward": 0.31328123807907104, "reward_std": 0.3454205095767975, "rewards/format_reward/mean": 0.3203125, "rewards/format_reward/std": 0.2576941251754761, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1649.0, "completions/mean_length": 505.78125, "completions/mean_terminated_length": 481.3016052246094, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0075, "frac_reward_zero_std": 0.0, "grad_norm": 11.198899269104004, "learning_rate": 4e-07, "loss": 0.0, "num_tokens": 390731.0, "reward": 0.1953124850988388, "reward_std": 0.2855938971042633, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.2741328477859497, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 1555.0, "completions/mean_length": 526.09375, "completions/mean_terminated_length": 501.9365234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.01, "frac_reward_zero_std": 0.125, "grad_norm": 3.1906931400299072, "learning_rate": 6e-07, "loss": -0.0, "num_tokens": 520417.0, "reward": 0.2984374761581421, "reward_std": 0.3686792850494385, "rewards/format_reward/mean": 0.328125, "rewards/format_reward/std": 0.25539806485176086, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1530.0, "completions/mean_length": 595.140625, "completions/mean_terminated_length": 523.6884765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0125, "frac_reward_zero_std": 0.125, "grad_norm": 3.78658127784729, "learning_rate": 8e-07, "loss": -0.0, "num_tokens": 663626.0, "reward": 0.16015625, "reward_std": 0.26552730798721313, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.3044600784778595, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 1356.0, "completions/max_terminated_length": 1356.0, "completions/mean_length": 512.671875, "completions/mean_terminated_length": 512.671875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.015, "frac_reward_zero_std": 0.0, "grad_norm": 6.359753131866455, "learning_rate": 1e-06, "loss": -0.0, "num_tokens": 777077.0, "reward": 0.3164062201976776, "reward_std": 0.40065521001815796, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.26246222853660583, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 6 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1754.0, "completions/max_terminated_length": 1754.0, "completions/mean_length": 577.96875, "completions/mean_terminated_length": 577.96875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0175, "frac_reward_zero_std": 0.0, "grad_norm": 7.725679874420166, "learning_rate": 9.99726628670463e-07, "loss": -0.0, "num_tokens": 896811.0, "reward": 0.36796873807907104, "reward_std": 0.42190927267074585, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.2387082874774933, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 1189.0, "completions/mean_length": 528.03125, "completions/mean_terminated_length": 503.90478515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.02, "frac_reward_zero_std": 0.125, "grad_norm": 3.6088428497314453, "learning_rate": 9.989068136093872e-07, "loss": 0.0, "num_tokens": 1030845.0, "reward": 0.16953124105930328, "reward_std": 0.2975313067436218, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.2203386276960373, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 0.125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1892.0, "completions/mean_length": 625.015625, "completions/mean_terminated_length": 421.7321472167969, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0225, "frac_reward_zero_std": 0.0, "grad_norm": 9.677778244018555, "learning_rate": 9.975414512725056e-07, "loss": 0.0, "num_tokens": 1163238.0, "reward": 0.10234375298023224, "reward_std": 0.18847742676734924, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.3354656398296356, "rewards/mcq_exact_match_reward/mean": 0.0625, "rewards/mcq_exact_match_reward/std": 0.24397502839565277, "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": 1197.0, "completions/max_terminated_length": 1197.0, "completions/mean_length": 298.65625, "completions/mean_terminated_length": 298.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.025, "frac_reward_zero_std": 0.125, "grad_norm": 12.062150955200195, "learning_rate": 9.956320346634875e-07, "loss": -0.0, "num_tokens": 1268568.0, "reward": 0.28046876192092896, "reward_std": 0.3143594264984131, "rewards/format_reward/mean": 0.3046875, "rewards/format_reward/std": 0.2615155577659607, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 1770.0, "completions/mean_length": 403.125, "completions/mean_terminated_length": 377.0158996582031, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0275, "frac_reward_zero_std": 0.0, "grad_norm": 16.18817710876465, "learning_rate": 9.931806517013612e-07, "loss": 0.0, "num_tokens": 1409344.0, "reward": 0.5226562023162842, "reward_std": 0.33827945590019226, "rewards/format_reward/mean": 0.3828125, "rewards/format_reward/std": 0.21347814798355103, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1433.0, "completions/mean_length": 303.390625, "completions/mean_terminated_length": 275.69842529296875, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.03, "frac_reward_zero_std": 0.0, "grad_norm": 19.480257034301758, "learning_rate": 9.901899829374047e-07, "loss": 0.0, "num_tokens": 1504889.0, "reward": 0.32109373807907104, "reward_std": 0.4039357900619507, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.28423789143562317, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 0.0, "completions/max_length": 1851.0, "completions/max_terminated_length": 1851.0, "completions/mean_length": 443.6875, "completions/mean_terminated_length": 443.6875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0325, "frac_reward_zero_std": 0.0, "grad_norm": 7.029141902923584, "learning_rate": 9.866632986240029e-07, "loss": 0.0, "num_tokens": 1600813.0, "reward": 0.28046873211860657, "reward_std": 0.3932330906391144, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.28597769141197205, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 1590.0, "completions/mean_length": 575.71875, "completions/mean_terminated_length": 552.3492431640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.035, "frac_reward_zero_std": 0.125, "grad_norm": 11.138574600219727, "learning_rate": 9.826044551386742e-07, "loss": 0.0, "num_tokens": 1771547.0, "reward": 0.27812498807907104, "reward_std": 0.2972943186759949, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 1340.0, "completions/max_terminated_length": 1340.0, "completions/mean_length": 