{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.11160714285714286, "eval_steps": 500, "global_step": 50, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1639.0, "completions/max_terminated_length": 1639.0, "completions/mean_length": 609.15625, "completions/mean_terminated_length": 609.15625, "completions/min_length": 22.0, "completions/min_terminated_length": 22.0, "epoch": 0.002232142857142857, "frac_reward_zero_std": 0.0, "grad_norm": 2.2023043632507324, "learning_rate": 0.0, "loss": -0.0, "num_tokens": 126514.0, "reward": 0.3804687261581421, "reward_std": 0.3518281579017639, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.22257846593856812, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1383.0, "completions/mean_length": 572.40625, "completions/mean_terminated_length": 524.8064575195312, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.004464285714285714, "frac_reward_zero_std": 0.125, "grad_norm": 25.689533233642578, "learning_rate": 2e-07, "loss": 0.0, "num_tokens": 260700.0, "reward": 0.21171873807907104, "reward_std": 0.3009757697582245, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.22146137058734894, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 2 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1475.0, "completions/max_terminated_length": 1475.0, "completions/mean_length": 496.328125, "completions/mean_terminated_length": 496.328125, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.006696428571428571, "frac_reward_zero_std": 0.0, "grad_norm": 12.14404582977295, "learning_rate": 4e-07, "loss": -0.0, "num_tokens": 366361.0, "reward": 0.2242187261581421, "reward_std": 0.3805117607116699, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.23974503576755524, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 1414.0, "completions/max_terminated_length": 1414.0, "completions/mean_length": 349.609375, "completions/mean_terminated_length": 349.609375, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.008928571428571428, "frac_reward_zero_std": 0.0, "grad_norm": 37.033538818359375, "learning_rate": 6e-07, "loss": -0.0, "num_tokens": 469360.0, "reward": 0.38359373807907104, "reward_std": 0.3727972209453583, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.25479042530059814, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 4 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1515.0, "completions/mean_length": 654.1875, "completions/mean_terminated_length": 585.6392822265625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.011160714285714286, "frac_reward_zero_std": 0.0, "grad_norm": 2.6391189098358154, "learning_rate": 8e-07, "loss": -0.0, "num_tokens": 606908.0, "reward": 0.2398437261581421, "reward_std": 0.3410371243953705, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.23974503576755524, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 2025.0, "completions/max_terminated_length": 2025.0, "completions/mean_length": 505.546875, "completions/mean_terminated_length": 505.546875, "completions/min_length": 8.0, "completions/min_terminated_length": 8.0, "epoch": 0.013392857142857142, "frac_reward_zero_std": 0.125, "grad_norm": 3.158848524093628, "learning_rate": 1e-06, "loss": -0.0, "num_tokens": 722519.0, "reward": 0.34765625, "reward_std": 0.35713839530944824, "rewards/format_reward/mean": 0.3515625, "rewards/format_reward/std": 0.26246222853660583, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 6 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1934.0, "completions/max_terminated_length": 1934.0, "completions/mean_length": 470.03125, "completions/mean_terminated_length": 470.03125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.015625, "frac_reward_zero_std": 0.0, "grad_norm": 13.845512390136719, "learning_rate": 9.99726628670463e-07, "loss": -0.0, "num_tokens": 845529.0, "reward": 0.4609374701976776, "reward_std": 0.43220406770706177, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.2592533528804779, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 7 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1473.0, "completions/mean_length": 453.65625, "completions/mean_terminated_length": 428.3492431640625, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.017857142857142856, "frac_reward_zero_std": 0.0, "grad_norm": 22.4105281829834, "learning_rate": 9.989068136093872e-07, "loss": 0.0, "num_tokens": 950339.0, "reward": 0.24531249701976776, "reward_std": 0.321063756942749, "rewards/format_reward/mean": 0.421875, "rewards/format_reward/std": 0.3238992393016815, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1569.0, "completions/mean_length": 697.828125, "completions/mean_terminated_length": 