{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.25, "eval_steps": 500, "global_step": 100, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 677.0, "completions/max_terminated_length": 677.0, "completions/mean_length": 187.359375, "completions/mean_terminated_length": 187.359375, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.0025, "frac_reward_zero_std": 0.125, "grad_norm": 50.62995910644531, "learning_rate": 0.0, "loss": 0.0, "num_tokens": 98191.0, "reward": 0.4453125, "reward_std": 0.33708566427230835, "rewards/format_reward/mean": 0.234375, "rewards/format_reward/std": 0.26679685711860657, "rewards/mcq_exact_match_reward/mean": 0.421875, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 1 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1398.0, "completions/mean_length": 261.8125, "completions/mean_terminated_length": 233.4603271484375, "completions/min_length": 2.0, "completions/min_terminated_length": 2.0, "epoch": 0.005, "frac_reward_zero_std": 0.25, "grad_norm": 20.401866912841797, "learning_rate": 2e-07, "loss": -0.0, "num_tokens": 213363.0, "reward": 0.3749999701976776, "reward_std": 0.24398061633110046, "rewards/format_reward/mean": 0.3125, "rewards/format_reward/std": 0.24397502839565277, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1416.0, "completions/max_terminated_length": 1416.0, "completions/mean_length": 227.703125, "completions/mean_terminated_length": 227.703125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0075, "frac_reward_zero_std": 0.0, "grad_norm": 15.30715560913086, "learning_rate": 4e-07, "loss": -0.0, "num_tokens": 324808.0, "reward": 0.40703123807907104, "reward_std": 0.3755374252796173, "rewards/format_reward/mean": 0.3203125, "rewards/format_reward/std": 0.24180518090724945, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 1488.0, "completions/max_terminated_length": 1488.0, "completions/mean_length": 186.9375, "completions/mean_terminated_length": 186.9375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.01, "frac_reward_zero_std": 0.125, "grad_norm": 24.156063079833984, "learning_rate": 6e-07, "loss": -0.0, "num_tokens": 436476.0, "reward": 0.2828125059604645, "reward_std": 0.19103410840034485, "rewards/format_reward/mean": 0.328125, "rewards/format_reward/std": 0.23935678601264954, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "step": 4 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1536.0, "completions/mean_length": 447.6875, "completions/mean_terminated_length": 422.2857360839844, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0125, "frac_reward_zero_std": 0.0, "grad_norm": 11.044807434082031, "learning_rate": 8e-07, "loss": -0.0, "num_tokens": 557704.0, "reward": 0.3046874701976776, "reward_std": 0.39356279373168945, "rewards/format_reward/mean": 0.390625, "rewards/format_reward/std": 0.24346621334552765, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 5 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1940.0, "completions/max_terminated_length": 1940.0, "completions/mean_length": 294.53125, "completions/mean_terminated_length": 294.53125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.015, "frac_reward_zero_std": 0.25, "grad_norm": 6.50140380859375, "learning_rate": 1e-06, "loss": 0.0, "num_tokens": 654738.0, "reward": 0.6945312023162842, "reward_std": 0.3751453161239624, "rewards/format_reward/mean": 0.3828125, "rewards/format_reward/std": 0.3052735924720764, "rewards/mcq_exact_match_reward/mean": 0.65625, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "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": 1237.0, "completions/max_terminated_length": 1237.0, "completions/mean_length": 224.375, "completions/mean_terminated_length": 224.375, "completions/min_length": 3.0, "completions/min_terminated_length": 3.0, "epoch": 0.0175, "frac_reward_zero_std": 0.125, "grad_norm": 10.862257957458496, "learning_rate": 9.99726628670463e-07, "loss": -0.0, "num_tokens": 755618.0, "reward": 0.41484373807907104, "reward_std": 0.35415294766426086, "rewards/format_reward/mean": 0.3984375, "rewards/format_reward/std": 0.2978719174861908, "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": 1277.0, "completions/max_terminated_length": 1277.0, "completions/mean_length": 246.34375, "completions/mean_terminated_length": 246.34375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.02, "frac_reward_zero_std": 0.0, "grad_norm": 13.488003730773926, "learning_rate": 9.989068136093872e-07, "loss": -0.0, "num_tokens": 868968.0, "reward": 0.37968745827674866, "reward_std": 0.3651529848575592, "rewards/format_reward/mean": 0.359375, "rewards/format_reward/std": 0.22658175230026245, "rewards/mcq_exact_match_reward/mean": 0.34375, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 8 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1013.0, "completions/max_terminated_length": 1013.0, "completions/mean_length": 308.84375, "completions/mean_terminated_length": 308.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0225, "frac_reward_zero_std": 0.125, "grad_norm": 15.022951126098633, "learning_rate": 9.975414512725056e-07, "loss": -0.0, "num_tokens": 974270.0, "reward": 0.49453121423721313, "reward_std": 0.32744407653808594, "rewards/format_reward/mean": 0.4140625, "rewards/format_reward/std": 0.19012710452079773, "rewards/mcq_exact_match_reward/mean": 0.453125, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 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": 1183.0, "completions/max_terminated_length": 1183.0, "completions/mean_length": 286.828125, "completions/mean_terminated_length": 286.828125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.025, "frac_reward_zero_std": 0.0, "grad_norm": 15.852792739868164, "learning_rate": 9.956320346634875e-07, "loss": -0.0, "num_tokens": 1078699.0, "reward": 0.3195312023162842, "reward_std": 0.3922736346721649, "rewards/format_reward/mean": 0.3828125, "rewards/format_reward/std": 0.2634054720401764, "rewards/mcq_exact_match_reward/mean": 0.28125, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 10 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1225.0, "completions/mean_length": 357.671875, "completions/mean_terminated_length": 330.8412780761719, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0275, "frac_reward_zero_std": 0.125, "grad_norm": 9.072699546813965, "learning_rate": 9.931806517013612e-07, "loss": -0.0, "num_tokens": 1184742.0, "reward": 0.4351561963558197, "reward_std": 0.40628623962402344, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.180765300989151, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 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": 1430.0, "completions/max_terminated_length": 1430.0, "completions/mean_length": 266.671875, "completions/mean_terminated_length": 266.671875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.03, "frac_reward_zero_std": 0.25, "grad_norm": 8.690740585327148, "learning_rate": 9.901899829374047e-07, "loss": -0.0, "num_tokens": 1288889.0, "reward": 0.4859374761581421, "reward_std": 0.30695563554763794, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.4375, "rewards/mcq_exact_match_reward/std": 0.5, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 810.0, "completions/mean_length": 209.5625, "completions/mean_terminated_length": 150.258056640625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0325, "frac_reward_zero_std": 0.25, "grad_norm": 16.069496154785156, "learning_rate": 9.866632986240029e-07, "loss": -0.0, "num_tokens": 1376117.0, "reward": 0.6921874284744263, "reward_std": 0.3327866792678833, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.23517554998397827, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "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": 1455.0, "completions/max_terminated_length": 1455.0, "completions/mean_length": 416.140625, "completions/mean_terminated_length": 416.140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.035, "frac_reward_zero_std": 0.375, "grad_norm": 5.533856391906738, "learning_rate": 9.826044551386742e-07, "loss": -0.0, "num_tokens": 1485550.0, "reward": 0.5640624761581421, "reward_std": 0.2898131012916565, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.125, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 14 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1879.0, "completions/max_terminated_length": 1879.0, "completions/mean_length": 278.84375, "completions/mean_terminated_length": 278.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0375, "frac_reward_zero_std": 0.125, "grad_norm": 9.030861854553223, "learning_rate": 9.780178907671788e-07, "loss": -0.0, "num_tokens": 1594252.0, "reward": 0.44218748807907104, "reward_std": 0.34988921880722046, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.1985812783241272, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 15 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1422.0, "completions/max_terminated_length": 1422.0, "completions/mean_length": 333.03125, "completions/mean_terminated_length": 333.03125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.04, "frac_reward_zero_std": 0.0, "grad_norm": 11.585250854492188, "learning_rate": 9.729086208503173e-07, "loss": -0.0, "num_tokens": 1703510.0, "reward": 0.2984374761581421, "reward_std": 0.32185253500938416, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.21764887869358063, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 768.0, "completions/max_terminated_length": 768.0, "completions/mean_length": 216.703125, "completions/mean_terminated_length": 216.703125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0425, "frac_reward_zero_std": 0.0, "grad_norm": 15.349666595458984, "learning_rate": 9.672822322997304e-07, "loss": -0.0, "num_tokens": 1793643.0, "reward": 0.3851562440395355, "reward_std": 0.35494449734687805, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.26528194546699524, "rewards/mcq_exact_match_reward/mean": 0.328125, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 17 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1036.0, "completions/max_terminated_length": 1036.0, "completions/mean_length": 218.65625, "completions/mean_terminated_length": 218.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.045, "frac_reward_zero_std": 0.125, "grad_norm": 14.17916488647461, "learning_rate": 9.611448774886923e-07, "loss": -0.0, "num_tokens": 1914501.0, "reward": 0.5640624761581421, "reward_std": 0.3361615836620331, "rewards/format_reward/mean": 0.484375, "rewards/format_reward/std": 0.17747680842876434, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1215.0, "completions/max_terminated_length": 1215.0, "completions/mean_length": 246.03125, "completions/mean_terminated_length": 246.03125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0475, "frac_reward_zero_std": 0.25, "grad_norm": 8.891441345214844, "learning_rate": 9.545032675245813e-07, "loss": -0.0, "num_tokens": 2008583.0, "reward": 0.5007811784744263, "reward_std": 0.2632700502872467, "rewards/format_reward/mean": 0.4765625, "rewards/format_reward/std": 0.1875000149011612, "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": 0.03125, "completions/max_length": 2048.0, "completions/max_terminated_length": 1567.0, "completions/mean_length": 348.375, "completions/mean_terminated_length": 293.5483703613281, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.05, "frac_reward_zero_std": 0.125, "grad_norm": 14.9910888671875, "learning_rate": 9.473646649103817e-07, "loss": 0.0, "num_tokens": 2132967.0, "reward": 0.2945312261581421, "reward_std": 0.18616712093353271, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.2203386276960373, "rewards/mcq_exact_match_reward/mean": 0.25, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 1420.0, "completions/max_terminated_length": 1420.0, "completions/mean_length": 290.40625, "completions/mean_terminated_length": 290.40625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0525, "frac_reward_zero_std": 0.25, "grad_norm": 9.075484275817871, "learning_rate": 9.397368756032444e-07, "loss": 0.0, "num_tokens": 2243057.0, "reward": 0.3218749761581421, "reward_std": 0.19791564345359802, "rewards/format_reward/mean": 0.5625, "rewards/format_reward/std": 0.208927720785141, "rewards/mcq_exact_match_reward/mean": 0.265625, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 699.0, "completions/max_terminated_length": 699.0, "completions/mean_length": 117.125, "completions/mean_terminated_length": 117.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.055, "frac_reward_zero_std": 0.5, "grad_norm": 16.495067596435547, "learning_rate": 9.316282404787869e-07, "loss": -0.0, "num_tokens": 2363217.0, "reward": 0.5992187261581421, "reward_std": 0.23310068249702454, "rewards/format_reward/mean": 0.5234375, "rewards/format_reward/std": 0.13886408507823944, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "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": 696.0, "completions/max_terminated_length": 696.0, "completions/mean_length": 114.03125, "completions/mean_terminated_length": 114.03125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0575, "frac_reward_zero_std": 0.25, "grad_norm": 25.43351936340332, "learning_rate": 9.230476262104676e-07, "loss": 0.0, "num_tokens": 2464699.0, "reward": 0.6007812023162842, "reward_std": 0.3113781809806824, "rewards/format_reward/mean": 0.5390625, "rewards/format_reward/std": 0.18483558297157288, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 23 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1635.0, "completions/max_terminated_length": 1635.0, "completions/mean_length": 105.3125, "completions/mean_terminated_length": 105.3125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.06, "frac_reward_zero_std": 0.125, "grad_norm": 22.568208694458008, "learning_rate": 9.1400441557401e-07, "loss": -0.0, "num_tokens": 2580535.0, "reward": 0.6695312261581421, "reward_std": 0.36984148621559143, "rewards/format_reward/mean": 0.4453125, "rewards/format_reward/std": 0.15728822350502014, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 627.0, "completions/max_terminated_length": 627.0, "completions/mean_length": 57.71875, "completions/mean_terminated_length": 57.71875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0625, "frac_reward_zero_std": 0.5, "grad_norm": 19.52919578552246, "learning_rate": 9.045084971874737e-07, "loss": 0.0, "num_tokens": 2698269.0, "reward": 0.34609371423721313, "reward_std": 0.19629064202308655, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.296875, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 25 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 1255.0, "completions/max_terminated_length": 1255.0, "completions/mean_length": 108.171875, "completions/mean_terminated_length": 108.171875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.065, "frac_reward_zero_std": 0.5, "grad_norm": 13.670044898986816, "learning_rate": 8.945702546981968e-07, "loss": 0.0, "num_tokens": 2791744.0, "reward": 0.5882812738418579, "reward_std": 0.17147785425186157, "rewards/format_reward/mean": 0.5703125, "rewards/format_reward/std": 0.21578919887542725, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 1056.0, "completions/mean_length": 214.453125, "completions/mean_terminated_length": 185.34921264648438, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0675, "frac_reward_zero_std": 0.25, "grad_norm": 10.611424446105957, "learning_rate": 8.842005554284295e-07, "loss": -0.0, "num_tokens": 2891421.0, "reward": 0.74609375, "reward_std": 0.26592713594436646, "rewards/format_reward/mean": 0.5859375, "rewards/format_reward/std": 0.2605654299259186, "rewards/mcq_exact_match_reward/mean": 0.6875, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 27 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 611.0, "completions/max_terminated_length": 611.0, "completions/mean_length": 104.203125, "completions/mean_terminated_length": 104.203125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.07, "frac_reward_zero_std": 0.625, "grad_norm": 21.459386825561523, "learning_rate": 8.734107384920769e-07, "loss": -0.0, "num_tokens": 2995762.0, "reward": 0.5671874284744263, "reward_std": 0.16976463794708252, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 2046.0, "completions/max_terminated_length": 2046.0, "completions/mean_length": 202.875, "completions/mean_terminated_length": 202.875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0725, "frac_reward_zero_std": 0.25, "grad_norm": 18.809906005859375, "learning_rate": 8.622126023955445e-07, "loss": -0.0, "num_tokens": 3106658.0, "reward": 0.539843738079071, "reward_std": 0.24837572872638702, "rewards/format_reward/mean": 0.5546875, "rewards/format_reward/std": 0.2538151443004608, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 1291.0, "completions/max_terminated_length": 1291.0, "completions/mean_length": 91.546875, "completions/mean_terminated_length": 91.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.075, "frac_reward_zero_std": 0.5, "grad_norm": 23.83727264404297, "learning_rate": 8.506183921362442e-07, "loss": 0.0, "num_tokens": 3205549.0, "reward": 0.5499999523162842, "reward_std": 0.1752062886953354, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.17817416787147522, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "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": 375.0, "completions/max_terminated_length": 375.0, "completions/mean_length": 19.21875, "completions/mean_terminated_length": 19.21875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0775, "frac_reward_zero_std": 0.75, "grad_norm": 7.594003200531006, "learning_rate": 8.386407858128706e-07, "loss": -0.0, "num_tokens": 3294923.0, "reward": 0.5679687261581421, "reward_std": 0.048349786549806595, "rewards/format_reward/mean": 0.5234375, "rewards/format_reward/std": 0.10652101784944534, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 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": 1506.0, "completions/max_terminated_length": 1506.0, "completions/mean_length": 74.6875, "completions/mean_terminated_length": 74.6875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.08, "frac_reward_zero_std": 0.5, "grad_norm": 22.89690399169922, "learning_rate": 8.262928807620843e-07, "loss": 0.0, "num_tokens": 3393431.0, "reward": 0.6609374284744263, "reward_std": 0.19887539744377136, "rewards/format_reward/mean": 0.515625, "rewards/format_reward/std": 0.08768405020236969, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 656.0, "completions/max_terminated_length": 656.0, "completions/mean_length": 63.1875, "completions/mean_terminated_length": 63.1875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0825, "frac_reward_zero_std": 0.875, "grad_norm": 5.9209513664245605, "learning_rate": 8.135881792367685e-07, "loss": 0.0, "num_tokens": 3480379.0, "reward": 0.7414062023162842, "reward_std": 0.0678049847483635, "rewards/format_reward/mean": 0.5390625, "rewards/format_reward/std": 0.16194961965084076, "rewards/mcq_exact_match_reward/mean": 0.6875, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 150.0, "completions/max_terminated_length": 150.0, "completions/mean_length": 8.25, "completions/mean_terminated_length": 8.25, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.085, "frac_reward_zero_std": 0.875, "grad_norm": 12.993154525756836, "learning_rate": 8.005405736415125e-07, "loss": 0.0, "num_tokens": 3579899.0, "reward": 0.5656249523162842, "reward_std": 0.04419416934251785, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 34 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 496.0, "completions/max_terminated_length": 496.0, "completions/mean_length": 84.78125, "completions/mean_terminated_length": 84.78125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0875, "frac_reward_zero_std": 0.875, "grad_norm": 14.276110649108887, "learning_rate": 7.871643313414718e-07, "loss": -0.0, "num_tokens": 3666677.0, "reward": 0.6437499523162842, "reward_std": 0.0578637532889843, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.59375, "rewards/mcq_exact_match_reward/std": 0.49501484632492065, "step": 35 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 546.0, "completions/max_terminated_length": 546.0, "completions/mean_length": 62.90625, "completions/mean_terminated_length": 62.90625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.09, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 7.734740790612136e-07, "loss": 0.0, "num_tokens": 3759271.0, "reward": 0.7999999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.75, "rewards/mcq_exact_match_reward/std": 0.4364357888698578, "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": 322.0, "completions/max_terminated_length": 322.0, "completions/mean_length": 37.765625, "completions/mean_terminated_length": 37.765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0925, "frac_reward_zero_std": 0.375, "grad_norm": 29.291526794433594, "learning_rate": 7.594847868906076e-07, "loss": -0.0, "num_tokens": 3865720.0, "reward": 0.5656249523162842, "reward_std": 0.3114553689956665, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "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": 0.015625, "completions/max_length": 2048.0, "completions/max_terminated_length": 500.0, "completions/mean_length": 106.46875, "completions/mean_terminated_length": 75.65079498291016, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.095, "frac_reward_zero_std": 0.5, "grad_norm": 19.967628479003906, "learning_rate": 7.452117519152541e-07, "loss": -0.0, "num_tokens": 3973950.0, "reward": 0.5304687023162842, "reward_std": 0.18365582823753357, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.13524486124515533, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 561.0, "completions/max_terminated_length": 561.0, "completions/mean_length": 39.5625, "completions/mean_terminated_length": 39.5625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.0975, "frac_reward_zero_std": 0.75, "grad_norm": 11.826974868774414, "learning_rate": 7.306705814893439e-07, "loss": 0.0, "num_tokens": 4069202.0, "reward": 0.4718749523162842, "reward_std": 0.10205791890621185, "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": 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": 727.0, "completions/max_terminated_length": 727.0, "completions/mean_length": 45.921875, "completions/mean_terminated_length": 45.921875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1, "frac_reward_zero_std": 0.625, "grad_norm": 33.241294860839844, "learning_rate": 7.158771761692464e-07, "loss": -0.0, "num_tokens": 4170741.0, "reward": 0.4093749523162842, "reward_std": 0.1804211586713791, "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": 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": 823.0, "completions/max_terminated_length": 823.0, "completions/mean_length": 59.265625, "completions/mean_terminated_length": 59.265625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1025, "frac_reward_zero_std": 0.625, "grad_norm": 17.169878005981445, "learning_rate": 7.008477123264847e-07, "loss": -0.0, "num_tokens": 4274854.0, "reward": 0.2796874940395355, "reward_std": 0.12545058131217957, "rewards/format_reward/mean": 0.453125, "rewards/format_reward/std": 0.14689241349697113, "rewards/mcq_exact_match_reward/mean": 0.234375, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "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": 623.0, "completions/max_terminated_length": 623.0, "completions/mean_length": 106.5, "completions/mean_terminated_length": 106.5, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.105, "frac_reward_zero_std": 0.875, "grad_norm": 12.11825942993164, "learning_rate": 6.855986244591103e-07, "loss": -0.0, "num_tokens": 4364534.0, "reward": 0.7687499523162842, "reward_std": 0.0578637532889843, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.71875, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "step": 42 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 459.0, "completions/max_terminated_length": 459.0, "completions/mean_length": 67.96875, "completions/mean_terminated_length": 67.96875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1075, "frac_reward_zero_std": 0.625, "grad_norm": 17.814006805419922, "learning_rate": 6.701465872208216e-07, "loss": -0.0, "num_tokens": 4460460.0, "reward": 0.5499999523162842, "reward_std": 0.1825428158044815, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 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": 618.0, "completions/max_terminated_length": 618.0, "completions/mean_length": 104.203125, "completions/mean_terminated_length": 104.203125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.11, "frac_reward_zero_std": 0.75, "grad_norm": 16.34816551208496, "learning_rate": 6.545084971874736e-07, "loss": -0.0, "num_tokens": 4565601.0, "reward": 0.7062499523162842, "reward_std": 0.12467905879020691, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.65625, "rewards/mcq_exact_match_reward/std": 0.4787135720252991, "step": 44 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 545.0, "completions/max_terminated_length": 545.0, "completions/mean_length": 99.53125, "completions/mean_terminated_length": 99.53125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1125, "frac_reward_zero_std": 0.875, "grad_norm": 5.380249500274658, "learning_rate": 6.387014543809223e-07, "loss": -0.0, "num_tokens": 4673595.0, "reward": 0.5812499523162842, "reward_std": 0.0578637532889843, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "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": 0.0, "completions/max_length": 304.0, "completions/max_terminated_length": 304.0, "completions/mean_length": 76.234375, "completions/mean_terminated_length": 76.234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.115, "frac_reward_zero_std": 0.5, "grad_norm": 14.324925422668457, "learning_rate": 6.227427435703995e-07, "loss": 0.0, "num_tokens": 4766802.0, "reward": 0.7843749523162842, "reward_std": 0.19044625759124756, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 927.0, "completions/max_terminated_length": 927.0, "completions/mean_length": 200.65625, "completions/mean_terminated_length": 200.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1175, "frac_reward_zero_std": 0.75, "grad_norm": 2.3821229934692383, "learning_rate": 6.066498153718734e-07, "loss": 0.0, "num_tokens": 4874484.0, "reward": 0.7374999523162842, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.6875, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "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": 528.0, "completions/max_terminated_length": 528.0, "completions/mean_length": 129.1875, "completions/mean_terminated_length": 129.1875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.12, "frac_reward_zero_std": 0.875, "grad_norm": 7.114138126373291, "learning_rate": 5.90440267166055e-07, "loss": 0.0, "num_tokens": 4971368.0, "reward": 0.9406249523162842, "reward_std": 0.04419416934251785, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.890625, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "step": 48 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 468.0, "completions/max_terminated_length": 468.0, "completions/mean_length": 98.65625, "completions/mean_terminated_length": 98.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1225, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 5.741318238559209e-07, "loss": 0.0, "num_tokens": 5067506.0, "reward": 0.9249999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.875, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 0.0, "completions/max_length": 795.0, "completions/max_terminated_length": 795.0, "completions/mean_length": 147.515625, "completions/mean_terminated_length": 147.