{ "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": 1464.0, "completions/mean_length": 477.234375, "completions/mean_terminated_length": 426.56451416015625, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.0025, "frac_reward_zero_std": 0.0, "grad_norm": 9.820798873901367, "learning_rate": 0.0, "loss": 0.0, "num_tokens": 128463.0, "reward": 0.30078125, "reward_std": 0.2949070334434509, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.3294980227947235, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 1 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1324.0, "completions/max_terminated_length": 1324.0, "completions/mean_length": 549.921875, "completions/mean_terminated_length": 549.921875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.005, "frac_reward_zero_std": 0.0, "grad_norm": 5.992785453796387, "learning_rate": 2e-07, "loss": -0.0, "num_tokens": 224890.0, "reward": 0.2890625, "reward_std": 0.39714252948760986, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.301698237657547, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 1812.0, "completions/mean_length": 556.71875, "completions/mean_terminated_length": 533.0476684570312, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0075, "frac_reward_zero_std": 0.0, "grad_norm": 6.250213146209717, "learning_rate": 4e-07, "loss": -0.0, "num_tokens": 361720.0, "reward": 0.1874999850988388, "reward_std": 0.3287465274333954, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.2597312331199646, "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": 2010.0, "completions/max_terminated_length": 2010.0, "completions/mean_length": 388.546875, "completions/mean_terminated_length": 388.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.01, "frac_reward_zero_std": 0.0, "grad_norm": 7.673406600952148, "learning_rate": 6e-07, "loss": -0.0, "num_tokens": 469803.0, "reward": 0.41718748211860657, "reward_std": 0.42704811692237854, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.23935678601264954, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 1047.0, "completions/max_terminated_length": 1047.0, "completions/mean_length": 299.203125, "completions/mean_terminated_length": 299.203125, "completions/min_length": 5.0, "completions/min_terminated_length": 5.0, "epoch": 0.0125, "frac_reward_zero_std": 0.0, "grad_norm": 9.532109260559082, "learning_rate": 8e-07, "loss": 0.0, "num_tokens": 581328.0, "reward": 0.32656246423721313, "reward_std": 0.25986582040786743, "rewards/format_reward/mean": 0.296875, "rewards/format_reward/std": 0.3177144229412079, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 1405.0, "completions/max_terminated_length": 1405.0, "completions/mean_length": 549.515625, "completions/mean_terminated_length": 549.515625, "completions/min_length": 19.0, "completions/min_terminated_length": 19.0, "epoch": 0.015, "frac_reward_zero_std": 0.0, "grad_norm": 6.551382064819336, "learning_rate": 1e-06, "loss": 0.0, "num_tokens": 695497.0, "reward": 0.25078123807907104, "reward_std": 0.3316580057144165, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.28770697116851807, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1751.0, "completions/mean_length": 496.078125, "completions/mean_terminated_length": 446.01611328125, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.0175, "frac_reward_zero_std": 0.0, "grad_norm": 14.735747337341309, "learning_rate": 9.99726628670463e-07, "loss": 0.0, "num_tokens": 819054.0, "reward": 0.41015625, "reward_std": 0.42701074481010437, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.24688033759593964, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 1369.0, "completions/max_terminated_length": 1369.0, "completions/mean_length": 461.125, "completions/mean_terminated_length": 461.125, "completions/min_length": 8.0, "completions/min_terminated_length": 8.0, "epoch": 0.02, "frac_reward_zero_std": 0.125, "grad_norm": 2.3624932765960693, "learning_rate": 9.989068136093872e-07, "loss": 0.0, "num_tokens": 918694.0, "reward": 0.42656248807907104, "reward_std": 0.3252020478248596, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.2741328477859497, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1707.0, "completions/mean_length": 575.140625, "completions/mean_terminated_length": 502.70489501953125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0225, "frac_reward_zero_std": 0.0, "grad_norm": 5.942022800445557, "learning_rate": 9.975414512725056e-07, "loss": -0.0, "num_tokens": 1059111.0, "reward": 0.32343748211860657, "reward_std": 0.38642236590385437, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.18298126757144928, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 9 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1615.0, "completions/mean_length": 552.0625, "completions/mean_terminated_length": 528.3175048828125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.025, "frac_reward_zero_std": 0.0, "grad_norm": 10.24199390411377, "learning_rate": 9.956320346634875e-07, "loss": -0.0, "num_tokens": 1199859.0, "reward": 0.29140621423721313, "reward_std": 0.28808674216270447, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.209963858127594, "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": 971.0, "completions/mean_length": 323.359375, "completions/mean_terminated_length": 295.984130859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0275, "frac_reward_zero_std": 0.125, "grad_norm": 11.983031272888184, "learning_rate": 9.931806517013612e-07, "loss": -0.0, "num_tokens": 1308242.0, "reward": 0.24374999105930328, "reward_std": 0.26579102873802185, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.19669894874095917, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 1475.0, "completions/max_terminated_length": 1475.0, "completions/mean_length": 494.640625, "completions/mean_terminated_length": 494.640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03, "frac_reward_zero_std": 0.125, "grad_norm": 13.534098625183105, "learning_rate": 9.901899829374047e-07, "loss": -0.0, "num_tokens": 1427139.0, "reward": 0.453125, "reward_std": 0.37656593322753906, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.21593283116817474, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 1420.0, "completions/max_terminated_length": 1420.0, "completions/mean_length": 591.890625, "completions/mean_terminated_length": 591.890625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0325, "frac_reward_zero_std": 