{ "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1297.0, "completions/mean_length": 468.421875, "completions/mean_terminated_length": 417.4677429199219, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.0025, "frac_reward_zero_std": 0.0, "grad_norm": 11.304153442382812, "learning_rate": 0.0, "loss": 0.0, "num_tokens": 127899.0, "reward": 0.23359373211860657, "reward_std": 0.3493765890598297, "rewards/format_reward/mean": 0.3046875, "rewards/format_reward/std": 0.3036443889141083, "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": 1568.0, "completions/max_terminated_length": 1568.0, "completions/mean_length": 457.0, "completions/mean_terminated_length": 457.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.005, "frac_reward_zero_std": 0.0, "grad_norm": 5.644176959991455, "learning_rate": 2e-07, "loss": 0.0, "num_tokens": 218379.0, "reward": 0.23906248807907104, "reward_std": 0.37608200311660767, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.3503824472427368, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 1727.0, "completions/mean_length": 573.15625, "completions/mean_terminated_length": 549.74609375, "completions/min_length": 67.0, "completions/min_terminated_length": 67.0, "epoch": 0.0075, "frac_reward_zero_std": 0.0, "grad_norm": 1.492297887802124, "learning_rate": 4e-07, "loss": 0.0, "num_tokens": 356261.0, "reward": 0.1874999850988388, "reward_std": 0.2862415313720703, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.24397502839565277, "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": 0.0, "completions/max_length": 1575.0, "completions/max_terminated_length": 1575.0, "completions/mean_length": 375.578125, "completions/mean_terminated_length": 375.578125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.01, "frac_reward_zero_std": 0.0, "grad_norm": 6.434549331665039, "learning_rate": 6e-07, "loss": -0.0, "num_tokens": 463514.0, "reward": 0.3179687261581421, "reward_std": 0.40387412905693054, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.255761981010437, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 4 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 999.0, "completions/max_terminated_length": 999.0, "completions/mean_length": 337.25, "completions/mean_terminated_length": 337.25, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "epoch": 0.0125, "frac_reward_zero_std": 0.125, "grad_norm": 5.192480564117432, "learning_rate": 8e-07, "loss": -0.0, "num_tokens": 577474.0, "reward": 0.29609373211860657, "reward_std": 0.33613812923431396, "rewards/format_reward/mean": 0.3046875, "rewards/format_reward/std": 0.32874447107315063, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 1841.0, "completions/max_terminated_length": 1841.0, "completions/mean_length": 670.390625, "completions/mean_terminated_length": 670.390625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.015, "frac_reward_zero_std": 0.25, "grad_norm": 4.872166156768799, "learning_rate": 1e-06, "loss": 0.0, "num_tokens": 699379.0, "reward": 0.21250000596046448, "reward_std": 0.28724104166030884, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.25, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 6 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1406.0, "completions/mean_length": 436.34375, "completions/mean_terminated_length": 410.7619323730469, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.0175, "frac_reward_zero_std": 0.0, "grad_norm": 17.72997283935547, "learning_rate": 9.99726628670463e-07, "loss": 0.0, "num_tokens": 819113.0, "reward": 0.2851562201976776, "reward_std": 0.2501617968082428, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.2302463799715042, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 7 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1215.0, "completions/max_terminated_length": 1215.0, "completions/mean_length": 459.46875, "completions/mean_terminated_length": 459.46875, "completions/min_length": 122.0, "completions/min_terminated_length": 122.0, "epoch": 0.02, "frac_reward_zero_std": 0.0, "grad_norm": 1.1645325422286987, "learning_rate": 9.989068136093872e-07, "loss": -0.0, "num_tokens": 918647.0, "reward": 0.39765623211860657, "reward_std": 0.4350404143333435, "rewards/format_reward/mean": 0.3828125, "rewards/format_reward/std": 0.2313210517168045, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.0625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1673.0, "completions/mean_length": 653.9375, "completions/mean_terminated_length": 561.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0225, "frac_reward_zero_std": 0.0, "grad_norm": 11.303831100463867, "learning_rate": 9.975414512725056e-07, "loss": 0.0, "num_tokens": 1064107.0, "reward": 0.23515623807907104, "reward_std": 0.3031618595123291, "rewards/format_reward/mean": 0.3203125, "rewards/format_reward/std": 0.24180518090724945, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 1701.0, "completions/max_terminated_length": 1701.0, "completions/mean_length": 616.46875, "completions/mean_terminated_length": 616.46875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.025, "frac_reward_zero_std": 0.0, "grad_norm": 6.167299747467041, "learning_rate": 9.956320346634875e-07, "loss": 0.0, "num_tokens": 1208977.0, "reward": 0.31953123211860657, "reward_std": 0.42240801453590393, "rewards/format_reward/mean": 0.3828125, "rewards/format_reward/std": 0.21347814798355103, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 10 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1045.0, "completions/max_terminated_length": 1045.0, "completions/mean_length": 411.8125, "completions/mean_terminated_length": 411.8125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0275, "frac_reward_zero_std": 0.125, "grad_norm": 5.666714668273926, "learning_rate": 9.931806517013612e-07, "loss": -0.0, "num_tokens": 1323021.0, "reward": 0.1953124850988388, "reward_std": 0.2733513414859772, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.2083333432674408, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 1725.0, "completions/max_terminated_length": 1725.0, "completions/mean_length": 579.484375, "completions/mean_terminated_length": 