308.53125, "completions/mean_terminated_length": 308.53125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0375, "frac_reward_zero_std": 0.0, "grad_norm": 7.62338399887085, "learning_rate": 9.780178907671788e-07, "loss": 0.0, "num_tokens": 1872757.0, "reward": 0.2789062261581421, "reward_std": 0.2728224992752075, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.20152568817138672, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 1745.0, "completions/max_terminated_length": 1745.0, "completions/mean_length": 413.875, "completions/mean_terminated_length": 413.875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04, "frac_reward_zero_std": 0.375, "grad_norm": 3.428014039993286, "learning_rate": 9.729086208503173e-07, "loss": -0.0, "num_tokens": 1976933.0, "reward": 0.5, "reward_std": 0.29212942719459534, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 16 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.078125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1455.0, "completions/mean_length": 481.984375, "completions/mean_terminated_length": 349.27117919921875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0425, "frac_reward_zero_std": 0.25, "grad_norm": 12.244830131530762, "learning_rate": 9.672822322997304e-07, "loss": -0.0, "num_tokens": 2126268.0, "reward": 0.2632812261581421, "reward_std": 0.2631150186061859, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.180765300989151, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 1530.0, "completions/max_terminated_length": 1530.0, "completions/mean_length": 378.28125, "completions/mean_terminated_length": 378.28125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.045, "frac_reward_zero_std": 0.125, "grad_norm": 6.67620849609375, "learning_rate": 9.611448774886923e-07, "loss": -0.0, "num_tokens": 2266390.0, "reward": 0.3085937201976776, "reward_std": 0.3328220248222351, "rewards/format_reward/mean": 0.4296875, "rewards/format_reward/std": 0.1751912236213684, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 1118.0, "completions/max_terminated_length": 1118.0, "completions/mean_length": 203.015625, "completions/mean_terminated_length": 203.015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0475, "frac_reward_zero_std": 0.5, "grad_norm": 11.600245475769043, "learning_rate": 9.545032675245813e-07, "loss": 0.0, "num_tokens": 2369271.0, "reward": 0.45625001192092896, "reward_std": 0.09559644758701324, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.19920477271080017, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 1107.0, "completions/max_terminated_length": 1107.0, "completions/mean_length": 342.953125, "completions/mean_terminated_length": 342.953125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05, "frac_reward_zero_std": 0.0, "grad_norm": 8.118581771850586, "learning_rate": 9.473646649103817e-07, "loss": -0.0, "num_tokens": 2482204.0, "reward": 0.5328124761581421, "reward_std": 0.44431811571121216, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.1985812783241272, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 20 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 2038.0, "completions/mean_length": 413.25, "completions/mean_terminated_length": 387.3016052246094, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0525, "frac_reward_zero_std": 0.125, "grad_norm": 12.09787654876709, "learning_rate": 9.397368756032444e-07, "loss": -0.0, "num_tokens": 2615236.0, "reward": 0.22499999403953552, "reward_std": 0.2952703833580017, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.233588308095932, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 21 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1083.0, "completions/max_terminated_length": 1083.0, "completions/mean_length": 229.15625, "completions/mean_terminated_length": 229.15625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.055, "frac_reward_zero_std": 0.5, "grad_norm": 10.867944717407227, "learning_rate": 9.316282404787869e-07, "loss": -0.0, "num_tokens": 2715126.0, "reward": 0.5328124761581421, "reward_std": 0.21166591346263885, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1308.0, "completions/max_terminated_length": 1308.0, "completions/mean_length": 309.8125, "completions/mean_terminated_length": 309.8125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0575, "frac_reward_zero_std": 0.25, "grad_norm": 9.249100685119629, "learning_rate": 9.230476262104676e-07, "loss": 0.0, "num_tokens": 2827538.0, "reward": 0.4976562261581421, "reward_std": 0.2265644371509552, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.180765300989151, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 1157.0, "completions/max_terminated_length": 1157.0, "completions/mean_length": 278.390625, "completions/mean_terminated_length": 278.390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06, "frac_reward_zero_std": 0.25, "grad_norm": 5.3155436515808105, "learning_rate": 9.1400441557401e-07, "loss": -0.0, "num_tokens": 2922003.0, "reward": 0.5679687261581421, "reward_std": 0.3106009364128113, "rewards/format_reward/mean": 0.5234375, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1905.0, "completions/max_terminated_length": 1905.0, "completions/mean_length": 267.96875, "completions/mean_terminated_length": 267.96875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0625, "frac_reward_zero_std": 0.375, "grad_norm": 17.019372940063477, "learning_rate": 9.045084971874737e-07, "loss": 0.0, "num_tokens": 3063369.0, "reward": 0.15390624105930328, "reward_std": 0.15965743362903595, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.109375, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "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": 1411.0, "completions/max_terminated_length": 1411.0, "completions/mean_length": 383.109375, "completions/mean_terminated_length": 383.109375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.065, "frac_reward_zero_std": 0.125, "grad_norm": 14.021422386169434, "learning_rate": 8.945702546981968e-07, "loss": -0.0, "num_tokens": 3169368.0, "reward": 0.48515620827674866, "reward_std": 0.37355077266693115, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 1282.0, "completions/max_terminated_length": 1282.0, "completions/mean_length": 251.25, "completions/mean_terminated_length": 251.25, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0675, "frac_reward_zero_std": 0.25, "grad_norm": 10.812904357910156, "learning_rate": 8.842005554284295e-07, "loss": -0.0, "num_tokens": 