0.125, "grad_norm": 3.6356875896453857, "learning_rate": 9.866632986240029e-07, "loss": 0.0, "num_tokens": 1555604.0, "reward": 0.24765624105930328, "reward_std": 0.3406350612640381, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.2538151443004608, "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": 0.0, "completions/max_length": 1100.0, "completions/max_terminated_length": 1100.0, "completions/mean_length": 327.9375, "completions/mean_terminated_length": 327.9375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.035, "frac_reward_zero_std": 0.0, "grad_norm": 19.802650451660156, "learning_rate": 9.826044551386742e-07, "loss": 0.0, "num_tokens": 1663176.0, "reward": 0.30781248211860657, "reward_std": 0.3470980226993561, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.20351573824882507, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1263.0, "completions/mean_length": 416.265625, "completions/mean_terminated_length": 390.3651123046875, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.0375, "frac_reward_zero_std": 0.0, "grad_norm": 10.966890335083008, "learning_rate": 9.780178907671788e-07, "loss": -0.0, "num_tokens": 1757385.0, "reward": 0.3671875, "reward_std": 0.37327155470848083, "rewards/format_reward/mean": 0.546875, "rewards/format_reward/std": 0.3299681544303894, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1576.0, "completions/mean_length": 447.671875, "completions/mean_terminated_length": 422.2698669433594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04, "frac_reward_zero_std": 0.125, "grad_norm": 10.119935989379883, "learning_rate": 9.729086208503173e-07, "loss": -0.0, "num_tokens": 1892308.0, "reward": 0.5914062261581421, "reward_std": 0.31337296962738037, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.546875, "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": 0.0, "completions/max_length": 1743.0, "completions/max_terminated_length": 1743.0, "completions/mean_length": 421.59375, "completions/mean_terminated_length": 421.59375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0425, "frac_reward_zero_std": 0.125, "grad_norm": 10.146256446838379, "learning_rate": 9.672822322997304e-07, "loss": 0.0, "num_tokens": 2011562.0, "reward": 0.29140621423721313, "reward_std": 0.3035487234592438, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.209963858127594, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 17 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1499.0, "completions/max_terminated_length": 1499.0, "completions/mean_length": 368.15625, "completions/mean_terminated_length": 368.15625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.045, "frac_reward_zero_std": 0.0, "grad_norm": 28.832923889160156, "learning_rate": 9.611448774886923e-07, "loss": 0.0, "num_tokens": 2107684.0, "reward": 0.31328123807907104, "reward_std": 0.32075032591819763, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.22589658200740814, "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": 1508.0, "completions/max_terminated_length": 1508.0, "completions/mean_length": 311.4375, "completions/mean_terminated_length": 311.4375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0475, "frac_reward_zero_std": 0.0, "grad_norm": 17.548686981201172, "learning_rate": 9.545032675245813e-07, "loss": -0.0, "num_tokens": 2220144.0, "reward": 0.5015624761581421, "reward_std": 0.41764065623283386, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1022.0, "completions/mean_length": 172.484375, "completions/mean_terminated_length": 142.71429443359375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05, "frac_reward_zero_std": 0.5, "grad_norm": 16.76396942138672, "learning_rate": 9.473646649103817e-07, "loss": 0.0, "num_tokens": 2318791.0, "reward": 0.5539062023162842, "reward_std": 0.17947588860988617, "rewards/format_reward/mean": 0.5390625, "rewards/format_reward/std": 0.16194961965084076, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 20 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1462.0, "completions/max_terminated_length": 1462.0, "completions/mean_length": 254.03125, "completions/mean_terminated_length": 254.03125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0525, "frac_reward_zero_std": 0.25, "grad_norm": 12.80144214630127, "learning_rate": 9.397368756032444e-07, "loss": -0.0, "num_tokens": 2429865.0, "reward": 0.25312498211860657, "reward_std": 0.33632034063339233, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.1259881556034088, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 954.0, "completions/max_terminated_length": 954.0, "completions/mean_length": 104.65625, "completions/mean_terminated_length": 104.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.055, "frac_reward_zero_std": 0.625, "grad_norm": 10.90204906463623, "learning_rate": 9.316282404787869e-07, "loss": -0.0, "num_tokens": 2523307.0, "reward": 0.4093749523162842, "reward_std": 0.17358146607875824, "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": 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": 1322.0, "completions/max_terminated_length": 1322.0, "completions/mean_length": 377.9375, "completions/mean_terminated_length": 377.9375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0575, "frac_reward_zero_std": 0.25, "grad_norm": 11.154988288879395, "learning_rate": 9.230476262104676e-07, "loss": -0.0, "num_tokens": 2626679.0, "reward": 0.22343748807907104, "reward_std": 0.28182199597358704, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.1534975916147232, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 398.0, "completions/max_terminated_length": 398.0, "completions/mean_length": 12.125, "completions/mean_terminated_length": 12.