579.484375, "completions/min_length": 192.0, "completions/min_terminated_length": 192.0, "epoch": 0.03, "frac_reward_zero_std": 0.0, "grad_norm": 0.9449681043624878, "learning_rate": 9.901899829374047e-07, "loss": -0.0, "num_tokens": 1447348.0, "reward": 0.4195312261581421, "reward_std": 0.4747834801673889, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.180765300989151, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 2011.0, "completions/mean_length": 714.265625, "completions/mean_terminated_length": 693.0952758789062, "completions/min_length": 290.0, "completions/min_terminated_length": 290.0, "epoch": 0.0325, "frac_reward_zero_std": 0.0, "grad_norm": 0.721289336681366, "learning_rate": 9.866632986240029e-07, "loss": 0.0, "num_tokens": 1583645.0, "reward": 0.23671874403953552, "reward_std": 0.31869614124298096, "rewards/format_reward/mean": 0.3359375, "rewards/format_reward/std": 0.25283610820770264, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 1182.0, "completions/mean_length": 457.390625, "completions/mean_terminated_length": 432.14288330078125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.035, "frac_reward_zero_std": 0.0, "grad_norm": 6.802702903747559, "learning_rate": 9.826044551386742e-07, "loss": -0.0, "num_tokens": 1699502.0, "reward": 0.21953123807907104, "reward_std": 0.3024585247039795, "rewards/format_reward/mean": 0.3203125, "rewards/format_reward/std": 0.24180518090724945, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 1272.0, "completions/max_terminated_length": 1272.0, "completions/mean_length": 497.671875, "completions/mean_terminated_length": 497.671875, "completions/min_length": 7.0, "completions/min_terminated_length": 7.0, "epoch": 0.0375, "frac_reward_zero_std": 0.0, "grad_norm": 6.346897125244141, "learning_rate": 9.780178907671788e-07, "loss": 0.0, "num_tokens": 1798921.0, "reward": 0.3671874701976776, "reward_std": 0.4178835153579712, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.22658175230026245, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 2019.0, "completions/max_terminated_length": 2019.0, "completions/mean_length": 602.859375, "completions/mean_terminated_length": 602.859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04, "frac_reward_zero_std": 0.0, "grad_norm": 5.839832305908203, "learning_rate": 9.729086208503173e-07, "loss": -0.0, "num_tokens": 1943776.0, "reward": 0.46328121423721313, "reward_std": 0.4243572950363159, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1925.0, "completions/mean_length": 508.171875, "completions/mean_terminated_length": 458.5, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0425, "frac_reward_zero_std": 0.125, "grad_norm": 2.3104865550994873, "learning_rate": 9.672822322997304e-07, "loss": 0.0, "num_tokens": 2068571.0, "reward": 0.19687499105930328, "reward_std": 0.2780441641807556, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.19669894874095917, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 1491.0, "completions/mean_length": 554.4375, "completions/mean_terminated_length": 530.7301635742188, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.045, "frac_reward_zero_std": 0.0, "grad_norm": 3.1332554817199707, "learning_rate": 9.611448774886923e-07, "loss": 0.0, "num_tokens": 2176615.0, "reward": 0.2499999701976776, "reward_std": 0.3216208219528198, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.2597312331199646, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1503.0, "completions/mean_length": 588.609375, "completions/mean_terminated_length": 516.8360595703125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0475, "frac_reward_zero_std": 0.0, "grad_norm": 5.148056507110596, "learning_rate": 9.545032675245813e-07, "loss": -0.0, "num_tokens": 2306814.0, "reward": 0.4007812440395355, "reward_std": 0.4721751809120178, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 1348.0, "completions/mean_length": 577.90625, "completions/mean_terminated_length": 530.4838256835938, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05, "frac_reward_zero_std": 0.0, "grad_norm": 4.143627166748047, "learning_rate": 9.473646649103817e-07, "loss": -0.0, "num_tokens": 2431408.0, "reward": 0.3843749761581421, "reward_std": 0.40651530027389526, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.233588308095932, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1748.0, "completions/mean_length": 596.796875, "completions/mean_terminated_length": 573.761962890625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0525, "frac_reward_zero_std": 0.125, "grad_norm": 2.0999743938446045, "learning_rate": 9.397368756032444e-07, "loss": -0.0, "num_tokens": 2564419.0, "reward": 0.3570312261581421, "reward_std": 0.42330649495124817, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 21 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1157.0, "completions/max_terminated_length": 1157.0, "completions/mean_length": 406.109375, "completions/mean_terminated_length": 406.109375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.055, "frac_reward_zero_std": 0.125, "grad_norm": 10.010024070739746, "learning_rate": 9.316282404787869e-07, "loss": 0.0, "num_tokens": 2677154.0, "reward": 0.30937498807907104, "reward_std": 0.3836524486541748, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.208927720785141, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 22 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1490.0, "completions/mean_length": 688.15625, "completions/mean_terminated_length": 666.5714721679688, "completions/min_length": 243.0, "completions/min_terminated_length": 243.0, "epoch": 0.0575, "frac_reward_zero_std": 0.0, "grad_norm": 0.798880398273468, "learning_rate": 9.230476262104676e-07, "loss": 0.0, "num_tokens": 2800380.0, "reward": 0.32499998807907104, "reward_std": 0.3335176408290863, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1417.0, "completions/max_terminated_length": 1417.0, "completions/mean_length": 435.09375, "completions/mean_terminated_length": 435.09375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06, "frac_reward_zero_std": 0.0, "grad_norm": 6.452167987823486, "learning_rate": 9.1400441557401e-07, "loss": -0.0, "num_tokens": 2934562.0, "reward": 0.4476562440395355, "reward_std": 0.4432813227176666, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 24 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1562.0, "completions/max_terminated_length": 1562.0, "completions/mean_length": 641.609375, "completions/mean_terminated_length": 641.