654.274169921875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.020089285714285716, "frac_reward_zero_std": 0.0, "grad_norm": 4.999773979187012, "learning_rate": 9.975414512725056e-07, "loss": 0.0, "num_tokens": 1106384.0, "reward": 0.2109374850988388, "reward_std": 0.27513793110847473, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 1300.0, "completions/max_terminated_length": 1300.0, "completions/mean_length": 368.015625, "completions/mean_terminated_length": 368.015625, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.022321428571428572, "frac_reward_zero_std": 0.0, "grad_norm": 29.1524600982666, "learning_rate": 9.956320346634875e-07, "loss": 0.0, "num_tokens": 1224897.0, "reward": 0.2734375, "reward_std": 0.30659258365631104, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.2592533528804779, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1506.0, "completions/mean_length": 439.46875, "completions/mean_terminated_length": 387.58062744140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.024553571428571428, "frac_reward_zero_std": 0.0, "grad_norm": 6.193141460418701, "learning_rate": 9.931806517013612e-07, "loss": -0.0, "num_tokens": 1337247.0, "reward": 0.39531245827674866, "reward_std": 0.3140166997909546, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.22658175230026245, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 1578.0, "completions/max_terminated_length": 1578.0, "completions/mean_length": 526.609375, "completions/mean_terminated_length": 526.609375, "completions/min_length": 11.0, "completions/min_terminated_length": 11.0, "epoch": 0.026785714285714284, "frac_reward_zero_std": 0.0, "grad_norm": 4.610779762268066, "learning_rate": 9.901899829374047e-07, "loss": -0.0, "num_tokens": 1471670.0, "reward": 0.3179687261581421, "reward_std": 0.38177454471588135, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.22257846593856812, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 12 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1653.0, "completions/mean_length": 558.546875, "completions/mean_terminated_length": 534.90478515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.029017857142857144, "frac_reward_zero_std": 0.0, "grad_norm": 3.6152477264404297, "learning_rate": 9.866632986240029e-07, "loss": 0.0, "num_tokens": 1588257.0, "reward": 0.24453124403953552, "reward_std": 0.3110152781009674, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.27537402510643005, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 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": 1594.0, "completions/max_terminated_length": 1594.0, "completions/mean_length": 475.703125, "completions/mean_terminated_length": 475.703125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03125, "frac_reward_zero_std": 0.125, "grad_norm": 4.853925704956055, "learning_rate": 9.826044551386742e-07, "loss": 0.0, "num_tokens": 1720934.0, "reward": 0.22890622913837433, "reward_std": 0.2894315719604492, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.2605654299259186, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1740.0, "completions/mean_length": 696.109375, "completions/mean_terminated_length": 629.6229248046875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.033482142857142856, "frac_reward_zero_std": 0.0, "grad_norm": 2.440049171447754, "learning_rate": 9.780178907671788e-07, "loss": -0.0, "num_tokens": 1862341.0, "reward": 0.39531248807907104, "reward_std": 0.3890523314476013, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.22658175230026245, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 1724.0, "completions/mean_length": 585.90625, "completions/mean_terminated_length": 562.6984252929688, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03571428571428571, "frac_reward_zero_std": 0.0, "grad_norm": 6.889235496520996, "learning_rate": 9.729086208503173e-07, "loss": 0.0, "num_tokens": 1991975.0, "reward": 0.3343749940395355, "reward_std": 0.3780812621116638, "rewards/format_reward/mean": 0.375, "rewards/format_reward/std": 0.26726123690605164, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 16 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1438.0, "completions/mean_length": 771.46875, "completions/mean_terminated_length": 730.290283203125, "completions/min_length": 293.0, "completions/min_terminated_length": 293.0, "epoch": 0.03794642857142857, "frac_reward_zero_std": 0.125, "grad_norm": 0.7261731028556824, "learning_rate": 9.672822322997304e-07, "loss": -0.0, "num_tokens": 2142765.0, "reward": 0.18203124403953552, "reward_std": 