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06, "frac_reward_zero_std": 0.375, "grad_norm": 11.924612045288086, "learning_rate": 9.1400441557401e-07, "loss": 0.0, "num_tokens": 2733791.0, "reward": 0.4718749523162842, "reward_std": 0.24831001460552216, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 979.0, "completions/max_terminated_length": 979.0, "completions/mean_length": 173.625, "completions/mean_terminated_length": 173.625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0625, "frac_reward_zero_std": 0.375, "grad_norm": 17.104806900024414, "learning_rate": 9.045084971874737e-07, "loss": -0.0, "num_tokens": 2827215.0, "reward": 0.51953125, "reward_std": 0.22175219655036926, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.1406387835741043, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 1075.0, "completions/max_terminated_length": 1075.0, "completions/mean_length": 161.796875, "completions/mean_terminated_length": 161.796875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.065, "frac_reward_zero_std": 0.375, "grad_norm": 24.93585968017578, "learning_rate": 8.945702546981968e-07, "loss": 0.0, "num_tokens": 2902970.0, "reward": 0.40859371423721313, "reward_std": 0.22210699319839478, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 928.0, "completions/max_terminated_length": 928.0, "completions/mean_length": 235.765625, "completions/mean_terminated_length": 235.765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0675, "frac_reward_zero_std": 0.25, "grad_norm": 16.02617645263672, "learning_rate": 8.842005554284295e-07, "loss": -0.0, "num_tokens": 3005379.0, "reward": 0.29999998211860657, "reward_std": 0.2879316806793213, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.08908708393573761, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 27 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 935.0, "completions/max_terminated_length": 935.0, "completions/mean_length": 141.0, "completions/mean_terminated_length": 141.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07, "frac_reward_zero_std": 0.125, "grad_norm": 29.071575164794922, "learning_rate": 8.734107384920769e-07, "loss": -0.0, "num_tokens": 3102595.0, "reward": 0.4562499523162842, "reward_std": 0.3846532702445984, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.08908708393573761, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 961.0, "completions/max_terminated_length": 961.0, "completions/mean_length": 109.046875, "completions/mean_terminated_length": 109.046875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0725, "frac_reward_zero_std": 0.5, "grad_norm": 14.398436546325684, "learning_rate": 8.622126023955445e-07, "loss": -0.0, "num_tokens": 3190334.0, "reward": 0.6898437142372131, "reward_std": 0.19096830487251282, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 29 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 2034.0, "completions/max_terminated_length": 2034.0, "completions/mean_length": 313.390625, "completions/mean_terminated_length": 313.390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.075, "frac_reward_zero_std": 0.25, "grad_norm": 23.27433967590332, "learning_rate": 8.506183921362442e-07, "loss": 0.0, "num_tokens": 3301191.0, "reward": 0.6148437261581421, "reward_std": 0.3767889142036438, "rewards/format_reward/mean": 0.5234375, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.5625, "rewards/mcq_exact_match_reward/std": 0.5, "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": 1409.0, "completions/max_terminated_length": 1409.0, "completions/mean_length": 302.03125, "completions/mean_terminated_length": 302.03125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0775, "frac_reward_zero_std": 0.375, "grad_norm": 9.451353073120117, "learning_rate": 8.386407858128706e-07, "loss": -0.0, "num_tokens": 3431513.0, "reward": 0.3273437023162842, "reward_std": 0.24055621027946472, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.16194961965084076, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 1716.0, "completions/max_terminated_length": 1716.0, "completions/mean_length": 140.75, "completions/mean_terminated_length": 140.75, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08, "frac_reward_zero_std": 0.5, "grad_norm": 19.545513153076172, "learning_rate": 8.262928807620843e-07, "loss": -0.0, "num_tokens": 3516601.0, "reward": 0.42890626192092896, "reward_std": 0.0956839770078659, "rewards/format_reward/mean": 0.5390625, "rewards/format_reward/std": 0.2236899733543396, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 333.0, "completions/max_terminated_length": 333.0, "completions/mean_length": 26.1875, "completions/mean_terminated_length": 26.1875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0825, "frac_reward_zero_std": 0.5, "grad_norm": 11.879921913146973, "learning_rate": 8.135881792367685e-07, "loss": 0.0, "num_tokens": 3610013.0, "reward": 0.7226561903953552, "reward_std": 0.19158241152763367, "rewards/format_reward/mean": 0.5078125, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.671875, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 1395.0, "completions/max_terminated_length": 1395.0, "completions/mean_length": 130.375, "completions/mean_terminated_length": 130.375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.085, "frac_reward_zero_std": 0.25, "grad_norm": 22.375093460083008, "learning_rate": 8.005405736415125e-07, "loss": 0.0, "num_tokens": 3706909.0, "reward": 0.4156249761581421, "reward_std": 0.34389790892601013, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.18898223340511322, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1169.0, "completions/mean_length": 228.734375, "completions/mean_terminated_length": 170.0483856201172, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0875, "frac_reward_zero_std": 0.25, "grad_norm": 17.84543228149414, "learning_rate": 7.871643313414718e-07, "loss": -0.0, "num_tokens": 3815884.0, "reward": 0.7546874284744263, "reward_std": 0.28800931572914124, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.1534975916147232, "rewards/mcq_exact_match_reward/mean": 0.703125, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 311.0, "completions/max_terminated_length": 311.0, "completions/mean_length": 27.609375, "completions/mean_terminated_length": 27.609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09, "frac_reward_zero_std": 0.25, "grad_norm": 31.98973274230957, "learning_rate": 7.734740790612136e-07, "loss": -0.0, "num_tokens": 3941163.0, "reward": 0.4898437261581421, "reward_std": 0.2993735373020172, "rewards/format_reward/mean": 0.5234375, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 1835.0, "completions/max_terminated_length": 1835.0, "completions/mean_length": 61.625, "completions/mean_terminated_length": 61.625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0925, "frac_reward_zero_std": 0.5, "grad_norm": 12.227835655212402, "learning_rate": 7.594847868906076e-07, "loss": 0.0, "num_tokens": 4027779.0, "reward": 0.528124988079071, "reward_std": 0.1930253505706787, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1583.0, "completions/max_terminated_length": 1583.0, "completions/mean_length": 239.546875, "completions/mean_terminated_length": 239.