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0625, "frac_reward_zero_std": 0.125, "grad_norm": 1.2295587062835693, "learning_rate": 9.045084971874737e-07, "loss": -0.0, "num_tokens": 3057937.0, "reward": 0.3414062261581421, "reward_std": 0.4102392792701721, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 1530.0, "completions/max_terminated_length": 1530.0, "completions/mean_length": 609.21875, "completions/mean_terminated_length": 609.21875, "completions/min_length": 10.0, "completions/min_terminated_length": 10.0, "epoch": 0.065, "frac_reward_zero_std": 0.25, "grad_norm": 3.503051519393921, "learning_rate": 8.945702546981968e-07, "loss": -0.0, "num_tokens": 3162327.0, "reward": 0.37578123807907104, "reward_std": 0.34600451588630676, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.1649840772151947, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1725.0, "completions/mean_length": 599.4375, "completions/mean_terminated_length": 552.7096557617188, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0675, "frac_reward_zero_std": 0.0, "grad_norm": 6.449777126312256, "learning_rate": 8.842005554284295e-07, "loss": -0.0, "num_tokens": 3288011.0, "reward": 0.3281249701976776, "reward_std": 0.45478343963623047, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.1510545015335083, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1223.0, "completions/mean_length": 420.03125, "completions/mean_terminated_length": 394.19049072265625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07, "frac_reward_zero_std": 0.125, "grad_norm": 9.595853805541992, "learning_rate": 8.734107384920769e-07, "loss": 0.0, "num_tokens": 3403085.0, "reward": 0.16484373807907104, "reward_std": 0.2631952166557312, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.20275264978408813, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 1400.0, "completions/max_terminated_length": 1400.0, "completions/mean_length": 542.390625, "completions/mean_terminated_length": 542.390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0725, "frac_reward_zero_std": 0.0, "grad_norm": 12.810264587402344, "learning_rate": 8.622126023955445e-07, "loss": -0.0, "num_tokens": 3518558.0, "reward": 0.5773437023162842, "reward_std": 0.39534229040145874, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.13524486124515533, "rewards/mcq_exact_match_reward/mean": 0.53125, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1365.0, "completions/mean_length": 588.53125, "completions/mean_terminated_length": 541.4515991210938, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.075, "frac_reward_zero_std": 0.0, "grad_norm": 1.7052180767059326, "learning_rate": 8.506183921362442e-07, "loss": -0.0, "num_tokens": 3647024.0, "reward": 0.44765621423721313, "reward_std": 0.4024611711502075, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 1290.0, "completions/max_terminated_length": 1290.0, "completions/mean_length": 605.84375, "completions/mean_terminated_length": 605.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0775, "frac_reward_zero_std": 0.125, "grad_norm": 1.3692716360092163, "learning_rate": 8.386407858128706e-07, "loss": 0.0, "num_tokens": 3796790.0, "reward": 0.24062499403953552, "reward_std": 0.24036644399166107, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.2182178944349289, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 1630.0, "completions/max_terminated_length": 1630.0, "completions/mean_length": 620.140625, "completions/mean_terminated_length": 620.140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08, "frac_reward_zero_std": 0.125, "grad_norm": 1.616221308708191, "learning_rate": 8.262928807620843e-07, "loss": 0.0, "num_tokens": 3912559.0, "reward": 0.46718746423721313, "reward_std": 0.313557505607605, "rewards/format_reward/mean": 0.453125, "rewards/format_reward/std": 0.14689241349697113, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 1188.0, "completions/max_terminated_length": 1188.0, "completions/mean_length": 407.609375, "completions/mean_terminated_length": 407.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0825, "frac_reward_zero_std": 0.0, "grad_norm": 10.067407608032227, "learning_rate": 8.135881792367685e-07, "loss": 0.0, "num_tokens": 4030382.0, "reward": 0.4351562261581421, "reward_std": 0.3817477226257324, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.180765300989151, "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": 1243.0, "completions/max_terminated_length": 1243.0, "completions/mean_length": 604.078125, "completions/mean_terminated_length": 604.078125, "completions/min_length": 20.0, "completions/min_terminated_length": 20.0, "epoch": 0.085, "frac_reward_zero_std": 0.0, "grad_norm": 3.3467695713043213, "learning_rate": 8.005405736415125e-07, "loss": 0.0, "num_tokens": 4157595.0, "reward": 0.28203123807907104, "reward_std": 0.3044944703578949, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.1875000149011612, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 1659.0, "completions/max_terminated_length": 1659.0, "completions/mean_length": 545.578125, "completions/mean_terminated_length": 545.578125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0875, "frac_reward_zero_std": 0.125, "grad_norm": 3.5687899589538574, "learning_rate": 7.871643313414718e-07, "loss": -0.0, "num_tokens": 4286848.0, "reward": 0.500781238079071, "reward_std": 0.4046732485294342, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 35 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1587.0, "completions/mean_length": 765.296875, "completions/mean_terminated_length": 702.2130737304688, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09, "frac_reward_zero_std": 0.0, "grad_norm": 6.329766750335693, "learning_rate": 7.734740790612136e-07, "loss": 0.0, "num_tokens": 4459339.0, "reward": 0.26093748211860657, "reward_std": 0.3628954291343689, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.18298126757144928, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 36 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1633.0, "completions/max_terminated_length": 1633.0, "completions/mean_length": 554.546875, "completions/mean_terminated_length": 554.