3264040.0, "reward": 0.5031249523162842, "reward_std": 0.38267990946769714, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.1259881556034088, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 1135.0, "completions/max_terminated_length": 1135.0, "completions/mean_length": 264.0625, "completions/mean_terminated_length": 264.0625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07, "frac_reward_zero_std": 0.375, "grad_norm": 19.734148025512695, "learning_rate": 8.734107384920769e-07, "loss": -0.0, "num_tokens": 3377820.0, "reward": 0.4374999701976776, "reward_std": 0.2899741530418396, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 28 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 981.0, "completions/max_terminated_length": 981.0, "completions/mean_length": 153.859375, "completions/mean_terminated_length": 153.859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0725, "frac_reward_zero_std": 0.5, "grad_norm": 5.822309970855713, "learning_rate": 8.622126023955445e-07, "loss": -0.0, "num_tokens": 3483099.0, "reward": 0.515625, "reward_std": 0.21751025319099426, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 830.0, "completions/mean_length": 313.484375, "completions/mean_terminated_length": 285.952392578125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.075, "frac_reward_zero_std": 0.25, "grad_norm": 11.583649635314941, "learning_rate": 8.506183921362442e-07, "loss": -0.0, "num_tokens": 3601034.0, "reward": 0.21249999105930328, "reward_std": 0.3205391764640808, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.19669894874095917, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 904.0, "completions/max_terminated_length": 904.0, "completions/mean_length": 281.390625, "completions/mean_terminated_length": 281.390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0775, "frac_reward_zero_std": 0.125, "grad_norm": 15.48983383178711, "learning_rate": 8.386407858128706e-07, "loss": 0.0, "num_tokens": 3693203.0, "reward": 0.42500001192092896, "reward_std": 0.29028111696243286, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.17817416787147522, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 1026.0, "completions/max_terminated_length": 1026.0, "completions/mean_length": 307.90625, "completions/mean_terminated_length": 307.90625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08, "frac_reward_zero_std": 0.125, "grad_norm": 14.248011589050293, "learning_rate": 8.262928807620843e-07, "loss": -0.0, "num_tokens": 3800549.0, "reward": 0.33515623211860657, "reward_std": 0.3557417690753937, "rewards/format_reward/mean": 0.5390625, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1425.0, "completions/max_terminated_length": 1425.0, "completions/mean_length": 289.796875, "completions/mean_terminated_length": 289.796875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0825, "frac_reward_zero_std": 0.5, "grad_norm": 7.1065592765808105, "learning_rate": 8.135881792367685e-07, "loss": -0.0, "num_tokens": 3920152.0, "reward": 0.44374996423721313, "reward_std": 0.210921972990036, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.1510545015335083, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 1083.0, "completions/max_terminated_length": 1083.0, "completions/mean_length": 325.71875, "completions/mean_terminated_length": 325.71875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.085, "frac_reward_zero_std": 0.375, "grad_norm": 2.761155128479004, "learning_rate": 8.005405736415125e-07, "loss": -0.0, "num_tokens": 4019150.0, "reward": 0.504687488079071, "reward_std": 0.274090051651001, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.1985812783241272, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 823.0, "completions/max_terminated_length": 823.0, "completions/mean_length": 132.390625, "completions/mean_terminated_length": 132.390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0875, "frac_reward_zero_std": 0.375, "grad_norm": 15.648600578308105, "learning_rate": 7.871643313414718e-07, "loss": 0.0, "num_tokens": 4109871.0, "reward": 0.36328125, "reward_std": 0.20542649924755096, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.10882443934679031, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 1283.0, "completions/max_terminated_length": 1283.0, "completions/mean_length": 297.0, "completions/mean_terminated_length": 297.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09, "frac_reward_zero_std": 0.25, "grad_norm": 18.359325408935547, "learning_rate": 7.734740790612136e-07, "loss": -0.0, "num_tokens": 4208727.0, "reward": 0.4531249403953552, "reward_std": 0.36290422081947327, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 1721.0, "completions/mean_length": 373.84375, "completions/mean_terminated_length": 347.2698669433594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0925, "frac_reward_zero_std": 0.375, "grad_norm": 11.33287239074707, "learning_rate": 7.594847868906076e-07, "loss": -0.0, "num_tokens": 4339405.0, "reward": 0.578125, "reward_std": 0.18085281550884247, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.21593283116817474, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 681.0, "completions/max_terminated_length": 681.0, "completions/mean_length": 152.390625, "completions/mean_terminated_length": 152.390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.095, "frac_reward_zero_std": 0.25, "grad_norm": 52.58133316040039, "learning_rate": 7.452117519152541e-07, "loss": 0.0, "num_tokens": 4416550.0, "reward": 0.6421874761581421, "reward_std": 0.161176860332489, "rewards/format_reward/mean": 0.640625, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 38 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1670.0, "completions/mean_length": 373.5625, "completions/mean_terminated_length": 291.2131042480469, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0975, "frac_reward_zero_std": 0.375, "grad_norm": 15.88608169555664, "learning_rate": 7.306705814893439e-07, "loss": 0.0, "num_tokens": 4542890.0, "reward": 0.15781249105930328, "reward_std": 0.2227596938610077, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.21764887869358063, "rewards/mcq_exact_match_reward/mean": 0.109375, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1166.0, "completions/mean_length": 401.953125, "completions/mean_terminated_length": 348.8548278808594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1, "frac_reward_zero_std": 0.0, "grad_norm": 12.829813003540039, "learning_rate": 7.158771761692464e-07, "loss": 0.0, "num_tokens": 4665815.0, "reward": 0.5539062023162842, "reward_std": 0.41745221614837646, "rewards/format_reward/mean": 0.5390625, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 1368.0, "completions/max_terminated_length": 1368.0, "completions/mean_length": 278.15625, "completions/mean_terminated_length": 278.15625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1025, "frac_reward_zero_std": 0.25, "grad_norm": 15.97768497467041, "learning_rate": 7.008477123264847e-07, "loss": -0.0, "num_tokens": 4792873.0, "reward": 0.668749988079071, "reward_std": 0.29939061403274536, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 41 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 954.0, "completions/max_terminated_length": 954.0, "completions/mean_length": 248.125, "completions/mean_terminated_length": 248.