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.095, "frac_reward_zero_std": 0.25, "grad_norm": 20.220304489135742, "learning_rate": 7.452117519152541e-07, "loss": 0.0, "num_tokens": 4120606.0, "reward": 0.39531248807907104, "reward_std": 0.3611350357532501, "rewards/format_reward/mean": 0.671875, "rewards/format_reward/std": 0.2847827076911926, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 38 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 926.0, "completions/max_terminated_length": 926.0, "completions/mean_length": 72.34375, "completions/mean_terminated_length": 72.34375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0975, "frac_reward_zero_std": 0.5, "grad_norm": 21.585006713867188, "learning_rate": 7.306705814893439e-07, "loss": -0.0, "num_tokens": 4210956.0, "reward": 0.37812501192092896, "reward_std": 0.23370197415351868, "rewards/format_reward/mean": 0.65625, "rewards/format_reward/std": 0.233588308095932, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 39 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1230.0, "completions/max_terminated_length": 1230.0, "completions/mean_length": 107.265625, "completions/mean_terminated_length": 107.265625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1, "frac_reward_zero_std": 0.375, "grad_norm": 19.986061096191406, "learning_rate": 7.158771761692464e-07, "loss": 0.0, "num_tokens": 4302349.0, "reward": 0.28593748807907104, "reward_std": 0.20377102494239807, "rewards/format_reward/mean": 0.671875, "rewards/format_reward/std": 0.23935678601264954, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 784.0, "completions/max_terminated_length": 784.0, "completions/mean_length": 170.578125, "completions/mean_terminated_length": 170.578125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1025, "frac_reward_zero_std": 0.25, "grad_norm": 23.84389305114746, "learning_rate": 7.008477123264847e-07, "loss": 0.0, "num_tokens": 4402810.0, "reward": 0.6890624761581421, "reward_std": 0.3438800573348999, "rewards/format_reward/mean": 0.640625, "rewards/format_reward/std": 0.24346621334552765, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1099.0, "completions/mean_length": 197.78125, "completions/mean_terminated_length": 106.78688049316406, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.105, "frac_reward_zero_std": 0.25, "grad_norm": 30.521644592285156, "learning_rate": 6.855986244591103e-07, "loss": -0.0, "num_tokens": 4512004.0, "reward": 0.5382812023162842, "reward_std": 0.25127214193344116, "rewards/format_reward/mean": 0.6953125, "rewards/format_reward/std": 0.31644338369369507, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 42 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1161.0, "completions/mean_length": 134.6875, "completions/mean_terminated_length": 104.31746673583984, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1075, "frac_reward_zero_std": 0.125, "grad_norm": 28.4834041595459, "learning_rate": 6.701465872208216e-07, "loss": 0.0, "num_tokens": 4623384.0, "reward": 0.2992187440395355, "reward_std": 0.23480820655822754, "rewards/format_reward/mean": 0.8046875, "rewards/format_reward/std": 0.29028159379959106, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 715.0, "completions/max_terminated_length": 715.0, "completions/mean_length": 76.203125, "completions/mean_terminated_length": 76.203125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.11, "frac_reward_zero_std": 0.25, "grad_norm": 14.868675231933594, "learning_rate": 6.545084971874736e-07, "loss": 0.0, "num_tokens": 4706005.0, "reward": 0.32734376192092896, "reward_std": 0.24423527717590332, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.1751912236213684, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 668.0, "completions/mean_length": 154.5625, "completions/mean_terminated_length": 61.44261932373047, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1125, "frac_reward_zero_std": 0.25, "grad_norm": 15.19470500946045, "learning_rate": 6.387014543809223e-07, "loss": 0.0, "num_tokens": 4798985.0, "reward": 0.5421874523162842, "reward_std": 0.3319449722766876, "rewards/format_reward/mean": 0.734375, "rewards/format_reward/std": 0.30820462107658386, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 543.0, "completions/mean_length": 72.453125, "completions/mean_terminated_length": 41.09524154663086, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.115, "frac_reward_zero_std": 0.5, "grad_norm": 15.99255084991455, "learning_rate": 6.227427435703995e-07, "loss": -0.0, "num_tokens": 4918510.0, "reward": 0.31406253576278687, "reward_std": 0.0530330091714859, "rewards/format_reward/mean": 0.796875, "rewards/format_reward/std": 0.3642643094062805, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 608.0, "completions/mean_length": 79.3125, "completions/mean_terminated_length": 48.06349563598633, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1175, "frac_reward_zero_std": 0.25, "grad_norm": 18.731386184692383, "learning_rate": 6.066498153718734e-07, "loss": 0.0, "num_tokens": 5018826.0, "reward": 0.6578124761581421, "reward_std": 0.29044055938720703, "rewards/format_reward/mean": 0.796875, "rewards/format_reward/std": 0.3177144229412079, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 855.0, "completions/max_terminated_length": 855.0, "completions/mean_length": 66.71875, "completions/mean_terminated_length": 66.71875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.12, "frac_reward_zero_std": 0.25, "grad_norm": 18.806800842285156, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 5152376.0, "reward": 0.3304687738418579, "reward_std": 0.30621567368507385, "rewards/format_reward/mean": 0.8046875, "rewards/format_reward/std": 0.37392371892929077, "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": 774.0, "completions/mean_length": 102.953125, "completions/mean_terminated_length": 72.0793685913086, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1225, "frac_reward_zero_std": 0.25, "grad_norm": 34.07569885253906, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 5259269.0, "reward": 0.45234376192092896, "reward_std": 0.2820115089416504, "rewards/format_reward/mean": 0.7734375, "rewards/format_reward/std": 0.2807259261608124, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 414.0, "completions/mean_length": 127.9375, "completions/mean_terminated_length": 