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0925, "frac_reward_zero_std": 0.125, "grad_norm": 4.2150654792785645, "learning_rate": 7.594847868906076e-07, "loss": -0.0, "num_tokens": 4577502.0, "reward": 0.3531249761581421, "reward_std": 0.3236146867275238, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.265398770570755, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1866.0, "completions/mean_length": 789.234375, "completions/mean_terminated_length": 748.6290283203125, "completions/min_length": 140.0, "completions/min_terminated_length": 140.0, "epoch": 0.095, "frac_reward_zero_std": 0.125, "grad_norm": 0.7523370981216431, "learning_rate": 7.452117519152541e-07, "loss": -0.0, "num_tokens": 4705509.0, "reward": 0.34218746423721313, "reward_std": 0.40115296840667725, "rewards/format_reward/mean": 0.453125, "rewards/format_reward/std": 0.14689241349697113, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 38 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1886.0, "completions/mean_length": 600.453125, "completions/mean_terminated_length": 577.4761962890625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0975, "frac_reward_zero_std": 0.25, "grad_norm": 4.145143032073975, "learning_rate": 7.306705814893439e-07, "loss": 0.0, "num_tokens": 4829658.0, "reward": 0.28125, "reward_std": 0.2720353305339813, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1634.0, "completions/mean_length": 660.015625, "completions/mean_terminated_length": 637.9841918945312, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1, "frac_reward_zero_std": 0.125, "grad_norm": 0.7042872905731201, "learning_rate": 7.158771761692464e-07, "loss": 0.0, "num_tokens": 4956427.0, "reward": 0.29609373211860657, "reward_std": 0.29663288593292236, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.13524486124515533, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 40 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1914.0, "completions/mean_length": 648.265625, "completions/mean_terminated_length": 626.0476684570312, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1025, "frac_reward_zero_std": 0.0, "grad_norm": 6.571728229522705, "learning_rate": 7.008477123264847e-07, "loss": -0.0, "num_tokens": 5087460.0, "reward": 0.46718746423721313, "reward_std": 0.4204556345939636, "rewards/format_reward/mean": 0.453125, "rewards/format_reward/std": 0.14689241349697113, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 1924.0, "completions/mean_length": 639.6875, "completions/mean_terminated_length": 617.3333740234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.105, "frac_reward_zero_std": 0.125, "grad_norm": 3.192551851272583, "learning_rate": 6.855986244591103e-07, "loss": -0.0, "num_tokens": 5224936.0, "reward": 0.49921876192092896, "reward_std": 0.33688807487487793, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.13524486124515533, "rewards/mcq_exact_match_reward/mean": 0.453125, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1979.0, "completions/mean_length": 773.40625, "completions/mean_terminated_length": 732.290283203125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1075, "frac_reward_zero_std": 0.125, "grad_norm": 2.8935868740081787, "learning_rate": 6.701465872208216e-07, "loss": -0.0, "num_tokens": 5377194.0, "reward": 0.3257812261581421, "reward_std": 0.3597572445869446, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1757.0, "completions/max_terminated_length": 1757.0, "completions/mean_length": 673.65625, "completions/mean_terminated_length": 673.65625, "completions/min_length": 243.0, "completions/min_terminated_length": 243.0, "epoch": 0.11, "frac_reward_zero_std": 0.25, "grad_norm": 0.6974867582321167, "learning_rate": 6.545084971874736e-07, "loss": -0.0, "num_tokens": 5498052.0, "reward": 0.3140624761581421, "reward_std": 0.36175572872161865, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "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": 1337.0, "completions/max_terminated_length": 1337.0, "completions/mean_length": 391.140625, "completions/mean_terminated_length": 391.140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1125, "frac_reward_zero_std": 0.25, "grad_norm": 7.897181987762451, "learning_rate": 6.387014543809223e-07, "loss": -0.0, "num_tokens": 5606173.0, "reward": 0.5343749523162842, "reward_std": 0.32660090923309326, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.08908708393573761, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1600.0, "completions/mean_length": 574.671875, "completions/mean_terminated_length": 551.2857666015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.115, "frac_reward_zero_std": 0.125, "grad_norm": 6.948278903961182, "learning_rate": 6.227427435703995e-07, "loss": -0.0, "num_tokens": 5757840.0, "reward": 0.2890625, "reward_std": 0.34619027376174927, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.2083333432674408, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1454.0, "completions/mean_length": 310.109375, "completions/mean_terminated_length": 282.5238342285156, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1175, "frac_reward_zero_std": 0.125, "grad_norm": 17.003774642944336, "learning_rate": 6.066498153718734e-07, "loss": -0.0, "num_tokens": 5872927.0, "reward": 0.49453121423721313, "reward_std": 0.26769620180130005, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 1465.0, "completions/mean_length": 520.15625, "completions/mean_terminated_length": 495.90478515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.12, "frac_reward_zero_std": 0.0, "grad_norm": 9.9341459274292, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 6035497.0, "reward": 0.29374998807907104, "reward_std": 0.3407875597476959, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 