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.105, "frac_reward_zero_std": 0.25, "grad_norm": 12.298758506774902, "learning_rate": 6.855986244591103e-07, "loss": -0.0, "num_tokens": 4893665.0, "reward": 0.6046874523162842, "reward_std": 0.3279338777065277, "rewards/format_reward/mean": 0.578125, "rewards/format_reward/std": 0.25539806485176086, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 1442.0, "completions/max_terminated_length": 1442.0, "completions/mean_length": 310.28125, "completions/mean_terminated_length": 310.28125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1075, "frac_reward_zero_std": 0.25, "grad_norm": 16.986190795898438, "learning_rate": 6.701465872208216e-07, "loss": 0.0, "num_tokens": 5023603.0, "reward": 0.56640625, "reward_std": 0.300856351852417, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.2744719088077545, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1049.0, "completions/mean_length": 360.46875, "completions/mean_terminated_length": 333.68255615234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.11, "frac_reward_zero_std": 0.25, "grad_norm": 7.303002834320068, "learning_rate": 6.545084971874736e-07, "loss": -0.0, "num_tokens": 5141313.0, "reward": 0.3218750059604645, "reward_std": 0.22298568487167358, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.3149704039096832, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 1649.0, "completions/max_terminated_length": 1649.0, "completions/mean_length": 293.15625, "completions/mean_terminated_length": 293.15625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1125, "frac_reward_zero_std": 0.25, "grad_norm": 11.653733253479004, "learning_rate": 6.387014543809223e-07, "loss": -0.0, "num_tokens": 5275643.0, "reward": 0.46406248211860657, "reward_std": 0.31110072135925293, "rewards/format_reward/mean": 0.578125, "rewards/format_reward/std": 0.2221602201461792, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 1389.0, "completions/mean_length": 485.296875, "completions/mean_terminated_length": 460.4920959472656, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.115, "frac_reward_zero_std": 0.0, "grad_norm": 10.199931144714355, "learning_rate": 6.227427435703995e-07, "loss": -0.0, "num_tokens": 5389678.0, "reward": 0.37187498807907104, "reward_std": 0.4432055354118347, "rewards/format_reward/mean": 0.59375, "rewards/format_reward/std": 0.279951810836792, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 1009.0, "completions/max_terminated_length": 1009.0, "completions/mean_length": 311.71875, "completions/mean_terminated_length": 311.71875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1175, "frac_reward_zero_std": 0.25, "grad_norm": 12.196937561035156, "learning_rate": 6.066498153718734e-07, "loss": -0.0, "num_tokens": 5498884.0, "reward": 0.4546874761581421, "reward_std": 0.20098435878753662, "rewards/format_reward/mean": 0.640625, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 1233.0, "completions/max_terminated_length": 1233.0, "completions/mean_length": 367.3125, "completions/mean_terminated_length": 367.3125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.12, "frac_reward_zero_std": 0.125, "grad_norm": 8.141874313354492, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 5607456.0, "reward": 0.3531249761581421, "reward_std": 0.34032151103019714, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.18898223340511322, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 48 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1439.0, "completions/mean_length": 326.671875, "completions/mean_terminated_length": 299.3492126464844, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1225, "frac_reward_zero_std": 0.25, "grad_norm": 17.232126235961914, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 5741531.0, "reward": 0.6468750238418579, "reward_std": 0.24827098846435547, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.19669894874095917, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 1148.0, "completions/mean_length": 298.546875, "completions/mean_terminated_length": 270.7778015136719, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.125, "frac_reward_zero_std": 0.125, "grad_norm": 18.573505401611328, "learning_rate": 5.577423184847931e-07, "loss": -0.0, "num_tokens": 5858302.0, "reward": 0.55078125, "reward_std": 0.3845759332180023, "rewards/format_reward/mean": 0.6640625, "rewards/format_reward/std": 0.2680720090866089, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 50 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1832.0, "completions/max_terminated_length": 1832.0, "completions/mean_length": 280.015625, "completions/mean_terminated_length": 280.015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1275, "frac_reward_zero_std": 0.375, "grad_norm": 24.255617141723633, "learning_rate": 5.412896727361662e-07, "loss": 0.0, "num_tokens": 5996783.0, "reward": 0.4609375, "reward_std": 0.18493039906024933, "rewards/format_reward/mean": 0.546875, "rewards/format_reward/std": 0.21304203569889069, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 51 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1010.0, "completions/max_terminated_length": 1010.0, "completions/mean_length": 221.34375, "completions/mean_terminated_length": 221.34375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.13, "frac_reward_zero_std": 0.125, "grad_norm": 35.651187896728516, "learning_rate": 5.247918773366111e-07, "loss": -0.0, "num_tokens": 6100277.0, "reward": 0.5257812142372131, "reward_std": 0.38004207611083984, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.19654127955436707, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 52 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1924.0, "completions/max_terminated_length": 1924.0, "completions/mean_length": 261.6875, "completions/mean_terminated_length": 261.6875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1325, "frac_reward_zero_std": 0.375, "grad_norm": 14.250286102294922, "learning_rate": 5.082669723831793e-07, "loss": 0.0, "num_tokens": 6237737.0, "reward": 0.6656249761581421, "reward_std": 0.18033984303474426, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.18898223340511322, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 53 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 825.0, "completions/max_terminated_length": 825.0, "completions/mean_length": 173.515625, "completions/mean_terminated_length": 173.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.135, "frac_reward_zero_std": 0.25, "grad_norm": 28.05312728881836, "learning_rate": 4.917330276168208e-07, "loss": 0.0, "num_tokens": 6343826.0, "reward": 0.5882812142372131, "reward_std": 0.25916147232055664, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.1751912236213684, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 643.0, "completions/mean_length": 130.421875, "completions/mean_terminated_length": 99.984130859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1375, "frac_reward_zero_std": 0.375, "grad_norm": 17.20871925354004, "learning_rate": 4.752081226633888e-07, "loss": 0.0, "num_tokens": 6433477.0, "reward": 0.4749999940395355, "reward_std": 0.27743715047836304, "rewards/format_reward/mean": 0.6875, "rewards/format_reward/std": 0.2745848298072815, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1023.0, "completions/mean_length": 257.734375, "completions/mean_terminated_length": 229.31747436523438, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.14, "frac_reward_zero_std": 0.375, "grad_norm": 28.584617614746094, "learning_rate": 4.5871032726383385e-07, "loss": -0.0, "num_tokens": 6542052.0, "reward": 0.37031251192092896, "reward_std": 0.20041994750499725, "rewards/format_reward/mean": 0.578125, "rewards/format_reward/std": 0.2221602201461792, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 56 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1153.0, "completions/max_terminated_length": 1153.0, "completions/mean_length": 298.546875, "completions/mean_terminated_length": 298.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1425, "frac_reward_zero_std": 0.375, "grad_norm": 2.868027448654175, "learning_rate": 4.4225768151520694e-07, "loss": -0.0, "num_tokens": 6649327.0, "reward": 0.5249999761581421, "reward_std": 0.20744971930980682, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.208927720785141, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 57 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1315.0, "completions/max_terminated_length": 1315.0, "completions/mean_length": 351.78125, "completions/mean_terminated_length": 351.78125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.145, "frac_reward_zero_std": 0.25, "grad_norm": 5.103922367095947, "learning_rate": 4.258681761440789e-07, "loss": 0.0, "num_tokens": 6767001.0, "reward": 0.32421875, "reward_std": 0.35478106141090393, "rewards/format_reward/mean": 0.5859375, "rewards/format_reward/std": 0.2605654299259186, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 58 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 874.0, "completions/max_terminated_length": 874.0, "completions/mean_length": 184.21875, "completions/mean_terminated_length": 184.21875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1475, "frac_reward_zero_std": 0.25, "grad_norm": 24.614654541015625, "learning_rate": 4.095597328339452e-07, "loss": 0.0, "num_tokens": 6862319.0, "reward": 0.6890624761581421, "reward_std": 0.15067152678966522, "rewards/format_reward/mean": 0.640625, "rewards/format_reward/std": 0.22658175230026245, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 59 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1203.0, "completions/max_terminated_length": 1203.0, "completions/mean_length": 206.96875, "completions/mean_terminated_length": 206.96875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.15, "frac_reward_zero_std": 0.25, "grad_norm": 14.782899856567383, "learning_rate": 3.9335018462812664e-07, "loss": 0.0, "num_tokens": 6964613.0, "reward": 0.328125, "reward_std": 0.2112889140844345, "rewards/format_reward/mean": 0.78125, "rewards/format_reward/std": 0.265398770570755, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 60 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1986.0, "completions/max_terminated_length": 1986.0, "completions/mean_length": 249.578125, "completions/mean_terminated_length": 249.578125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1525, "frac_reward_zero_std": 0.25, "grad_norm": 38.279579162597656, "learning_rate": 3.772572564296004e-07, "loss": 0.0, "num_tokens": 7072938.0, "reward": 0.3515625, "reward_std": 0.24058622121810913, "rewards/format_reward/mean": 0.703125, "rewards/format_reward/std": 0.24750742316246033, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1140.0, "completions/mean_length": 185.34375, "completions/mean_terminated_length": 155.7777862548828, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.155, "frac_reward_zero_std": 0.125, "grad_norm": 21.222848892211914, "learning_rate": 3.612985456190778e-07, "loss": 0.0, "num_tokens": 7167584.0, "reward": 0.6851562261581421, "reward_std": 0.2669914662837982, "rewards/format_reward/mean": 0.7578125, "rewards/format_reward/std": 0.26714521646499634, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 1354.0, "completions/max_terminated_length": 1354.0, "completions/mean_length": 236.25, "completions/mean_terminated_length": 236.25, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1575, "frac_reward_zero_std": 0.375, "grad_norm": 13.26902961730957, "learning_rate": 3.454915028125263e-07, "loss": 0.0, "num_tokens": 7278392.0, "reward": 0.40937501192092896, "reward_std": 0.20505425333976746, "rewards/format_reward/mean": 0.8125, "rewards/format_reward/std": 0.2597312331199646, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 63 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1760.0, "completions/max_terminated_length": 1760.0, "completions/mean_length": 298.8125, "completions/mean_terminated_length": 