66.0, "completions/min_length": 11.0, "completions/min_terminated_length": 11.0, "epoch": 0.125, "frac_reward_zero_std": 0.25, "grad_norm": 27.46225929260254, "learning_rate": 5.577423184847931e-07, "loss": 0.0, "num_tokens": 5354257.0, "reward": 0.592968761920929, "reward_std": 0.22307650744915009, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.23345555365085602, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 50 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1015.0, "completions/mean_length": 120.328125, "completions/mean_terminated_length": 89.73016357421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1275, "frac_reward_zero_std": 0.125, "grad_norm": 18.311309814453125, "learning_rate": 5.412896727361662e-07, "loss": 0.0, "num_tokens": 5452862.0, "reward": 0.510937511920929, "reward_std": 0.3758324980735779, "rewards/format_reward/mean": 0.890625, "rewards/format_reward/std": 0.2592533528804779, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 51 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1425.0, "completions/mean_length": 101.140625, "completions/mean_terminated_length": 70.23809814453125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.13, "frac_reward_zero_std": 0.25, "grad_norm": 15.757895469665527, "learning_rate": 5.247918773366111e-07, "loss": -0.0, "num_tokens": 5554471.0, "reward": 0.43046876788139343, "reward_std": 0.20688433945178986, "rewards/format_reward/mean": 0.8671875, "rewards/format_reward/std": 0.28510910272598267, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1296.0, "completions/max_terminated_length": 1296.0, "completions/mean_length": 67.234375, "completions/mean_terminated_length": 67.234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1325, "frac_reward_zero_std": 0.5, "grad_norm": 11.062511444091797, "learning_rate": 5.082669723831793e-07, "loss": -0.0, "num_tokens": 5651566.0, "reward": 0.42109376192092896, "reward_std": 0.17108294367790222, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.1751912236213684, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 0.0625, "completions/max_length": 2048.0, "completions/max_terminated_length": 447.0, "completions/mean_length": 153.5, "completions/mean_terminated_length": 27.200000762939453, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.135, "frac_reward_zero_std": 0.125, "grad_norm": 16.75969123840332, "learning_rate": 4.917330276168208e-07, "loss": 0.0, "num_tokens": 5753262.0, "reward": 0.6187499761581421, "reward_std": 0.307129442691803, "rewards/format_reward/mean": 0.875, "rewards/format_reward/std": 0.28171807527542114, "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": 0.0, "completions/max_length": 462.0, "completions/max_terminated_length": 462.0, "completions/mean_length": 48.203125, "completions/mean_terminated_length": 48.203125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1375, "frac_reward_zero_std": 0.5, "grad_norm": 14.472973823547363, "learning_rate": 4.752081226633888e-07, "loss": -0.0, "num_tokens": 5829763.0, "reward": 0.45390626788139343, "reward_std": 0.1621313989162445, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.15728822350502014, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 488.0, "completions/mean_length": 70.390625, "completions/mean_terminated_length": 39.000003814697266, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.14, "frac_reward_zero_std": 0.5, "grad_norm": 9.199593544006348, "learning_rate": 4.5871032726383385e-07, "loss": 0.0, "num_tokens": 5927572.0, "reward": 0.18906250596046448, "reward_std": 0.11857090145349503, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.21304203569889069, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 872.0, "completions/mean_length": 107.125, "completions/mean_terminated_length": 44.51612854003906, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1425, "frac_reward_zero_std": 0.25, "grad_norm": 14.753751754760742, "learning_rate": 4.4225768151520694e-07, "loss": 0.0, "num_tokens": 6032756.0, "reward": 0.45000001788139343, "reward_std": 0.28659987449645996, "rewards/format_reward/mean": 0.90625, "rewards/format_reward/std": 0.233588308095932, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 553.0, "completions/max_terminated_length": 553.0, "completions/mean_length": 36.59375, "completions/mean_terminated_length": 36.59375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.145, "frac_reward_zero_std": 0.5, "grad_norm": 18.922006607055664, "learning_rate": 4.258681761440789e-07, "loss": 0.0, "num_tokens": 6120570.0, "reward": 0.37968751788139343, "reward_std": 0.24809977412223816, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 58 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 667.0, "completions/mean_length": 143.96875, "completions/mean_terminated_length": 50.32786560058594, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1475, "frac_reward_zero_std": 0.5, "grad_norm": 6.868103504180908, "learning_rate": 4.095597328339452e-07, "loss": -0.0, "num_tokens": 6236432.0, "reward": 0.5914062261581421, "reward_std": 0.20316563546657562, "rewards/format_reward/mean": 0.9140625, "rewards/format_reward/std": 0.27537402510643005, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 407.0, "completions/max_terminated_length": 407.0, "completions/mean_length": 21.84375, "completions/mean_terminated_length": 21.84375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.15, "frac_reward_zero_std": 0.375, "grad_norm": 15.219839096069336, "learning_rate": 3.9335018462812664e-07, "loss": -0.0, "num_tokens": 6327390.0, "reward": 0.7085937857627869, "reward_std": 0.2691788673400879, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 164.0, "completions/mean_length": 80.140625, "completions/mean_terminated_length": 16.66128921508789, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.1525, "frac_reward_zero_std": 0.375, "grad_norm": 10.40705394744873, "learning_rate": 3.772572564296004e-07, "loss": 0.0, "num_tokens": 6440111.0, "reward": 0.5804687738418579, "reward_std": 0.24087271094322205, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 366.0, "completions/mean_length": 58.875, "completions/mean_terminated_length": 27.30158805847168, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.155, "frac_reward_zero_std": 0.375, "grad_norm": 16.986087799072266, "learning_rate": 3.612985456190778e-07, "loss": -0.0, "num_tokens": 6518671.0, "reward": 0.7054687738418579, "reward_std": 