1932.0, "completions/mean_length": 370.703125, "completions/mean_terminated_length": 344.0793762207031, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1225, "frac_reward_zero_std": 0.25, "grad_norm": 14.393939971923828, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 6159526.0, "reward": 0.3906249701976776, "reward_std": 0.25874659419059753, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1324.0, "completions/mean_length": 315.03125, "completions/mean_terminated_length": 287.5238342285156, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.125, "frac_reward_zero_std": 0.25, "grad_norm": 21.79771614074707, "learning_rate": 5.577423184847931e-07, "loss": -0.0, "num_tokens": 6266488.0, "reward": 0.4492187201976776, "reward_std": 0.29772311449050903, "rewards/format_reward/mean": 0.4296875, "rewards/format_reward/std": 0.1751912236213684, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1364.0, "completions/mean_length": 422.046875, "completions/mean_terminated_length": 342.0819396972656, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1275, "frac_reward_zero_std": 0.25, "grad_norm": 13.080544471740723, "learning_rate": 5.412896727361662e-07, "loss": 0.0, "num_tokens": 6384403.0, "reward": 0.29765623807907104, "reward_std": 0.22971822321414948, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 1457.0, "completions/max_terminated_length": 1457.0, "completions/mean_length": 266.25, "completions/mean_terminated_length": 266.25, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.13, "frac_reward_zero_std": 0.0, "grad_norm": 18.36222267150879, "learning_rate": 5.247918773366111e-07, "loss": -0.0, "num_tokens": 6496579.0, "reward": 0.4859374761581421, "reward_std": 0.4036346673965454, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 1142.0, "completions/max_terminated_length": 1142.0, "completions/mean_length": 364.578125, "completions/mean_terminated_length": 364.578125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1325, "frac_reward_zero_std": 0.5, "grad_norm": 11.033051490783691, "learning_rate": 5.082669723831793e-07, "loss": -0.0, "num_tokens": 6612704.0, "reward": 0.18906249105930328, "reward_std": 0.22160840034484863, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "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": 1600.0, "completions/max_terminated_length": 1600.0, "completions/mean_length": 235.90625, "completions/mean_terminated_length": 235.90625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.135, "frac_reward_zero_std": 0.125, "grad_norm": 15.763660430908203, "learning_rate": 4.917330276168208e-07, "loss": -0.0, "num_tokens": 6719674.0, "reward": 0.5320311784744263, "reward_std": 0.3425878882408142, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1638.0, "completions/max_terminated_length": 1638.0, "completions/mean_length": 229.125, "completions/mean_terminated_length": 229.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1375, "frac_reward_zero_std": 0.5, "grad_norm": 9.227850914001465, "learning_rate": 4.752081226633888e-07, "loss": -0.0, "num_tokens": 6807754.0, "reward": 0.4093749523162842, "reward_std": 0.2109457403421402, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 1620.0, "completions/max_terminated_length": 1620.0, "completions/mean_length": 344.546875, "completions/mean_terminated_length": 344.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.14, "frac_reward_zero_std": 0.125, "grad_norm": 11.433066368103027, "learning_rate": 4.5871032726383385e-07, "loss": 0.0, "num_tokens": 6923109.0, "reward": 0.14140625298023224, "reward_std": 0.26811346411705017, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.09375, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "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": 1071.0, "completions/mean_length": 377.546875, "completions/mean_terminated_length": 351.0317687988281, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1425, "frac_reward_zero_std": 0.375, "grad_norm": 11.500009536743164, "learning_rate": 4.4225768151520694e-07, "loss": -0.0, "num_tokens": 7045600.0, "reward": 0.4703124761581421, "reward_std": 0.2971927225589752, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 1814.0, "completions/max_terminated_length": 1814.0, "completions/mean_length": 307.3125, "completions/mean_terminated_length": 307.3125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.145, "frac_reward_zero_std": 0.25, "grad_norm": 8.84330940246582, "learning_rate": 4.258681761440789e-07, "loss": -0.0, "num_tokens": 7150740.0, "reward": 0.34375, "reward_std": 0.3180304169654846, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 0.0625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1981.0, "completions/mean_length": 501.765625, "completions/mean_terminated_length": 398.683349609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1475, "frac_reward_zero_std": 0.125, "grad_norm": 7.77143669128418, "learning_rate": 4.095597328339452e-07, "loss": -0.0, "num_tokens": 7289501.0, "reward": 0.16718748211860657, "reward_std": 0.26362812519073486, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.18298126757144928, "rewards/mcq_exact_match_reward/mean": 0.125, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1667.0, "completions/mean_length": 393.03125, "completions/mean_terminated_length": 339.6451416015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.15, "frac_reward_zero_std": 0.125, "grad_norm": 14.823182106018066, "learning_rate": 3.9335018462812664e-07, "loss": -0.0, "num_tokens": 7404215.0, "reward": 0.45234373211860657, "reward_std": 0.41068345308303833, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.13524486124515533, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 1333.0, "completions/max_terminated_length": 1333.0, "completions/mean_length": 237.46875, "completions/mean_terminated_length": 237.46875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1525, "frac_reward_zero_std": 0.25, "grad_norm": 14.556740760803223, "learning_rate": 3.772572564296004e-07, "loss": 0.0, "num_tokens": 7527005.0, "reward": 0.34453123807907104, "reward_std": 0.22420553863048553, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 