0.26640385389328003, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.2280818521976471, "rewards/mcq_exact_match_reward/mean": 0.140625, "rewards/mcq_exact_match_reward/std": 0.3503824472427368, "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": 1903.0, "completions/max_terminated_length": 1903.0, "completions/mean_length": 623.953125, "completions/mean_terminated_length": 623.953125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04017857142857143, "frac_reward_zero_std": 0.0, "grad_norm": 12.646529197692871, "learning_rate": 9.611448774886923e-07, "loss": 0.0, "num_tokens": 2280298.0, "reward": 0.23203124105930328, "reward_std": 0.29715561866760254, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1689.0, "completions/mean_length": 585.21875, "completions/mean_terminated_length": 562.0000610351562, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04241071428571429, "frac_reward_zero_std": 0.0, "grad_norm": 5.6680755615234375, "learning_rate": 9.545032675245813e-07, "loss": 0.0, "num_tokens": 2429904.0, "reward": 0.4281249940395355, "reward_std": 0.40475884079933167, "rewards/format_reward/mean": 0.53125, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1988.0, "completions/mean_length": 737.34375, "completions/mean_terminated_length": 672.8851928710938, "completions/min_length": 55.0, "completions/min_terminated_length": 55.0, "epoch": 0.044642857142857144, "frac_reward_zero_std": 0.0, "grad_norm": 1.117641806602478, "learning_rate": 9.473646649103817e-07, "loss": -0.0, "num_tokens": 2581766.0, "reward": 0.4140625, "reward_std": 0.42353588342666626, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.2741328477859497, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 2032.0, "completions/mean_length": 682.84375, "completions/mean_terminated_length": 661.1746215820312, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.046875, "frac_reward_zero_std": 0.0, "grad_norm": 8.76787281036377, "learning_rate": 9.397368756032444e-07, "loss": 0.0, "num_tokens": 2745580.0, "reward": 0.3460937440395355, "reward_std": 0.3644231855869293, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.16648055613040924, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 1910.0, "completions/max_terminated_length": 1910.0, "completions/mean_length": 636.734375, "completions/mean_terminated_length": 636.734375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.049107142857142856, "frac_reward_zero_std": 0.0, "grad_norm": 3.8484814167022705, "learning_rate": 9.316282404787869e-07, "loss": 0.0, "num_tokens": 2883659.0, "reward": 0.3906249701976776, "reward_std": 0.42049989104270935, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.25, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 2016.0, "completions/mean_length": 648.96875, "completions/mean_terminated_length": 580.1638793945312, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05133928571428571, "frac_reward_zero_std": 0.0, "grad_norm": 13.19649887084961, "learning_rate": 9.230476262104676e-07, "loss": 0.0, "num_tokens": 3047465.0, "reward": 0.11484374850988388, "reward_std": 0.23332995176315308, "rewards/format_reward/mean": 0.3671875, "rewards/format_reward/std": 0.22257846593856812, "rewards/mcq_exact_match_reward/mean": 0.078125, "rewards/mcq_exact_match_reward/std": 0.27048972249031067, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1814.0, "completions/mean_length": 641.75, "completions/mean_terminated_length": 596.3870849609375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05357142857142857, "frac_reward_zero_std": 0.125, "grad_norm": 10.830889701843262, "learning_rate": 9.1400441557401e-07, "loss": 0.0, "num_tokens": 3210817.0, "reward": 0.21328124403953552, "reward_std": 0.2633717656135559, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.30282649397850037, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 24 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1800.0, "completions/mean_length": 710.109375, "completions/mean_terminated_length": 688.873046875, "completions/min_length": 221.0, "completions/min_terminated_length": 221.0, "epoch": 0.05580357142857143, "frac_reward_zero_std": 0.125, "grad_norm": 0.8190599083900452, "learning_rate": 9.045084971874737e-07, "loss": -0.0, "num_tokens": 3349960.0, "reward": 0.3906249701976776, "reward_std": 0.348865270614624, "rewards/format_reward/mean": 0.46875, "rewards/format_reward/std": 0.17536810040473938, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1323.0, "completions/max_terminated_length": 1323.0, "completions/mean_length": 530.28125, "completions/mean_terminated_length": 530.28125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05803571428571429, "frac_reward_zero_std": 