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.125, "frac_reward_zero_std": 0.625, "grad_norm": 16.074617385864258, "learning_rate": 5.577423184847931e-07, "loss": -0.0, "num_tokens": 5184419.0, "reward": 0.7531249523162842, "reward_std": 0.15992167592048645, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.703125, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "step": 50 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 664.0, "completions/max_terminated_length": 664.0, "completions/mean_length": 69.0, "completions/mean_terminated_length": 69.0, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1275, "frac_reward_zero_std": 0.5, "grad_norm": 10.014348983764648, "learning_rate": 5.412896727361662e-07, "loss": 0.0, "num_tokens": 5298763.0, "reward": 0.6281249523162842, "reward_std": 0.19044625759124756, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.578125, "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": 0.0, "completions/max_length": 605.0, "completions/max_terminated_length": 605.0, "completions/mean_length": 198.21875, "completions/mean_terminated_length": 198.21875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.13, "frac_reward_zero_std": 0.625, "grad_norm": 8.749409675598145, "learning_rate": 5.247918773366111e-07, "loss": 0.0, "num_tokens": 5415705.0, "reward": 0.25312498211860657, "reward_std": 0.1530819833278656, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.203125, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "step": 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": 1417.0, "completions/max_terminated_length": 1417.0, "completions/mean_length": 136.421875, "completions/mean_terminated_length": 136.421875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1325, "frac_reward_zero_std": 0.75, "grad_norm": 7.897878646850586, "learning_rate": 5.082669723831793e-07, "loss": -0.0, "num_tokens": 5518652.0, "reward": 0.6281249523162842, "reward_std": 0.10205792635679245, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.578125, "rewards/mcq_exact_match_reward/std": 0.49776285886764526, "step": 53 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 606.0, "completions/max_terminated_length": 606.0, "completions/mean_length": 131.671875, "completions/mean_terminated_length": 131.671875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.135, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 4.917330276168208e-07, "loss": 0.0, "num_tokens": 5631223.0, "reward": 0.9249999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.875, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 1410.0, "completions/max_terminated_length": 1410.0, "completions/mean_length": 82.34375, "completions/mean_terminated_length": 82.34375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1375, "frac_reward_zero_std": 0.625, "grad_norm": 23.53481674194336, "learning_rate": 4.752081226633888e-07, "loss": 0.0, "num_tokens": 5732949.0, "reward": 0.5179687142372131, "reward_std": 0.13159699738025665, "rewards/format_reward/mean": 0.4921875, "rewards/format_reward/std": 0.0625, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 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": 799.0, "completions/mean_length": 164.328125, "completions/mean_terminated_length": 134.42857360839844, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.14, "frac_reward_zero_std": 0.75, "grad_norm": 1.8375636339187622, "learning_rate": 4.5871032726383385e-07, "loss": -0.0, "num_tokens": 5842530.0, "reward": 0.8624999523162842, "reward_std": 0.10888781398534775, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.8125, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "step": 56 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 380.0, "completions/max_terminated_length": 380.0, "completions/mean_length": 52.484375, "completions/mean_terminated_length": 52.484375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1425, "frac_reward_zero_std": 0.75, "grad_norm": 9.857487678527832, "learning_rate": 4.4225768151520694e-07, "loss": -0.0, "num_tokens": 5934169.0, "reward": 0.5656249523162842, "reward_std": 0.10205792635679245, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.515625, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "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": 826.0, "completions/max_terminated_length": 826.0, "completions/mean_length": 83.515625, "completions/mean_terminated_length": 83.515625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.145, "frac_reward_zero_std": 0.625, "grad_norm": 24.82467269897461, "learning_rate": 4.258681761440789e-07, "loss": 0.0, "num_tokens": 6036130.0, "reward": 0.6906249523162842, "reward_std": 0.13258251547813416, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 58 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 452.0, "completions/max_terminated_length": 452.0, "completions/mean_length": 53.078125, "completions/mean_terminated_length": 53.078125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1475, "frac_reward_zero_std": 0.5, "grad_norm": 29.634475708007812, "learning_rate": 4.095597328339452e-07, "loss": -0.0, "num_tokens": 6141695.0, "reward": 0.5812499523162842, "reward_std": 0.24511480331420898, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 506.0, "completions/max_terminated_length": 506.0, "completions/mean_length": 73.109375, "completions/mean_terminated_length": 73.109375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.15, "frac_reward_zero_std": 0.625, "grad_norm": 23.178203582763672, "learning_rate": 3.9335018462812664e-07, "loss": 0.0, "num_tokens": 6236950.0, "reward": 0.7843749523162842, "reward_std": 0.13258251547813416, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "step": 60 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 562.0, "completions/max_terminated_length": 562.0, "completions/mean_length": 62.15625, "completions/mean_terminated_length": 62.15625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1525, "frac_reward_zero_std": 0.875, "grad_norm": 10.943224906921387, "learning_rate": 3.772572564296004e-07, "loss": -0.0, "num_tokens": 6325400.0, "reward": 0.6593749523162842, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 61 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 627.0, "completions/max_terminated_length": 627.0, "completions/mean_length": 89.953125, "completions/mean_terminated_length": 89.953125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.155, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 3.612985456190778e-07, "loss": 0.0, "num_tokens": 6431429.0, "reward": 0.9249999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.875, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "step": 62 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 494.0, "completions/max_terminated_length": 494.0, "completions/mean_length": 96.796875, "completions/mean_terminated_length": 96.796875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1575, "frac_reward_zero_std": 0.625, "grad_norm": 18.425609588623047, "learning_rate": 3.454915028125263e-07, "loss": -0.0, "num_tokens": 6538240.0, "reward": 0.6124999523162842, "reward_std": 0.16675157845020294, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5625, "rewards/mcq_exact_match_reward/std": 0.5, "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": 331.0, "completions/max_terminated_length": 331.0, "completions/mean_length": 81.953125, "completions/mean_terminated_length": 81.953125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.16, "frac_reward_zero_std": 0.75, "grad_norm": 13.189492225646973, "learning_rate": 3.2985341277917846e-07, "loss": 0.0, "num_tokens": 6618469.0, "reward": 0.6749999523162842, "reward_std": 0.0883883386850357, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 360.0, "completions/max_terminated_length": 360.0, "completions/mean_length": 39.796875, "completions/mean_terminated_length": 39.796875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1625, "frac_reward_zero_std": 0.5, "grad_norm": 44.135902404785156, "learning_rate": 3.1440137554088953e-07, "loss": -0.0, "num_tokens": 6707040.0, "reward": 0.5187499523162842, "reward_std": 0.22461532056331635, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 65 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 539.0, "completions/max_terminated_length": 539.0, "completions/mean_length": 52.625, "completions/mean_terminated_length": 52.625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.165, "frac_reward_zero_std": 0.75, "grad_norm": 29.948780059814453, "learning_rate": 2.9915228767351535e-07, "loss": -0.0, "num_tokens": 6800072.0, "reward": 0.9406249523162842, "reward_std": 0.12255740165710449, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.890625, "rewards/mcq_exact_match_reward/std": 0.3145764470100403, "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": 830.0, "completions/max_terminated_length": 830.0, "completions/mean_length": 71.75, "completions/mean_terminated_length": 71.75, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1675, "frac_reward_zero_std": 0.875, "grad_norm": 12.17775821685791, "learning_rate": 2.841228238307536e-07, "loss": 0.0, "num_tokens": 6914024.0, "reward": 0.4406249523162842, "reward_std": 0.04419416934251785, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 67 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 536.0, "completions/max_terminated_length": 536.0, "completions/mean_length": 89.359375, "completions/mean_terminated_length": 89.359375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.17, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 2.6932941851065615e-07, "loss": 0.0, "num_tokens": 6989311.0, "reward": 0.9249999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.875, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 248.0, "completions/max_terminated_length": 248.0, "completions/mean_length": 65.75, "completions/mean_terminated_length": 65.75, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1725, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 2.547882480847461e-07, "loss": 0.0, "num_tokens": 7090607.0, "reward": 0.9249999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.875, "rewards/mcq_exact_match_reward/std": 0.3333333432674408, "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": 411.0, "completions/max_terminated_length": 411.0, "completions/mean_length": 83.4375, "completions/mean_terminated_length": 83.4375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.175, "frac_reward_zero_std": 0.5, "grad_norm": 31.62489891052246, "learning_rate": 2.4051521310939254e-07, "loss": 0.0, "num_tokens": 7197507.0, "reward": 0.7648437023162842, "reward_std": 0.22706486284732819, "rewards/format_reward/mean": 0.4609375, "rewards/format_reward/std": 0.13524486124515533, "rewards/mcq_exact_match_reward/mean": 0.71875, "rewards/mcq_exact_match_reward/std": 0.4531635046005249, "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": 419.0, "completions/max_terminated_length": 419.0, "completions/mean_length": 68.9375, "completions/mean_terminated_length": 68.9375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1775, "frac_reward_zero_std": 0.75, "grad_norm": 8.030557632446289, "learning_rate": 2.2652592093878665e-07, "loss": -0.0, "num_tokens": 7288655.0, "reward": 0.5968749523162842, "reward_std": 0.12255740165710449, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.546875, "rewards/mcq_exact_match_reward/std": 0.501733124256134, "step": 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": 500.0, "completions/max_terminated_length": 500.0, "completions/mean_length": 99.546875, "completions/mean_terminated_length": 99.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.18, "frac_reward_zero_std": 0.875, "grad_norm": 0.696971595287323, "learning_rate": 2.128356686585282e-07, "loss": 0.0, "num_tokens": 7375754.0, "reward": 0.6593749523162842, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 299.0, "completions/max_terminated_length": 299.0, "completions/mean_length": 52.859375, "completions/mean_terminated_length": 52.859375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1825, "frac_reward_zero_std": 0.75, "grad_norm": 18.39797592163086, "learning_rate": 1.9945942635848745e-07, "loss": -0.0, "num_tokens": 7473969.0, "reward": 0.9718749523162842, "reward_std": 0.12255740165710449, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.921875, "rewards/mcq_exact_match_reward/std": 0.27048972249031067, "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": 458.0, "completions/max_terminated_length": 458.0, "completions/mean_length": 63.8125, "completions/mean_terminated_length": 63.8125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.185, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 1.8641182076323148e-07, "loss": 0.0, "num_tokens": 7565637.0, "reward": 0.6749999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 74 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 664.0, "completions/max_terminated_length": 664.0, "completions/mean_length": 81.140625, "completions/mean_terminated_length": 81.140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1875, "frac_reward_zero_std": 0.75, "grad_norm": 21.489856719970703, "learning_rate": 1.7370711923791564e-07, "loss": -0.0, "num_tokens": 7658422.0, "reward": 0.9562499523162842, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.90625, "rewards/mcq_exact_match_reward/std": 0.29378482699394226, "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": 532.0, "completions/max_terminated_length": 532.0, "completions/mean_length": 105.5625, "completions/mean_terminated_length": 105.5625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.19, "frac_reward_zero_std": 0.625, "grad_norm": 6.992222785949707, "learning_rate": 1.6135921418712955e-07, "loss": 0.0, "num_tokens": 7759138.0, "reward": 0.43437498807907104, "reward_std": 0.15992167592048645, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.390625, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 76 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 335.0, "completions/max_terminated_length": 335.0, "completions/mean_length": 39.5, "completions/mean_terminated_length": 39.5, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1925, "frac_reward_zero_std": 0.75, "grad_norm": 19.49919891357422, "learning_rate": 1.493816078637557e-07, "loss": -0.0, "num_tokens": 7850794.0, "reward": 0.7218749523162842, "reward_std": 0.11100947856903076, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.671875, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 77 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 906.0, "completions/max_terminated_length": 906.0, "completions/mean_length": 108.796875, "completions/mean_terminated_length": 108.796875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.195, "frac_reward_zero_std": 0.875, "grad_norm": 1.4245766401290894, "learning_rate": 1.3778739760445552e-07, "loss": -0.0, "num_tokens": 7954501.0, "reward": 0.7843749523162842, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 374.0, "completions/max_terminated_length": 374.0, "completions/mean_length": 34.84375, "completions/mean_terminated_length": 34.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.1975, "frac_reward_zero_std": 0.875, "grad_norm": 5.065796852111816, "learning_rate": 1.2658926150792322e-07, "loss": -0.0, "num_tokens": 8053659.0, "reward": 0.8312499523162842, "reward_std": 0.0578637532889843, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.78125, "rewards/mcq_exact_match_reward/std": 0.4166666865348816, "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": 657.0, "completions/max_terminated_length": 657.0, "completions/mean_length": 84.140625, "completions/mean_terminated_length": 84.140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2, "frac_reward_zero_std": 0.875, "grad_norm": 2.4845361709594727, "learning_rate": 1.1579944457157059e-07, "loss": -0.0, "num_tokens": 8134372.0, "reward": 1.0343749523162842, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.984375, "rewards/mcq_exact_match_reward/std": 0.125, "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": 913.0, "completions/max_terminated_length": 913.0, "completions/mean_length": 127.09375, "completions/mean_terminated_length": 127.09375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2025, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 1.0542974530180327e-07, "loss": 0.0, "num_tokens": 8233274.0, "reward": 0.5499999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.5, "rewards/mcq_exact_match_reward/std": 0.5039526224136353, "step": 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": 598.0, "completions/max_terminated_length": 598.0, "completions/mean_length": 124.234375, "completions/mean_terminated_length": 124.234375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.205, "frac_reward_zero_std": 0.75, "grad_norm": 8.355081558227539, "learning_rate": 9.549150281252632e-08, "loss": -0.0, "num_tokens": 8325609.0, "reward": 0.5187499523162842, "reward_std": 0.10888781398534775, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.46875, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "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": 688.0, "completions/max_terminated_length": 688.0, "completions/mean_length": 49.9375, "completions/mean_terminated_length": 49.9375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2075, "frac_reward_zero_std": 0.625, "grad_norm": 18.9002742767334, "learning_rate": 8.599558442598998e-08, "loss": -0.0, "num_tokens": 8443061.0, "reward": 0.6593749523162842, "reward_std": 0.13258251547813416, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "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": 709.0, "completions/max_terminated_length": 709.0, "completions/mean_length": 147.140625, "completions/mean_terminated_length": 147.140625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.21, "frac_reward_zero_std": 0.75, "grad_norm": 6.830162525177002, "learning_rate": 7.695237378953224e-08, "loss": -0.0, "num_tokens": 8537230.0, "reward": 0.8156249523162842, "reward_std": 0.10205792635679245, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.765625, "rewards/mcq_exact_match_reward/std": 0.42695629596710205, "step": 84 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 283.0, "completions/max_terminated_length": 283.0, "completions/mean_length": 55.84375, "completions/mean_terminated_length": 55.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2125, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 6.837175952121304e-08, "loss": 0.0, "num_tokens": 8617452.0, "reward": 0.6749999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "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": 755.0, "completions/max_terminated_length": 755.0, "completions/mean_length": 104.546875, "completions/mean_terminated_length": 104.546875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.215, "frac_reward_zero_std": 0.625, "grad_norm": 27.913728713989258, "learning_rate": 6.026312439675551e-08, "loss": -0.0, "num_tokens": 8714511.0, "reward": 0.7374999523162842, "reward_std": 0.1825428158044815, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.6875, "rewards/mcq_exact_match_reward/std": 0.467176616191864, "step": 86 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 549.0, "completions/max_terminated_length": 549.0, "completions/mean_length": 83.453125, "completions/mean_terminated_length": 83.453125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2175, "frac_reward_zero_std": 0.75, "grad_norm": 29.583234786987305, "learning_rate": 5.263533508961826e-08, "loss": 0.0, "num_tokens": 8804092.0, "reward": 0.7687499523162842, "reward_std": 0.10888781398534775, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "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": 1104.0, "completions/max_terminated_length": 1104.0, "completions/mean_length": 131.484375, "completions/mean_terminated_length": 131.484375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.22, "frac_reward_zero_std": 0.75, "grad_norm": 9.109688758850098, "learning_rate": 4.549673247541874e-08, "loss": -0.0, "num_tokens": 8911835.0, "reward": 0.5343749523162842, "reward_std": 0.1315089464187622, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.484375, "rewards/mcq_exact_match_reward/std": 0.5037065148353577, "step": 88 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 567.0, "completions/max_terminated_length": 567.0, "completions/mean_length": 83.53125, "completions/mean_terminated_length": 83.53125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2225, "frac_reward_zero_std": 0.875, "grad_norm": 20.39177703857422, "learning_rate": 3.8855122511307626e-08, "loss": -0.0, "num_tokens": 8995933.0, "reward": 0.7468749284744263, "reward_std": 0.0646936446428299, "rewards/format_reward/mean": 0.4375, "rewards/format_reward/std": 0.1666666716337204, "rewards/mcq_exact_match_reward/mean": 0.703125, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 248.0, "completions/max_terminated_length": 248.0, "completions/mean_length": 27.4375, "completions/mean_terminated_length": 27.4375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.225, "frac_reward_zero_std": 0.5, "grad_norm": 28.867332458496094, "learning_rate": 3.271776770026963e-08, "loss": -0.0, "num_tokens": 9112537.0, "reward": 0.8468749523162842, "reward_std": 0.19044627249240875, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.796875, "rewards/mcq_exact_match_reward/std": 0.40550529956817627, "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": 173.0, "completions/max_terminated_length": 173.0, "completions/mean_length": 30.375, "completions/mean_terminated_length": 30.375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2275, "frac_reward_zero_std": 0.875, "grad_norm": 5.908293724060059, "learning_rate": 2.7091379149682682e-08, "loss": 0.0, "num_tokens": 9214753.0, "reward": 0.8781249523162842, "reward_std": 0.0646936446428299, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.828125, "rewards/mcq_exact_match_reward/std": 0.38025420904159546, "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": 581.0, "completions/max_terminated_length": 581.0, "completions/mean_length": 79.765625, "completions/mean_terminated_length": 79.765625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.23, "frac_reward_zero_std": 0.75, "grad_norm": 19.826080322265625, "learning_rate": 2.1982109232821176e-08, "loss": -0.0, "num_tokens": 9319474.0, "reward": 0.5812499523162842, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.53125, "rewards/mcq_exact_match_reward/std": 0.5029674172401428, "step": 92 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 354.0, "completions/max_terminated_length": 354.0, "completions/mean_length": 59.921875, "completions/mean_terminated_length": 59.921875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2325, "frac_reward_zero_std": 0.625, "grad_norm": 11.2923583984375, "learning_rate": 1.7395544861325718e-08, "loss": -0.0, "num_tokens": 9417053.0, "reward": 0.7218749523162842, "reward_std": 0.1530819833278656, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.671875, "rewards/mcq_exact_match_reward/std": 0.4732423722743988, "step": 93 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 372.0, "completions/max_terminated_length": 372.0, "completions/mean_length": 36.125, "completions/mean_terminated_length": 36.125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.235, "frac_reward_zero_std": 0.75, "grad_norm": 10.971405029296875, "learning_rate": 1.3336701375997127e-08, "loss": -0.0, "num_tokens": 9510501.0, "reward": 0.7531249523162842, "reward_std": 0.10205792635679245, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.703125, "rewards/mcq_exact_match_reward/std": 0.4604927599430084, "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": 491.0, "completions/max_terminated_length": 491.0, "completions/mean_length": 130.46875, "completions/mean_terminated_length": 130.46875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2375, "frac_reward_zero_std": 0.75, "grad_norm": 1.9149724245071411, "learning_rate": 9.810017062595321e-09, "loss": -0.0, "num_tokens": 9595035.0, "reward": 0.8624999523162842, "reward_std": 0.10888782143592834, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.8125, "rewards/mcq_exact_match_reward/std": 0.39339789748191833, "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": 382.0, "completions/max_terminated_length": 382.0, "completions/mean_length": 75.1875, "completions/mean_terminated_length": 75.1875, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.24, "frac_reward_zero_std": 0.5, "grad_norm": 19.724637985229492, "learning_rate": 6.819348298638839e-09, "loss": -0.0, "num_tokens": 9706103.0, "reward": 0.6593749523162842, "reward_std": 0.2472364455461502, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.609375, "rewards/mcq_exact_match_reward/std": 0.4917473793029785, "step": 96 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 824.0, "completions/max_terminated_length": 824.0, "completions/mean_length": 114.0625, "completions/mean_terminated_length": 114.0625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2425, "frac_reward_zero_std": 0.75, "grad_norm": 10.95627212524414, "learning_rate": 4.367965336512403e-09, "loss": 0.0, "num_tokens": 9807051.0, "reward": 0.42499998211860657, "reward_std": 0.0883883386850357, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.375, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 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": 837.0, "completions/max_terminated_length": 837.0, "completions/mean_length": 125.703125, "completions/mean_terminated_length": 125.703125, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.245, "frac_reward_zero_std": 1.0, "grad_norm": 0.0, "learning_rate": 2.458548727494292e-09, "loss": 0.0, "num_tokens": 9900232.0, "reward": 0.6749999523162842, "reward_std": 0.0, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.625, "rewards/mcq_exact_match_reward/std": 0.48795005679130554, "step": 98 }, { "clip_ratio/high_max": 0.0, "clip_ratio/high_mean": 0.0, "clip_ratio/low_mean": 0.0, "clip_ratio/low_min": 0.0, "clip_ratio/region_mean": 0.0, "completions/clipped_ratio": 0.0, "completions/max_length": 301.0, "completions/max_terminated_length": 301.0, "completions/mean_length": 33.65625, "completions/mean_terminated_length": 33.65625, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.2475, "frac_reward_zero_std": 0.875, "grad_norm": 12.734614372253418, "learning_rate": 1.0931863906127325e-09, "loss": -0.0, "num_tokens": 9988786.0, "reward": 0.7843749523162842, "reward_std": 0.04419417306780815, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.734375, "rewards/mcq_exact_match_reward/std": 0.44515693187713623, "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": 434.0, "completions/max_terminated_length": 434.0, "completions/mean_length": 51.84375, "completions/mean_terminated_length": 51.84375, "completions/min_length": 6.0, "completions/min_terminated_length": 6.0, "epoch": 0.25, "frac_reward_zero_std": 0.875, "grad_norm": 1.3988230228424072, "learning_rate": 2.733713295369755e-10, "loss": 0.0, "num_tokens": 10087824.0, "reward": 0.6906249523162842, "reward_std": 0.04419416934251785, "rewards/format_reward/mean": 0.5, "rewards/format_reward/std": 0.0, "rewards/mcq_exact_match_reward/mean": 0.640625, "rewards/mcq_exact_match_reward/std": 0.4836103618144989, "step": 100 } ], "logging_steps": 1, "max_steps": 100, "num_input_tokens_seen": 10087824, "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 }