0.0, "completions/max_length": 1878.0, "completions/max_terminated_length": 1878.0, "completions/mean_length": 364.421875, "completions/mean_terminated_length": 364.421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.155, "frac_reward_zero_std": 0.125, "grad_norm": 9.945978164672852, "learning_rate": 3.612985456190778e-07, "loss": -0.0, "num_tokens": 7625120.0, "reward": 0.5453124642372131, "reward_std": 0.40885093808174133, "rewards/format_reward/mean": 0.453125, "rewards/format_reward/std": 0.14689241349697113, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 1189.0, "completions/max_terminated_length": 1189.0, "completions/mean_length": 228.28125, "completions/mean_terminated_length": 228.28125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1575, "frac_reward_zero_std": 0.375, "grad_norm": 10.00886344909668, "learning_rate": 3.454915028125263e-07, "loss": 0.0, "num_tokens": 7725042.0, "reward": 0.47109371423721313, "reward_std": 0.2691788673400879, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 1563.0, "completions/max_terminated_length": 1563.0, "completions/mean_length": 320.328125, "completions/mean_terminated_length": 320.328125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.16, "frac_reward_zero_std": 0.375, "grad_norm": 26.35382843017578, "learning_rate": 3.2985341277917846e-07, "loss": -0.0, "num_tokens": 7828679.0, "reward": 0.2828124761581421, "reward_std": 0.2784997820854187, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 1084.0, "completions/mean_length": 293.53125, "completions/mean_terminated_length": 265.68255615234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1625, "frac_reward_zero_std": 0.125, "grad_norm": 13.174885749816895, "learning_rate": 3.1440137554088953e-07, "loss": 0.0, "num_tokens": 7937721.0, "reward": 0.2953124940395355, "reward_std": 0.347170889377594, "rewards/format_reward/mean": 0.453125, "rewards/format_reward/std": 0.14689241349697113, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 419.0, "completions/max_terminated_length": 419.0, "completions/mean_length": 19.59375, "completions/mean_terminated_length": 19.59375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.165, "frac_reward_zero_std": 0.625, "grad_norm": 11.680887222290039, "learning_rate": 2.9915228767351535e-07, "loss": 0.0, "num_tokens": 8048871.0, "reward": 0.4406249523162842, "reward_std": 0.13258251547813416, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 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": 825.0, "completions/max_terminated_length": 825.0, "completions/mean_length": 193.359375, "completions/mean_terminated_length": 193.359375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1675, "frac_reward_zero_std": 0.25, "grad_norm": 12.663226127624512, "learning_rate": 2.841228238307536e-07, "loss": -0.0, "num_tokens": 8129878.0, "reward": 0.5171874761581421, "reward_std": 0.3607522249221802, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 1236.0, "completions/max_terminated_length": 1236.0, "completions/mean_length": 151.171875, "completions/mean_terminated_length": 151.171875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.17, "frac_reward_zero_std": 0.25, "grad_norm": 17.396196365356445, "learning_rate": 2.6932941851065615e-07, "loss": -0.0, "num_tokens": 8237217.0, "reward": 0.43984371423721313, "reward_std": 0.32888293266296387, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1072.0, "completions/mean_length": 280.390625, "completions/mean_terminated_length": 223.37095642089844, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1725, "frac_reward_zero_std": 0.375, "grad_norm": 19.55508041381836, "learning_rate": 2.547882480847461e-07, "loss": -0.0, "num_tokens": 8352442.0, "reward": 0.3921874761581421, "reward_std": 0.3110659718513489, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1640.0, "completions/max_terminated_length": 1640.0, "completions/mean_length": 217.453125, "completions/mean_terminated_length": 217.453125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.175, "frac_reward_zero_std": 0.375, "grad_norm": 18.596508026123047, "learning_rate": 2.4051521310939254e-07, "loss": -0.0, "num_tokens": 8462631.0, "reward": 0.46953120827674866, "reward_std": 0.3010781407356262, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 1443.0, "completions/max_terminated_length": 1443.0, "completions/mean_length": 216.296875, "completions/mean_terminated_length": 216.296875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1775, "frac_reward_zero_std": 0.375, "grad_norm": 17.10657501220703, "learning_rate": 2.2652592093878665e-07, "loss": -0.0, "num_tokens": 8552530.0, "reward": 0.37812498211860657, "reward_std": 0.2777610421180725, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 1248.0, "completions/max_terminated_length": 1248.0, "completions/mean_length": 174.625, "completions/mean_terminated_length": 174.625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.18, "frac_reward_zero_std": 0.5, "grad_norm": 13.355034828186035, "learning_rate": 2.128356686585282e-07, "loss": -0.0, "num_tokens": 8645914.0, "reward": 0.25312498211860657, "reward_std": 0.2446400672197342, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "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": 1594.0, "completions/max_terminated_length": 1594.0, "completions/mean_length": 300.984375, "completions/mean_terminated_length": 300.984375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1825, "frac_reward_zero_std": 0.375, "grad_norm": 11.026819229125977, "learning_rate": 1.9945942635848745e-07, "loss": -0.0, "num_tokens": 8736817.0, "reward": 0.45390623807907104, "reward_std": 0.26449376344680786, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 1856.0, "completions/max_terminated_length": 1856.0, "completions/mean_length": 69.359375, "completions/mean_terminated_length": 69.359375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.185, "frac_reward_zero_std": 0.375, "grad_norm": 19.433134078979492, "learning_rate": 1.8641182076323148e-07, "loss": -0.0, "num_tokens": 8837448.0, "reward": 0.42421871423721313, "reward_std": 