0.25, "grad_norm": 2.7582437992095947, "learning_rate": 8.945702546981968e-07, "loss": 0.0, "num_tokens": 3480610.0, "reward": 0.3960937559604645, "reward_std": 0.3427093029022217, "rewards/format_reward/mean": 0.5234375, "rewards/format_reward/std": 0.24282869696617126, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1864.0, "completions/mean_length": 717.171875, "completions/mean_terminated_length": 674.241943359375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.060267857142857144, "frac_reward_zero_std": 0.0, "grad_norm": 3.3076305389404297, "learning_rate": 8.842005554284295e-07, "loss": 0.0, "num_tokens": 3663341.0, "reward": 0.2437499761581421, "reward_std": 0.3142412304878235, "rewards/format_reward/mean": 0.40625, "rewards/format_reward/std": 0.21593283116817474, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1616.0, "completions/mean_length": 695.453125, "completions/mean_terminated_length": 628.9343872070312, "completions/min_length": 215.0, "completions/min_terminated_length": 215.0, "epoch": 0.0625, "frac_reward_zero_std": 0.25, "grad_norm": 0.7634695172309875, "learning_rate": 8.734107384920769e-07, "loss": -0.0, "num_tokens": 3812410.0, "reward": 0.2945312261581421, "reward_std": 0.32147979736328125, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.2538151443004608, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 28 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1797.0, "completions/mean_length": 509.3125, "completions/mean_terminated_length": 484.888916015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06473214285714286, "frac_reward_zero_std": 0.125, "grad_norm": 6.607220649719238, "learning_rate": 8.622126023955445e-07, "loss": -0.0, "num_tokens": 3937998.0, "reward": 0.47109371423721313, "reward_std": 0.4023832082748413, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.16648055613040924, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "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": 1948.0, "completions/max_terminated_length": 1948.0, "completions/mean_length": 672.265625, "completions/mean_terminated_length": 672.265625, "completions/min_length": 72.0, "completions/min_terminated_length": 72.0, "epoch": 0.06696428571428571, "frac_reward_zero_std": 0.125, "grad_norm": 1.1837975978851318, "learning_rate": 8.506183921362442e-07, "loss": 0.0, "num_tokens": 4069807.0, "reward": 0.42265623807907104, "reward_std": 0.24412530660629272, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.28770697116851807, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1799.0, "completions/mean_length": 536.515625, "completions/mean_terminated_length": 512.5238647460938, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06919642857142858, "frac_reward_zero_std": 0.0, "grad_norm": 14.385038375854492, "learning_rate": 8.386407858128706e-07, "loss": 0.0, "num_tokens": 4179424.0, "reward": 0.3453125059604645, "reward_std": 0.3491568863391876, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.23517554998397827, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1616.0, "completions/mean_length": 537.859375, "completions/mean_terminated_length": 513.888916015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07142857142857142, "frac_reward_zero_std": 0.125, "grad_norm": 10.334357261657715, "learning_rate": 8.262928807620843e-07, "loss": 0.0, "num_tokens": 4308543.0, "reward": 0.3101562261581421, "reward_std": 0.37167930603027344, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1829.0, "completions/mean_length": 727.734375, "completions/mean_terminated_length": 685.1451416015625, "completions/min_length": 207.0, "completions/min_terminated_length": 207.0, "epoch": 0.07366071428571429, "frac_reward_zero_std": 0.125, "grad_norm": 0.7631356120109558, "learning_rate": 8.135881792367685e-07, "loss": 0.0, "num_tokens": 4472694.0, "reward": 0.31953126192092896, "reward_std": 0.35187065601348877, "rewards/format_reward/mean": 0.5390625, "rewards/format_reward/std": 0.20518454909324646, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 33 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1818.0, "completions/mean_length": 534.390625, "completions/mean_terminated_length": 485.56451416015625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07589285714285714, "frac_reward_zero_std": 0.25, "grad_norm": 5.339401721954346, "learning_rate": 8.005405736415125e-07, "loss": 0.0, "num_tokens": 4603871.0, "reward": 0.4007812738418579, "reward_std": 0.26095426082611084, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.26528194546699524, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 0.09375, "completions/max_length": 