0.26613086462020874, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.18483558297157288, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 460.0, "completions/mean_length": 122.21875, "completions/mean_terminated_length": 27.508195877075195, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.1575, "frac_reward_zero_std": 0.375, "grad_norm": 16.218698501586914, "learning_rate": 3.454915028125263e-07, "loss": -0.0, "num_tokens": 6611805.0, "reward": 0.561718761920929, "reward_std": 0.14248578250408173, "rewards/format_reward/mean": 0.9296875, "rewards/format_reward/std": 0.2498759627342224, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 876.0, "completions/max_terminated_length": 876.0, "completions/mean_length": 53.1875, "completions/mean_terminated_length": 53.1875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.16, "frac_reward_zero_std": 0.75, "grad_norm": 12.511017799377441, "learning_rate": 3.2985341277917846e-07, "loss": 0.0, "num_tokens": 6698345.0, "reward": 0.32890626788139343, "reward_std": 0.04640388861298561, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.15728822350502014, "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": 0.0, "completions/max_length": 20.0, "completions/max_terminated_length": 20.0, "completions/mean_length": 13.28125, "completions/mean_terminated_length": 13.28125, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.1625, "frac_reward_zero_std": 0.875, "grad_norm": 4.530981063842773, "learning_rate": 3.1440137554088953e-07, "loss": 0.0, "num_tokens": 6789451.0, "reward": 0.3968750238418579, "reward_std": 0.0646936446428299, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 65 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1223.0, "completions/mean_length": 134.359375, "completions/mean_terminated_length": 72.6290283203125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.165, "frac_reward_zero_std": 0.5, "grad_norm": 16.526830673217773, "learning_rate": 2.9915228767351535e-07, "loss": 0.0, "num_tokens": 6907946.0, "reward": 0.48593753576278687, "reward_std": 0.18089531362056732, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.19352105259895325, "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": 481.0, "completions/max_terminated_length": 481.0, "completions/mean_length": 35.515625, "completions/mean_terminated_length": 35.515625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.1675, "frac_reward_zero_std": 0.5, "grad_norm": 9.258106231689453, "learning_rate": 2.841228238307536e-07, "loss": -0.0, "num_tokens": 6978851.0, "reward": 0.8031250238418579, "reward_std": 0.19044628739356995, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.703125, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 67 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 618.0, "completions/mean_length": 63.296875, "completions/mean_terminated_length": 31.79365348815918, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.17, "frac_reward_zero_std": 0.25, "grad_norm": 18.417194366455078, "learning_rate": 2.6932941851065615e-07, "loss": 0.0, "num_tokens": 7080566.0, "reward": 0.6421874761581421, "reward_std": 0.29052332043647766, "rewards/format_reward/mean": 0.953125, "rewards/format_reward/std": 0.21304203569889069, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 550.0, "completions/max_terminated_length": 550.0, "completions/mean_length": 33.65625, "completions/mean_terminated_length": 33.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1725, "frac_reward_zero_std": 0.125, "grad_norm": 28.758838653564453, "learning_rate": 2.547882480847461e-07, "loss": 0.0, "num_tokens": 7180000.0, "reward": 0.534375011920929, "reward_std": 0.33078908920288086, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.12198751419782639, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 480.0, "completions/max_terminated_length": 480.0, "completions/mean_length": 20.453125, "completions/mean_terminated_length": 20.453125, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.175, "frac_reward_zero_std": 0.625, "grad_norm": 11.625492095947266, "learning_rate": 2.4051521310939254e-07, "loss": 0.0, "num_tokens": 7277581.0, "reward": 0.5843750238418579, "reward_std": 0.15992169082164764, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 192.0, "completions/max_terminated_length": 192.0, "completions/mean_length": 18.796875, "completions/mean_terminated_length": 18.796875, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.1775, "frac_reward_zero_std": 0.625, "grad_norm": 8.98841667175293, "learning_rate": 2.2652592093878665e-07, "loss": 0.0, "num_tokens": 7354840.0, "reward": 0.41093751788139343, "reward_std": 0.11330723762512207, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 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": 336.0, "completions/max_terminated_length": 336.0, "completions/mean_length": 20.78125, "completions/mean_terminated_length": 20.78125, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.18, "frac_reward_zero_std": 0.5, "grad_norm": 11.35882568359375, "learning_rate": 2.128356686585282e-07, "loss": 0.0, "num_tokens": 7438378.0, "reward": 0.45859378576278687, "reward_std": 0.17108294367790222, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 15.0, "completions/max_terminated_length": 15.0, "completions/mean_length": 13.03125, "completions/mean_terminated_length": 13.03125, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.1825, "frac_reward_zero_std": 0.625, "grad_norm": 11.875134468078613, "learning_rate": 1.9945942635848745e-07, "loss": -0.0, "num_tokens": 7510852.0, "reward": 0.7390625476837158, "reward_std": 0.17329266667366028, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 370.0, "completions/max_terminated_length": 370.0, "completions/mean_length": 19.203125, "completions/mean_terminated_length": 19.203125, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.185, "frac_reward_zero_std": 0.5, "grad_norm": 13.027864456176758, "learning_rate": 1.8641182076323148e-07, "loss": 0.0, "num_tokens": 7608273.0, "reward": 0.7390625476837158, "reward_std": 0.18484057486057281, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 1156.0, "completions/max_terminated_length": 1156.0, "completions/mean_length": 50.65625, "completions/mean_terminated_length": 50.65625, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.1875, "frac_reward_zero_std": 0.625, "grad_norm": 8.088542938232422, "learning_rate": 1.7370711923791564e-07, "loss": 0.0, "num_tokens": 7691619.0, "reward": 0.5843750238418579, "reward_std": 0.16887325048446655, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "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": 1471.0, "completions/max_terminated_length": 1471.0, "completions/mean_length": 43.625, "completions/mean_terminated_length": 43.625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.19, "frac_reward_zero_std": 0.625, "grad_norm": 10.211929321289062, "learning_rate": 1.6135921418712955e-07, "loss": -0.0, "num_tokens": 7776755.0, "reward": 0.8812500238418579, "reward_std": 0.1552036553621292, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.78125, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 249.0, "completions/max_terminated_length": 249.0, "completions/mean_length": 16.90625, "completions/mean_terminated_length": 16.90625, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.1925, "frac_reward_zero_std": 0.625, "grad_norm": 12.14816665649414, "learning_rate": 1.493816078637557e-07, "loss": -0.0, "num_tokens": 7856829.0, "reward": 0.4437500238418579, "reward_std": 0.1552036553621292, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 400.0, "completions/max_terminated_length": 400.0, "completions/mean_length": 17.484375, "completions/mean_terminated_length": 17.484375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.195, "frac_reward_zero_std": 0.75, "grad_norm": 8.064270973205566, "learning_rate": 1.3778739760445552e-07, "loss": -0.0, "num_tokens": 7957444.0, "reward": 0.6195312738418579, "reward_std": 0.06007346510887146, "rewards/format_reward/mean": 0.8828125, "rewards/format_reward/std": 0.21347814798355103, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 862.0, "completions/max_terminated_length": 862.0, "completions/mean_length": 28.59375, "completions/mean_terminated_length": 28.59375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1975, "frac_reward_zero_std": 0.5, "grad_norm": 15.456476211547852, "learning_rate": 1.2658926150792322e-07, "loss": 0.0, "num_tokens": 8047770.0, "reward": 0.753125011920929, "reward_std": 0.23930205404758453, "rewards/format_reward/mean": 0.96875, "rewards/format_reward/std": 0.1510545015335083, "rewards/mcq_exact_match_reward/mean": 0.65625, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 413.0, "completions/max_terminated_length": 413.0, "completions/mean_length": 23.6875, "completions/mean_terminated_length": 23.6875, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.2, "frac_reward_zero_std": 0.75, "grad_norm": 6.343238353729248, "learning_rate": 1.1579944457157059e-07, "loss": -0.0, "num_tokens": 8158614.0, "reward": 0.6000000238418579, "reward_std": 0.0883883461356163, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 80 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 19.0, "completions/max_terminated_length": 19.0, "completions/mean_length": 12.328125, "completions/mean_terminated_length": 12.328125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2025, "frac_reward_zero_std": 0.5, "grad_norm": 19.15117073059082, "learning_rate": 1.0542974530180327e-07, "loss": -0.0, "num_tokens": 8261035.0, "reward": 0.5007812976837158, "reward_std": 0.1689612865447998, "rewards/format_reward/mean": 0.9453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "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": 118.0, "completions/max_terminated_length": 118.0, "completions/mean_length": 16.375, "completions/mean_terminated_length": 16.375, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.205, "frac_reward_zero_std": 0.625, "grad_norm": 12.861918449401855, "learning_rate": 9.549150281252632e-08, "loss": 0.0, "num_tokens": 8359787.0, "reward": 0.20937500894069672, "reward_std": 0.16887325048446655, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.109375, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "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": 119.0, "completions/max_terminated_length": 119.0, "completions/mean_length": 15.03125, "completions/mean_terminated_length": 15.03125, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.2075, "frac_reward_zero_std": 0.75, "grad_norm": 14.756561279296875, "learning_rate": 8.599558442598998e-08, "loss": 0.0, "num_tokens": 8441949.0, "reward": 0.6156250238418579, "reward_std": 0.1315089464187622, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 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": 310.0, "completions/max_terminated_length": 310.0, "completions/mean_length": 19.09375, "completions/mean_terminated_length": 19.09375, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.21, "frac_reward_zero_std": 0.5, "grad_norm": 10.06808090209961, "learning_rate": 7.695237378953224e-08, "loss": 0.0, "num_tokens": 8543755.0, "reward": 0.8500000238418579, "reward_std": 0.19727616012096405, "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": 84 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 18.0, "completions/mean_length": 75.015625, "completions/mean_terminated_length": 11.370967864990234, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2125, "frac_reward_zero_std": 0.25, "grad_norm": 14.676437377929688, "learning_rate": 6.837175952121304e-08, "loss": 0.0, "num_tokens": 8640956.0, "reward": 0.4742187559604645, "reward_std": 0.34538891911506653, "rewards/format_reward/mean": 0.8359375, "rewards/format_reward/std": 0.2824873626232147, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 22.0, "completions/max_terminated_length": 22.0, "completions/mean_length": 13.515625, "completions/mean_terminated_length": 13.515625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.215, "frac_reward_zero_std": 0.5, "grad_norm": 14.535361289978027, "learning_rate": 6.026312439675551e-08, "loss": 0.0, "num_tokens": 8729309.0, "reward": 0.55078125, "reward_std": 0.1658112108707428, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 101.0, "completions/mean_length": 46.234375, "completions/mean_terminated_length": 14.460318565368652, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.2175, "frac_reward_zero_std": 0.375, "grad_norm": 18.44989776611328, "learning_rate": 5.263533508961826e-08, "loss": 0.0, "num_tokens": 8833324.0, "reward": 0.81640625, "reward_std": 0.30399811267852783, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.71875, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 347.0, "completions/max_terminated_length": 347.0, "completions/mean_length": 22.5625, "completions/mean_terminated_length": 22.5625, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.22, "frac_reward_zero_std": 0.5, "grad_norm": 14.351069450378418, "learning_rate": 4.549673247541874e-08, "loss": 0.0, "num_tokens": 8932336.0, "reward": 0.6781250238418579, "reward_std": 0.2519446909427643, "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": 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": 388.0, "completions/max_terminated_length": 388.0, "completions/mean_length": 27.359375, "completions/mean_terminated_length": 27.359375, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.2225, "frac_reward_zero_std": 0.25, "grad_norm": 14.434761047363281, "learning_rate": 3.8855122511307626e-08, "loss": 0.0, "num_tokens": 9044767.0, "reward": 0.6312500238418579, "reward_std": 0.31300368905067444, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 1093.0, "completions/max_terminated_length": 1093.0, "completions/mean_length": 30.03125, "completions/mean_terminated_length": 30.03125, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.225, "frac_reward_zero_std": 0.625, "grad_norm": 15.73192310333252, "learning_rate": 3.271776770026963e-08, "loss": 0.0, "num_tokens": 9157089.0, "reward": 0.5367187857627869, "reward_std": 0.1834089457988739, "rewards/format_reward/mean": 0.9921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 499.0, "completions/max_terminated_length": 499.0, "completions/mean_length": 26.140625, "completions/mean_terminated_length": 26.140625, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.2275, "frac_reward_zero_std": 0.625, "grad_norm": 15.007508277893066, "learning_rate": 2.7091379149682682e-08, "loss": 0.0, "num_tokens": 9224186.0, "reward": 0.7718750238418579, "reward_std": 0.15992169082164764, "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": 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": 305.0, "completions/max_terminated_length": 305.0, "completions/mean_length": 19.234375, "completions/mean_terminated_length": 19.234375, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.23, "frac_reward_zero_std": 0.625, "grad_norm": 7.761956691741943, "learning_rate": 2.1982109232821176e-08, "loss": -0.0, "num_tokens": 9321929.0, "reward": 0.5218750238418579, "reward_std": 0.1530819982290268, "rewards/format_reward/mean": 1.0, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 20.0, "completions/mean_length": 45.453125, "completions/mean_terminated_length": 13.666666984558105, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.2325, "frac_reward_zero_std": 0.625, "grad_norm": 12.21143627166748, "learning_rate": 1.7395544861325718e-08, "loss": -0.0, "num_tokens": 9410454.0, "reward": 0.64453125, "reward_std": 0.07132276892662048, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 321.0, "completions/mean_length": 87.828125, "completions/mean_terminated_length": 24.596773147583008, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.235, "frac_reward_zero_std": 0.375, "grad_norm": 16.356138229370117, "learning_rate": 1.3336701375997127e-08, "loss": -0.0, "num_tokens": 9519363.0, "reward": 0.7523437738418579, "reward_std": 0.23261141777038574, "rewards/format_reward/mean": 0.9609375, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.65625, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 400.0, "completions/max_terminated_length": 400.0, "completions/mean_length": 19.109375, "completions/mean_terminated_length": 19.109375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2375, "frac_reward_zero_std": 0.75, "grad_norm": 7.05694055557251, "learning_rate": 9.810017062595321e-09, "loss": 0.0, "num_tokens": 9619178.0, "reward": 0.5835937261581421, "reward_std": 0.12194531410932541, "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": 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": 326.0, "completions/max_terminated_length": 326.0, "completions/mean_length": 23.65625, "completions/mean_terminated_length": 23.65625, "completions/min_length": 12.0, "completions/min_terminated_length": 12.0, "epoch": 0.24, "frac_reward_zero_std": 0.25, "grad_norm": 13.755544662475586, "learning_rate": 6.819348298638839e-09, "loss": 0.0, "num_tokens": 9730556.0, "reward": 0.3476562798023224, "reward_std": 0.2659573554992676, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 96 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 278.0, "completions/mean_length": 52.09375, "completions/mean_terminated_length": 20.41269874572754, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.2425, "frac_reward_zero_std": 0.125, "grad_norm": 16.00373077392578, "learning_rate": 4.367965336512403e-09, "loss": 0.0, "num_tokens": 9832362.0, "reward": 0.6765625476837158, "reward_std": 0.303753525018692, "rewards/format_reward/mean": 0.984375, "rewards/format_reward/std": 0.125, "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": 43.0, "completions/max_terminated_length": 43.0, "completions/mean_length": 13.59375, "completions/mean_terminated_length": 13.59375, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.245, "frac_reward_zero_std": 0.625, "grad_norm": 11.894449234008789, "learning_rate": 2.458548727494292e-09, "loss": -0.0, "num_tokens": 9954448.0, "reward": 0.37890625, "reward_std": 0.09131823480129242, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 584.0, "completions/max_terminated_length": 584.0, "completions/mean_length": 38.921875, "completions/mean_terminated_length": 38.921875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2475, "frac_reward_zero_std": 0.375, "grad_norm": 21.100183486938477, "learning_rate": 1.0931863906127325e-09, "loss": 0.0, "num_tokens": 10048931.0, "reward": 0.4414062798023224, "reward_std": 0.2259259968996048, "rewards/format_reward/mean": 0.9765625, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 99 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 18.0, "completions/max_terminated_length": 18.0, "completions/mean_length": 13.078125, "completions/mean_terminated_length": 13.078125, "completions/min_length": 13.0, "completions/min_terminated_length": 13.0, "epoch": 0.25, "frac_reward_zero_std": 0.875, "grad_norm": 9.943309783935547, "learning_rate": 2.733713295369755e-10, "loss": 0.0, "num_tokens": 10141176.0, "reward": 0.7718750238418579, "reward_std": 0.0646936446428299, "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": 100 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 10141176, "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 }