298.8125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.16, "frac_reward_zero_std": 0.25, "grad_norm": 15.441954612731934, "learning_rate": 3.2985341277917846e-07, "loss": 0.0, "num_tokens": 7408788.0, "reward": 0.23593749105930328, "reward_std": 0.2559191882610321, "rewards/format_reward/mean": 0.640625, "rewards/format_reward/std": 0.301698237657547, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 64 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1933.0, "completions/max_terminated_length": 1933.0, "completions/mean_length": 197.765625, "completions/mean_terminated_length": 197.765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1625, "frac_reward_zero_std": 0.25, "grad_norm": 18.9749698638916, "learning_rate": 3.1440137554088953e-07, "loss": 0.0, "num_tokens": 7505557.0, "reward": 0.637499988079071, "reward_std": 0.3514561057090759, "rewards/format_reward/mean": 0.75, "rewards/format_reward/std": 0.2519763112068176, "rewards/mcq_exact_match_reward/mean": 0.5625, "rewards/mcq_exact_match_reward/std": 0.5, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 634.0, "completions/mean_length": 125.34375, "completions/mean_terminated_length": 94.82540130615234, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.165, "frac_reward_zero_std": 0.125, "grad_norm": 41.26137924194336, "learning_rate": 2.9915228767351535e-07, "loss": -0.0, "num_tokens": 7602339.0, "reward": 0.45234376192092896, "reward_std": 0.18625041842460632, "rewards/format_reward/mean": 0.7734375, "rewards/format_reward/std": 0.250866562128067, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 66 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 694.0, "completions/max_terminated_length": 694.0, "completions/mean_length": 125.875, "completions/mean_terminated_length": 125.875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1675, "frac_reward_zero_std": 0.125, "grad_norm": 45.19137191772461, "learning_rate": 2.841228238307536e-07, "loss": -0.0, "num_tokens": 7714491.0, "reward": 0.49921876192092896, "reward_std": 0.29034745693206787, "rewards/format_reward/mean": 0.7734375, "rewards/format_reward/std": 0.2662152051925659, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 67 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1291.0, "completions/max_terminated_length": 1291.0, "completions/mean_length": 120.421875, "completions/mean_terminated_length": 120.421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.17, "frac_reward_zero_std": 0.375, "grad_norm": 18.962387084960938, "learning_rate": 2.6932941851065615e-07, "loss": 0.0, "num_tokens": 7799758.0, "reward": 0.5257812738418579, "reward_std": 0.16320057213306427, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.2313210517168045, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 1747.0, "completions/max_terminated_length": 1747.0, "completions/mean_length": 172.296875, "completions/mean_terminated_length": 172.296875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1725, "frac_reward_zero_std": 0.375, "grad_norm": 12.699563026428223, "learning_rate": 2.547882480847461e-07, "loss": -0.0, "num_tokens": 7887265.0, "reward": 0.39531251788139343, "reward_std": 0.2821016311645508, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1078.0, "completions/mean_length": 179.875, "completions/mean_terminated_length": 150.22222900390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.175, "frac_reward_zero_std": 0.125, "grad_norm": 40.29182052612305, "learning_rate": 2.4051521310939254e-07, "loss": 0.0, "num_tokens": 7998809.0, "reward": 0.6945312023162842, "reward_std": 0.2864573299884796, "rewards/format_reward/mean": 0.8515625, "rewards/format_reward/std": 0.24688033759593964, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 70 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 625.0, "completions/max_terminated_length": 625.0, "completions/mean_length": 77.765625, "completions/mean_terminated_length": 77.765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1775, "frac_reward_zero_std": 0.25, "grad_norm": 22.098108291625977, "learning_rate": 2.2652592093878665e-07, "loss": 0.0, "num_tokens": 8102746.0, "reward": 0.5023437738418579, "reward_std": 0.30273789167404175, "rewards/format_reward/mean": 0.8046875, "rewards/format_reward/std": 0.2615155577659607, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 71 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 875.0, "completions/max_terminated_length": 875.0, "completions/mean_length": 134.25, "completions/mean_terminated_length": 134.25, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.18, "frac_reward_zero_std": 0.625, "grad_norm": 7.783851623535156, "learning_rate": 2.128356686585282e-07, "loss": 0.0, "num_tokens": 8212594.0, "reward": 0.29843753576278687, "reward_std": 0.16265606880187988, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.21304203569889069, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 719.0, "completions/max_terminated_length": 719.0, "completions/mean_length": 22.90625, "completions/mean_terminated_length": 22.90625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1825, "frac_reward_zero_std": 0.125, "grad_norm": 31.63098907470703, "learning_rate": 1.9945942635848745e-07, "loss": 0.0, "num_tokens": 8329172.0, "reward": 0.649218738079071, "reward_std": 0.3588632047176361, "rewards/format_reward/mean": 0.7109375, "rewards/format_reward/std": 0.24888142943382263, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 73 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 790.0, "completions/max_terminated_length": 790.0, "completions/mean_length": 123.84375, "completions/mean_terminated_length": 123.84375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.185, "frac_reward_zero_std": 0.25, "grad_norm": 9.745734214782715, "learning_rate": 1.8641182076323148e-07, "loss": -0.0, "num_tokens": 8431330.0, "reward": 0.5492187738418579, "reward_std": 0.3164989948272705, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 0.0, "completions/max_length": 451.0, "completions/max_terminated_length": 451.0, "completions/mean_length": 45.109375, "completions/mean_terminated_length": 45.109375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1875, "frac_reward_zero_std": 0.875, "grad_norm": 2.067138671875, "learning_rate": 1.7370711923791564e-07, "loss": -0.0, "num_tokens": 8493945.0, "reward": 0.5835937857627869, "reward_std": 0.04640388488769531, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 75 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1033.0, "completions/max_terminated_length": 1033.0, "completions/mean_length": 105.765625, "completions/mean_terminated_length": 105.765625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.19, "frac_reward_zero_std": 0.375, "grad_norm": 14.228410720825195, "learning_rate": 1.6135921418712955e-07, "loss": 0.0, "num_tokens": 8591890.0, "reward": 0.6625000238418579, "reward_std": 0.2798827290534973, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5625, "rewards/mcq_exact_match_reward/std": 0.5, "step": 76 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 312.0, "completions/max_terminated_length": 312.0, "completions/mean_length": 35.734375, "completions/mean_terminated_length": 35.734375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1925, "frac_reward_zero_std": 0.5, "grad_norm": 67.87721252441406, "learning_rate": 1.493816078637557e-07, "loss": 0.0, "num_tokens": 8663641.0, "reward": 0.58203125, "reward_std": 0.19731561839580536, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 631.0, "completions/mean_length": 82.34375, "completions/mean_terminated_length": 51.142860412597656, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.195, "frac_reward_zero_std": 0.5, "grad_norm": 15.133502960205078, "learning_rate": 1.3778739760445552e-07, "loss": 0.0, "num_tokens": 8768623.0, "reward": 0.6921875476837158, "reward_std": 0.19591379165649414, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 78 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 735.0, "completions/max_terminated_length": 735.0, "completions/mean_length": 43.484375, "completions/mean_terminated_length": 43.484375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1975, "frac_reward_zero_std": 0.625, "grad_norm": 10.814410209655762, "learning_rate": 1.2658926150792322e-07, "loss": 0.0, "num_tokens": 8856382.0, "reward": 0.71875, "reward_std": 0.1462521106004715, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 240.0, "completions/max_terminated_length": 240.0, "completions/mean_length": 18.71875, "completions/mean_terminated_length": 18.71875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2, "frac_reward_zero_std": 0.625, "grad_norm": 12.850092887878418, "learning_rate": 1.1579944457157059e-07, "loss": -0.0, "num_tokens": 8937236.0, "reward": 0.4593750238418579, "reward_std": 0.15992169082164764, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 80 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 567.0, "completions/max_terminated_length": 567.0, "completions/mean_length": 48.34375, "completions/mean_terminated_length": 48.34375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2025, "frac_reward_zero_std": 0.625, "grad_norm": 9.853850364685059, "learning_rate": 1.0542974530180327e-07, "loss": 0.0, "num_tokens": 9021906.0, "reward": 0.7406250238418579, "reward_std": 0.17358146607875824, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 81 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 350.0, "completions/max_terminated_length": 350.0, "completions/mean_length": 21.921875, "completions/mean_terminated_length": 21.921875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.205, "frac_reward_zero_std": 0.625, "grad_norm": 6.610433578491211, "learning_rate": 9.549150281252632e-08, "loss": 0.0, "num_tokens": 9123549.0, "reward": 0.3656250238418579, "reward_std": 0.15992169082164764, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 931.0, "completions/mean_length": 66.65625, "completions/mean_terminated_length": 35.20635223388672, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2075, "frac_reward_zero_std": 0.625, "grad_norm": 14.159682273864746, "learning_rate": 8.599558442598998e-08, "loss": 0.0, "num_tokens": 9225087.0, "reward": 0.565625011920929, "reward_std": 0.1536564975976944, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 83 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 972.0, "completions/max_terminated_length": 972.0, "completions/mean_length": 27.84375, "completions/mean_terminated_length": 27.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.21, "frac_reward_zero_std": 0.5, "grad_norm": 32.14434814453125, "learning_rate": 7.695237378953224e-08, "loss": -0.0, "num_tokens": 9349853.0, "reward": 0.6140625476837158, "reward_std": 0.10647734999656677, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 84 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 901.0, "completions/max_terminated_length": 901.0, "completions/mean_length": 43.84375, "completions/mean_terminated_length": 43.84375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2125, "frac_reward_zero_std": 0.875, "grad_norm": 2.7934324741363525, "learning_rate": 6.837175952121304e-08, "loss": 0.0, "num_tokens": 9432515.0, "reward": 0.36484378576278687, "reward_std": 0.04198446497321129, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 85 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1178.0, "completions/max_terminated_length": 1178.0, "completions/mean_length": 30.21875, "completions/mean_terminated_length": 30.21875, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.215, "frac_reward_zero_std": 0.75, "grad_norm": 7.896993637084961, "learning_rate": 6.026312439675551e-08, "loss": 0.0, "num_tokens": 9516033.0, "reward": 0.7718750238418579, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.671875, "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": 0.0, "completions/max_length": 466.0, "completions/max_terminated_length": 466.0, "completions/mean_length": 24.28125, "completions/mean_terminated_length": 24.28125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2175, "frac_reward_zero_std": 0.5, "grad_norm": 19.25350570678711, "learning_rate": 5.263533508961826e-08, "loss": 0.0, "num_tokens": 9609595.0, "reward": 0.9234375357627869, "reward_std": 0.17176643013954163, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.14689241349697113, "rewards/mcq_exact_match_reward/mean": 0.828125, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 87 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 672.0, "completions/max_terminated_length": 672.0, "completions/mean_length": 48.21875, "completions/mean_terminated_length": 48.21875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.22, "frac_reward_zero_std": 0.625, "grad_norm": 13.40875244140625, "learning_rate": 4.549673247541874e-08, "loss": 0.0, "num_tokens": 9713649.0, "reward": 0.46875, "reward_std": 0.1872510462999344, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 88 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 788.0, "completions/max_terminated_length": 788.0, "completions/mean_length": 32.109375, "completions/mean_terminated_length": 32.109375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2225, "frac_reward_zero_std": 0.375, "grad_norm": 15.456236839294434, "learning_rate": 3.8855122511307626e-08, "loss": 0.0, "num_tokens": 9810704.0, "reward": 0.6148437857627869, "reward_std": 0.26135772466659546, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 128.0, "completions/max_terminated_length": 128.0, "completions/mean_length": 20.53125, "completions/mean_terminated_length": 20.53125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.225, "frac_reward_zero_std": 0.75, "grad_norm": 6.198591232299805, "learning_rate": 3.271776770026963e-08, "loss": -0.0, "num_tokens": 9907490.0, "reward": 0.8500000238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.75, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 90 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 92.0, "completions/max_terminated_length": 92.0, "completions/mean_length": 13.125, "completions/mean_terminated_length": 13.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2275, "frac_reward_zero_std": 0.625, "grad_norm": 16.619112014770508, "learning_rate": 2.7091379149682682e-08, "loss": -0.0, "num_tokens": 10013050.0, "reward": 0.81640625, "reward_std": 0.13371865451335907, "rewards/format_reward/mean": 0.8203125, "rewards/format_reward/std": 0.3376763164997101, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 91 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 538.0, "completions/max_terminated_length": 538.0, "completions/mean_length": 40.125, "completions/mean_terminated_length": 40.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.23, "frac_reward_zero_std": 0.625, "grad_norm": 7.519993782043457, "learning_rate": 2.1982109232821176e-08, "loss": 0.0, "num_tokens": 10103962.0, "reward": 0.359375, "reward_std": 0.15992167592048645, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 208.0, "completions/max_terminated_length": 208.0, "completions/mean_length": 16.125, "completions/mean_terminated_length": 16.125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2325, "frac_reward_zero_std": 0.75, "grad_norm": 8.49631118774414, "learning_rate": 1.7395544861325718e-08, "loss": 0.0, "num_tokens": 10182930.0, "reward": 0.7718750238418579, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.671875, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 1528.0, "completions/max_terminated_length": 1528.0, "completions/mean_length": 80.78125, "completions/mean_terminated_length": 80.78125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.235, "frac_reward_zero_std": 0.5, "grad_norm": 25.295228958129883, "learning_rate": 1.3336701375997127e-08, "loss": 0.0, "num_tokens": 10285604.0, "reward": 0.671093761920929, "reward_std": 0.11763861775398254, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.19654127955436707, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 94 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 47.0, "completions/max_terminated_length": 47.0, "completions/mean_length": 11.875, "completions/mean_terminated_length": 11.875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2375, "frac_reward_zero_std": 0.625, "grad_norm": 16.39482307434082, "learning_rate": 9.810017062595321e-09, "loss": 0.0, "num_tokens": 10400204.0, "reward": 0.640625, "reward_std": 0.15992169082164764, "rewards/format_reward/mean": 0.9375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 95 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 114.0, "completions/max_terminated_length": 114.0, "completions/mean_length": 18.890625, "completions/mean_terminated_length": 18.890625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.24, "frac_reward_zero_std": 0.75, "grad_norm": 16.500764846801758, "learning_rate": 6.819348298638839e-09, "loss": 0.0, "num_tokens": 10489605.0, "reward": 0.9007812738418579, "reward_std": 0.11430564522743225, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.21347814798355103, "rewards/mcq_exact_match_reward/mean": 0.8125, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 96 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 186.0, "completions/max_terminated_length": 186.0, "completions/mean_length": 29.734375, "completions/mean_terminated_length": 29.734375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2425, "frac_reward_zero_std": 0.75, "grad_norm": 10.579672813415527, "learning_rate": 4.367965336512403e-09, "loss": 0.0, "num_tokens": 10580684.0, "reward": 0.6781250238418579, "reward_std": 0.10205793380737305, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 97 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 612.0, "completions/max_terminated_length": 612.0, "completions/mean_length": 35.953125, "completions/mean_terminated_length": 35.953125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.245, "frac_reward_zero_std": 0.375, "grad_norm": 8.055426597595215, "learning_rate": 2.458548727494292e-09, "loss": 0.0, "num_tokens": 10671393.0, "reward": 0.7078125476837158, "reward_std": 0.22539028525352478, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 98 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 761.0, "completions/max_terminated_length": 761.0, "completions/mean_length": 78.578125, "completions/mean_terminated_length": 78.578125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2475, "frac_reward_zero_std": 0.625, "grad_norm": 10.974425315856934, "learning_rate": 1.0931863906127325e-09, "loss": -0.0, "num_tokens": 10756950.0, "reward": 0.5375000238418579, "reward_std": 0.16675157845020294, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "step": 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": 863.0, "completions/max_terminated_length": 863.0, "completions/mean_length": 53.359375, "completions/mean_terminated_length": 53.359375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.25, "frac_reward_zero_std": 0.5, "grad_norm": 10.117315292358398, "learning_rate": 2.733713295369755e-10, "loss": 0.0, "num_tokens": 10859229.0, "reward": 0.5687500238418579, "reward_std": 0.24511480331420898, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 100 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 10859229, "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 }