2048.0, "completions/max_terminated_length": 1676.0, "completions/mean_length": 726.609375, "completions/mean_terminated_length": 589.913818359375, "completions/min_length": 100.0, "completions/min_terminated_length": 100.0, "epoch": 0.078125, "frac_reward_zero_std": 0.0, "grad_norm": 1.0368614196777344, "learning_rate": 7.871643313414718e-07, "loss": 0.0, "num_tokens": 4731646.0, "reward": 0.3101562261581421, "reward_std": 0.3630598783493042, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.20152568817138672, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1353.0, "completions/mean_length": 511.03125, "completions/mean_terminated_length": 486.63494873046875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08035714285714286, "frac_reward_zero_std": 0.125, "grad_norm": 10.782816886901855, "learning_rate": 7.734740790612136e-07, "loss": -0.0, "num_tokens": 4839608.0, "reward": 0.22343748807907104, "reward_std": 0.3053799867630005, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.17747680842876434, "rewards/mcq_exact_match_reward/mean": 0.171875, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "step": 36 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1972.0, "completions/mean_length": 674.5625, "completions/mean_terminated_length": 652.761962890625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08258928571428571, "frac_reward_zero_std": 0.0, "grad_norm": 5.231216907501221, "learning_rate": 7.594847868906076e-07, "loss": 0.0, "num_tokens": 4977508.0, "reward": 0.23671874403953552, "reward_std": 0.35034993290901184, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.2744719088077545, "rewards/mcq_exact_match_reward/mean": 0.1875, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 0.078125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1850.0, "completions/mean_length": 762.984375, "completions/mean_terminated_length": 654.084716796875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08482142857142858, "frac_reward_zero_std": 0.125, "grad_norm": 3.4671003818511963, "learning_rate": 7.452117519152541e-07, "loss": -0.0, "num_tokens": 5138179.0, "reward": 0.2828124761581421, "reward_std": 0.387776255607605, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.2812775671482086, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 1828.0, "completions/max_terminated_length": 1828.0, "completions/mean_length": 557.0, "completions/mean_terminated_length": 557.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08705357142857142, "frac_reward_zero_std": 0.0, "grad_norm": 4.978475570678711, "learning_rate": 7.306705814893439e-07, "loss": 0.0, "num_tokens": 5254915.0, "reward": 0.5828124284744263, "reward_std": 0.37641221284866333, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.1985812783241272, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 1996.0, "completions/mean_length": 549.765625, "completions/mean_terminated_length": 525.984130859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08928571428571429, "frac_reward_zero_std": 0.25, "grad_norm": 14.763313293457031, "learning_rate": 7.158771761692464e-07, "loss": 0.0, "num_tokens": 5366588.0, "reward": 0.2992187440395355, "reward_std": 0.2617577314376831, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.2744719088077545, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 2036.0, "completions/mean_length": 695.21875, "completions/mean_terminated_length": 628.6884765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09151785714285714, "frac_reward_zero_std": 0.0, "grad_norm": 4.990268230438232, "learning_rate": 7.008477123264847e-07, "loss": -0.0, "num_tokens": 5525970.0, "reward": 0.3609374761581421, "reward_std": 0.28916043043136597, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.2514837086200714, "rewards/mcq_exact_match_reward/mean": 0.3125, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 41 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1187.0, "completions/max_terminated_length": 1187.0, "completions/mean_length": 503.046875, "completions/mean_terminated_length": 503.046875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09375, "frac_reward_zero_std": 0.0, "grad_norm": 5.825297832489014, "learning_rate": 6.855986244591103e-07, "loss": -0.0, "num_tokens": 5652813.0, "reward": 0.3109374940395355, "reward_std": 0.27216827869415283, "rewards/format_reward/mean": 0.453125, "rewards/format_reward/std": 0.23091863095760345, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 0.046875, "completions/max_length": 2048.0, "completions/max_terminated_length": 1779.0, "completions/mean_length": 698.46875, "completions/mean_terminated_length": 