0.2760087847709656, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 835.0, "completions/max_terminated_length": 835.0, "completions/mean_length": 131.484375, "completions/mean_terminated_length": 131.484375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1875, "frac_reward_zero_std": 0.5, "grad_norm": 21.523534774780273, "learning_rate": 1.7370711923791564e-07, "loss": -0.0, "num_tokens": 8925967.0, "reward": 0.4554687440395355, "reward_std": 0.25517284870147705, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 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": 1546.0, "completions/max_terminated_length": 1546.0, "completions/mean_length": 180.875, "completions/mean_terminated_length": 180.875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.19, "frac_reward_zero_std": 0.5, "grad_norm": 13.826128005981445, "learning_rate": 1.6135921418712955e-07, "loss": -0.0, "num_tokens": 9019887.0, "reward": 0.6578124761581421, "reward_std": 0.21383655071258545, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 838.0, "completions/max_terminated_length": 838.0, "completions/mean_length": 134.609375, "completions/mean_terminated_length": 134.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1925, "frac_reward_zero_std": 0.5, "grad_norm": 8.817794799804688, "learning_rate": 1.493816078637557e-07, "loss": 0.0, "num_tokens": 9107494.0, "reward": 0.3765624761581421, "reward_std": 0.16265341639518738, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 983.0, "completions/max_terminated_length": 983.0, "completions/mean_length": 94.328125, "completions/mean_terminated_length": 94.328125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.195, "frac_reward_zero_std": 0.375, "grad_norm": 25.716712951660156, "learning_rate": 1.3778739760445552e-07, "loss": -0.0, "num_tokens": 9213027.0, "reward": 0.5968749523162842, "reward_std": 0.2688094973564148, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 1001.0, "completions/max_terminated_length": 1001.0, "completions/mean_length": 72.140625, "completions/mean_terminated_length": 72.140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1975, "frac_reward_zero_std": 0.375, "grad_norm": 18.667165756225586, "learning_rate": 1.2658926150792322e-07, "loss": -0.0, "num_tokens": 9306140.0, "reward": 0.5335937142372131, "reward_std": 0.2907790541648865, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1508.0, "completions/max_terminated_length": 1508.0, "completions/mean_length": 101.40625, "completions/mean_terminated_length": 101.40625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2, "frac_reward_zero_std": 0.625, "grad_norm": 13.560271263122559, "learning_rate": 1.1579944457157059e-07, "loss": 0.0, "num_tokens": 9421958.0, "reward": 0.5031249523162842, "reward_std": 0.1530819833278656, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 990.0, "completions/max_terminated_length": 990.0, "completions/mean_length": 140.671875, "completions/mean_terminated_length": 140.671875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2025, "frac_reward_zero_std": 0.5, "grad_norm": 22.05609130859375, "learning_rate": 1.0542974530180327e-07, "loss": -0.0, "num_tokens": 9532593.0, "reward": 0.20624998211860657, "reward_std": 0.23356688022613525, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.15625, "rewards/mcq_exact_match_reward/std": 0.36596253514289856, "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": 999.0, "completions/max_terminated_length": 999.0, "completions/mean_length": 110.5, "completions/mean_terminated_length": 110.5, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.205, "frac_reward_zero_std": 0.25, "grad_norm": 26.485273361206055, "learning_rate": 9.549150281252632e-08, "loss": -0.0, "num_tokens": 9637369.0, "reward": 0.2523437440395355, "reward_std": 0.342986524105072, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 0.0, "completions/max_length": 763.0, "completions/max_terminated_length": 763.0, "completions/mean_length": 70.328125, "completions/mean_terminated_length": 70.328125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2075, "frac_reward_zero_std": 0.5, "grad_norm": 20.73296546936035, "learning_rate": 8.599558442598998e-08, "loss": -0.0, "num_tokens": 9723070.0, "reward": 0.5499999523162842, "reward_std": 0.21306739747524261, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 952.0, "completions/max_terminated_length": 952.0, "completions/mean_length": 101.515625, "completions/mean_terminated_length": 101.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.21, "frac_reward_zero_std": 0.375, "grad_norm": 17.785524368286133, "learning_rate": 7.695237378953224e-08, "loss": -0.0, "num_tokens": 9830151.0, "reward": 0.5499999523162842, "reward_std": 0.28247907757759094, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 1695.0, "completions/max_terminated_length": 1695.0, "completions/mean_length": 325.59375, "completions/mean_terminated_length": 325.59375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2125, "frac_reward_zero_std": 0.25, "grad_norm": 10.511614799499512, "learning_rate": 6.837175952121304e-08, "loss": -0.0, "num_tokens": 9943389.0, "reward": 0.36171871423721313, "reward_std": 0.3338724970817566, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 1268.0, "completions/mean_length": 182.234375, "completions/mean_terminated_length": 152.61904907226562, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.215, "frac_reward_zero_std": 0.5, "grad_norm": 26.823604583740234, "learning_rate": 6.026312439675551e-08, "loss": -0.0, "num_tokens": 10042540.0, "reward": 0.4078124761581421, "reward_std": 0.24163225293159485, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1333.0, "completions/mean_length": 167.09375, "completions/mean_terminated_length": 137.23809814453125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2175, "frac_reward_zero_std": 0.375, "grad_norm": 27.927860260009766, "learning_rate": 5.263533508961826e-08, "loss": 0.0, "num_tokens": 10154290.0, "reward": 0.5179687142372131, "reward_std": 0.2404848039150238, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 972.0, "completions/max_terminated_length": 972.0, "completions/mean_length": 132.875, "completions/mean_terminated_length": 132.875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.22, "frac_reward_zero_std": 0.5, "grad_norm": 17.77516746520996, "learning_rate": 4.549673247541874e-08, "loss": -0.0, "num_tokens": 10260362.0, "reward": 0.5031249523162842, "reward_std": 0.24039675295352936, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 567.0, "completions/max_terminated_length": 567.0, "completions/mean_length": 20.546875, "completions/mean_terminated_length": 20.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2225, "frac_reward_zero_std": 0.375, "grad_norm": 46.047019958496094, "learning_rate": 3.8855122511307626e-08, "loss": -0.0, "num_tokens": 10372357.0, "reward": 0.36249998211860657, "reward_std": 0.29143062233924866, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 1015.0, "completions/max_terminated_length": 1015.0, "completions/mean_length": 125.6875, "completions/mean_terminated_length": 125.6875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.225, "frac_reward_zero_std": 0.375, "grad_norm": 18.907434463500977, "learning_rate": 3.271776770026963e-08, "loss": -0.0, "num_tokens": 10490801.0, "reward": 0.2828124761581421, "reward_std": 0.2582891583442688, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 679.0, "completions/max_terminated_length": 679.0, "completions/mean_length": 71.125, "completions/mean_terminated_length": 71.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2275, "frac_reward_zero_std": 0.75, "grad_norm": 8.110417366027832, "learning_rate": 2.7091379149682682e-08, "loss": -0.0, "num_tokens": 10560777.0, "reward": 0.7062499523162842, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.65625, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1576.0, "completions/mean_length": 308.875, "completions/mean_terminated_length": 281.2698669433594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.23, "frac_reward_zero_std": 0.5, "grad_norm": 11.587091445922852, "learning_rate": 2.1982109232821176e-08, "loss": -0.0, "num_tokens": 10677057.0, "reward": 0.48671871423721313, "reward_std": 0.213933527469635, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "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": 898.0, "completions/max_terminated_length": 898.0, "completions/mean_length": 149.921875, "completions/mean_terminated_length": 149.921875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2325, "frac_reward_zero_std": 0.5, "grad_norm": 10.026474952697754, "learning_rate": 1.7395544861325718e-08, "loss": -0.0, "num_tokens": 10772268.0, "reward": 0.33046871423721313, "reward_std": 0.2251373678445816, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1856.0, "completions/max_terminated_length": 1856.0, "completions/mean_length": 172.03125, "completions/mean_terminated_length": 172.03125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.235, "frac_reward_zero_std": 0.625, "grad_norm": 15.831273078918457, "learning_rate": 1.3336701375997127e-08, "loss": 0.0, "num_tokens": 10886566.0, "reward": 0.5632811784744263, "reward_std": 0.10529523342847824, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1191.0, "completions/max_terminated_length": 1191.0, "completions/mean_length": 83.21875, "completions/mean_terminated_length": 83.21875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2375, "frac_reward_zero_std": 0.75, "grad_norm": 10.345685958862305, "learning_rate": 9.810017062595321e-09, "loss": -0.0, "num_tokens": 10990484.0, "reward": 0.4874999523162842, "reward_std": 0.10888781398534775, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 1230.0, "completions/max_terminated_length": 1230.0, "completions/mean_length": 198.84375, "completions/mean_terminated_length": 198.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.24, "frac_reward_zero_std": 0.5, "grad_norm": 14.650867462158203, "learning_rate": 6.819348298638839e-09, "loss": -0.0, "num_tokens": 11113074.0, "reward": 0.14218749105930328, "reward_std": 0.20483601093292236, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.09375, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "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": 1262.0, "completions/max_terminated_length": 1262.0, "completions/mean_length": 155.1875, "completions/mean_terminated_length": 155.1875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2425, "frac_reward_zero_std": 0.25, "grad_norm": 25.82438087463379, "learning_rate": 4.367965336512403e-09, "loss": -0.0, "num_tokens": 11221478.0, "reward": 0.453125, "reward_std": 0.3423337936401367, "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": 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": 1588.0, "completions/max_terminated_length": 1588.0, "completions/mean_length": 184.015625, "completions/mean_terminated_length": 184.015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.245, "frac_reward_zero_std": 0.5, "grad_norm": 10.68353271484375, "learning_rate": 2.458548727494292e-09, "loss": 0.0, "num_tokens": 11354471.0, "reward": 0.36171871423721313, "reward_std": 0.14846178889274597, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 1418.0, "completions/max_terminated_length": 1418.0, "completions/mean_length": 279.609375, "completions/mean_terminated_length": 279.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2475, "frac_reward_zero_std": 0.25, "grad_norm": 17.142242431640625, "learning_rate": 1.0931863906127325e-09, "loss": -0.0, "num_tokens": 11464358.0, "reward": 0.4249999523162842, "reward_std": 0.36295413970947266, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 963.0, "completions/max_terminated_length": 963.0, "completions/mean_length": 146.21875, "completions/mean_terminated_length": 146.21875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.25, "frac_reward_zero_std": 0.75, "grad_norm": 2.886613368988037, "learning_rate": 2.733713295369755e-10, "loss": 0.0, "num_tokens": 11565124.0, "reward": 0.6742187142372131, "reward_std": 0.0887577086687088, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 100 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 11565124, "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 }