632.0983276367188, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09598214285714286, "frac_reward_zero_std": 0.0, "grad_norm": 13.259655952453613, "learning_rate": 6.701465872208216e-07, "loss": -0.0, "num_tokens": 5805515.0, "reward": 0.4039062261581421, "reward_std": 0.42280226945877075, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.29703813791275024, "rewards/mcq_exact_match_reward/mean": 0.359375, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1473.0, "completions/mean_length": 426.15625, "completions/mean_terminated_length": 373.83868408203125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09821428571428571, "frac_reward_zero_std": 0.0, "grad_norm": 17.04229164123535, "learning_rate": 6.545084971874736e-07, "loss": 0.0, "num_tokens": 5926829.0, "reward": 0.3999999761581421, "reward_std": 0.3521348237991333, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.208927720785141, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 44 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1995.0, "completions/mean_length": 559.859375, "completions/mean_terminated_length": 511.8548278808594, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10044642857142858, "frac_reward_zero_std": 0.0, "grad_norm": 16.93348503112793, "learning_rate": 6.387014543809223e-07, "loss": 0.0, "num_tokens": 6053444.0, "reward": 0.40468746423721313, "reward_std": 0.3729129731655121, "rewards/format_reward/mean": 0.609375, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 45 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1565.0, "completions/max_terminated_length": 1565.0, "completions/mean_length": 373.53125, "completions/mean_terminated_length": 373.53125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10267857142857142, "frac_reward_zero_std": 0.125, "grad_norm": 16.590978622436523, "learning_rate": 6.227427435703995e-07, "loss": -0.0, "num_tokens": 6187174.0, "reward": 0.27421873807907104, "reward_std": 0.33643338084220886, "rewards/format_reward/mean": 0.5546875, "rewards/format_reward/std": 0.20152568817138672, "rewards/mcq_exact_match_reward/mean": 0.21875, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "step": 46 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1133.0, "completions/max_terminated_length": 1133.0, "completions/mean_length": 301.3125, "completions/mean_terminated_length": 301.3125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10491071428571429, "frac_reward_zero_std": 0.0, "grad_norm": 40.43541717529297, "learning_rate": 6.066498153718734e-07, "loss": -0.0, "num_tokens": 6299226.0, "reward": 0.38515621423721313, "reward_std": 0.33590227365493774, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.19654127955436707, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "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": 1598.0, "completions/mean_length": 501.921875, "completions/mean_terminated_length": 477.3809814453125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.10714285714285714, "frac_reward_zero_std": 0.125, "grad_norm": 8.04982852935791, "learning_rate": 5.90440267166055e-07, "loss": -0.0, "num_tokens": 6423469.0, "reward": 0.40156251192092896, "reward_std": 0.3936265707015991, "rewards/format_reward/mean": 0.578125, "rewards/format_reward/std": 0.25539806485176086, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1493.0, "completions/mean_length": 285.203125, "completions/mean_terminated_length": 257.22222900390625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.109375, "frac_reward_zero_std": 0.25, "grad_norm": 21.848068237304688, "learning_rate": 5.741318238559209e-07, "loss": -0.0, "num_tokens": 6559442.0, "reward": 0.46171873807907104, "reward_std": 0.25157496333122253, "rewards/format_reward/mean": 0.5546875, "rewards/format_reward/std": 0.20152568817138672, "rewards/mcq_exact_match_reward/mean": 0.40625, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 49 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1993.0, "completions/mean_length": 242.203125, "completions/mean_terminated_length": 183.9516143798828, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.11160714285714286, "frac_reward_zero_std": 0.125, "grad_norm": 18.545976638793945, "learning_rate": 5.577423184847931e-07, "loss": 0.0, "num_tokens": 6663423.0, "reward": 0.27421873807907104, "reward_std": 0.30959123373031616, "rewards/format_reward/mean": 0.7109375, "rewards/format_reward/std": 0.2928335666656494, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 50 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 6663423, "num_train_epochs": 1, "save_steps": 50, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }