| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9996190476190476, | |
| "eval_steps": 50, | |
| "global_step": 164, | |
| "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": 0.76337890625, | |
| "completions/max_length": 128.0, | |
| "completions/max_terminated_length": 128.0, | |
| "completions/mean_length": 116.70947265625, | |
| "completions/mean_terminated_length": 80.29782409667969, | |
| "completions/min_length": 2.0, | |
| "completions/min_terminated_length": 2.0, | |
| "epoch": 0.030476190476190476, | |
| "grad_norm": 0.48198601603507996, | |
| "learning_rate": 5.555555555555555e-07, | |
| "loss": 0.3641, | |
| "num_tokens": 3784033.0, | |
| "reward": 1.6605166435241698, | |
| "reward_std": 3.4095802307128906, | |
| "rewards/accuracy_reward": 0.03017578125, | |
| "rewards/brier_reward": 0.130406653881073, | |
| "rewards/confidence_one_or_zero": 0.105078125, | |
| "rewards/format_reward": 0.1703125, | |
| "rewards/log_2_reward": 1.5602725982666015, | |
| "rewards/mean_confidence_reward": 0.10646367371082306, | |
| "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": 0.42197265625, | |
| "completions/max_length": 128.0, | |
| "completions/max_terminated_length": 128.0, | |
| "completions/mean_length": 97.47744140625, | |
| "completions/mean_terminated_length": 76.52580413818359, | |
| "completions/min_length": 5.2, | |
| "completions/min_terminated_length": 5.2, | |
| "epoch": 0.06095238095238095, | |
| "grad_norm": 0.38170281052589417, | |
| "learning_rate": 1e-06, | |
| "loss": 0.7106, | |
| "num_tokens": 7371066.0, | |
| "reward": 5.354305839538574, | |
| "reward_std": 4.306017875671387, | |
| "rewards/accuracy_reward": 0.12158203125, | |
| "rewards/brier_reward": 0.41428526639938357, | |
| "rewards/confidence_one_or_zero": 0.2091796875, | |
| "rewards/format_reward": 0.54599609375, | |
| "rewards/log_2_reward": 5.020516633987427, | |
| "rewards/mean_confidence_reward": 0.27076671421527865, | |
| "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": 0.04169921875, | |
| "completions/max_length": 128.0, | |
| "completions/max_terminated_length": 124.4, | |
| "completions/mean_length": 63.0056640625, | |
| "completions/mean_terminated_length": 60.36174392700195, | |
| "completions/min_length": 16.4, | |
| "completions/min_terminated_length": 16.4, | |
| "epoch": 0.09142857142857143, | |
| "grad_norm": 0.1799696832895279, | |
| "learning_rate": 1e-06, | |
| "loss": 0.2204, | |
| "num_tokens": 10604244.0, | |
| "reward": 9.318659210205078, | |
| "reward_std": 2.0989513874053953, | |
| "rewards/accuracy_reward": 0.19326171875, | |
| "rewards/brier_reward": 0.7086999773979187, | |
| "rewards/confidence_one_or_zero": 0.30234375, | |
| "rewards/format_reward": 0.95224609375, | |
| "rewards/log_2_reward": 8.745905876159668, | |
| "rewards/mean_confidence_reward": 0.4371440947055817, | |
| "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": 0.000390625, | |
| "completions/max_length": 112.0, | |
| "completions/max_terminated_length": 99.6, | |
| "completions/mean_length": 43.30595703125, | |
| "completions/mean_terminated_length": 43.27299423217774, | |
| "completions/min_length": 17.6, | |
| "completions/min_terminated_length": 17.6, | |
| "epoch": 0.1219047619047619, | |
| "grad_norm": 0.0386660136282444, | |
| "learning_rate": 1e-06, | |
| "loss": -0.0043, | |
| "num_tokens": 13634321.0, | |
| "reward": 10.072226905822754, | |
| "reward_std": 0.6315600275993347, | |
| "rewards/accuracy_reward": 0.16630859375, | |
| "rewards/brier_reward": 0.8290799617767334, | |
| "rewards/confidence_one_or_zero": 0.44169921875, | |
| "rewards/format_reward": 0.9970703125, | |
| "rewards/log_2_reward": 9.490537071228028, | |
| "rewards/mean_confidence_reward": 0.31124078631401064, | |
| "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": 90.0, | |
| "completions/max_terminated_length": 90.0, | |
| "completions/mean_length": 42.37373046875, | |
| "completions/mean_terminated_length": 42.37373046875, | |
| "completions/min_length": 22.0, | |
| "completions/min_terminated_length": 22.0, | |
| "epoch": 0.1523809523809524, | |
| "grad_norm": 0.013290881179273129, | |
| "learning_rate": 1e-06, | |
| "loss": -0.001, | |
| "num_tokens": 16657508.0, | |
| "reward": 10.345937538146973, | |
| "reward_std": 0.27599466145038604, | |
| "rewards/accuracy_reward": 0.11845703125, | |
| "rewards/brier_reward": 0.9262194037437439, | |
| "rewards/confidence_one_or_zero": 0.70439453125, | |
| "rewards/format_reward": 0.99912109375, | |
| "rewards/log_2_reward": 9.787148475646973, | |
| "rewards/mean_confidence_reward": 0.13568951636552812, | |
| "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": 77.6, | |
| "completions/max_terminated_length": 77.6, | |
| "completions/mean_length": 40.87314453125, | |
| "completions/mean_terminated_length": 40.87314453125, | |
| "completions/min_length": 24.0, | |
| "completions/min_terminated_length": 24.0, | |
| "epoch": 0.18285714285714286, | |
| "grad_norm": 0.01713799685239792, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0003, | |
| "num_tokens": 19663153.0, | |
| "reward": 10.48781967163086, | |
| "reward_std": 0.0398749703541398, | |
| "rewards/accuracy_reward": 0.00849609375, | |
| "rewards/brier_reward": 0.9944719195365905, | |
| "rewards/confidence_one_or_zero": 0.98203125, | |
| "rewards/format_reward": 0.9998046875, | |
| "rewards/log_2_reward": 9.98366928100586, | |
| "rewards/mean_confidence_reward": 0.005668945552315563, | |
| "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": 60.6, | |
| "completions/max_terminated_length": 60.6, | |
| "completions/mean_length": 40.10283203125, | |
| "completions/mean_terminated_length": 40.10283203125, | |
| "completions/min_length": 24.2, | |
| "completions/min_terminated_length": 24.2, | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 22658990.0, | |
| "reward": 10.499842834472656, | |
| "reward_std": 0.0008906088653020561, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.9999728918075561, | |
| "rewards/confidence_one_or_zero": 0.99912109375, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 9.999842834472656, | |
| "rewards/mean_confidence_reward": 0.00014160156424622983, | |
| "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": 60.4, | |
| "completions/max_terminated_length": 60.4, | |
| "completions/mean_length": 39.8552734375, | |
| "completions/mean_terminated_length": 39.8552734375, | |
| "completions/min_length": 25.8, | |
| "completions/min_terminated_length": 25.8, | |
| "epoch": 0.2438095238095238, | |
| "grad_norm": 9.549452079227194e-05, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0001, | |
| "num_tokens": 25658020.0, | |
| "reward": 10.498812484741212, | |
| "reward_std": 0.006671252334490419, | |
| "rewards/accuracy_reward": 9.765625e-05, | |
| "rewards/brier_reward": 0.9998310446739197, | |
| "rewards/confidence_one_or_zero": 0.999609375, | |
| "rewards/format_reward": 0.99990234375, | |
| "rewards/log_2_reward": 9.998812484741212, | |
| "rewards/mean_confidence_reward": 6.83593752910383e-05, | |
| "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": 58.2, | |
| "completions/max_terminated_length": 58.2, | |
| "completions/mean_length": 39.4888671875, | |
| "completions/mean_terminated_length": 39.4888671875, | |
| "completions/min_length": 26.2, | |
| "completions/min_terminated_length": 26.2, | |
| "epoch": 0.2742857142857143, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": -0.0001, | |
| "num_tokens": 28652306.0, | |
| "reward": 10.498974609375, | |
| "reward_std": 0.005800485238432884, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.99990234375, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 0.99990234375, | |
| "rewards/log_2_reward": 9.9990234375, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "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": 56.6, | |
| "completions/max_terminated_length": 56.6, | |
| "completions/mean_length": 39.37822265625, | |
| "completions/mean_terminated_length": 39.37822265625, | |
| "completions/min_length": 26.6, | |
| "completions/min_terminated_length": 26.6, | |
| "epoch": 0.3047619047619048, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": -0.0, | |
| "num_tokens": 31639059.0, | |
| "reward": 10.498974609375, | |
| "reward_std": 0.005800485238432884, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.99990234375, | |
| "rewards/confidence_one_or_zero": 0.99990234375, | |
| "rewards/format_reward": 0.99990234375, | |
| "rewards/log_2_reward": 9.9990234375, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3047619047619048, | |
| "eval_completions/clipped_ratio": 0.0, | |
| "eval_completions/max_length": 48.25, | |
| "eval_completions/max_terminated_length": 48.25, | |
| "eval_completions/mean_length": 39.21484375, | |
| "eval_completions/mean_terminated_length": 39.21484375, | |
| "eval_completions/min_length": 29.25, | |
| "eval_completions/min_terminated_length": 29.25, | |
| "eval_loss": 0.0, | |
| "eval_num_tokens": 31639059.0, | |
| "eval_reward": 10.5, | |
| "eval_reward_std": 0.0, | |
| "eval_rewards/accuracy_reward": 0.0, | |
| "eval_rewards/brier_reward": 1.0, | |
| "eval_rewards/confidence_one_or_zero": 1.0, | |
| "eval_rewards/format_reward": 1.0, | |
| "eval_rewards/log_2_reward": 10.0, | |
| "eval_rewards/mean_confidence_reward": 0.0, | |
| "eval_runtime": 7.2463, | |
| "eval_samples_per_second": 69.001, | |
| "eval_steps_per_second": 0.552, | |
| "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": 55.0, | |
| "completions/max_terminated_length": 55.0, | |
| "completions/mean_length": 39.50654296875, | |
| "completions/mean_terminated_length": 39.50654296875, | |
| "completions/min_length": 26.0, | |
| "completions/min_terminated_length": 26.0, | |
| "epoch": 0.3352380952380952, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 34630902.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 55.4, | |
| "completions/max_terminated_length": 55.4, | |
| "completions/mean_length": 39.666015625, | |
| "completions/mean_terminated_length": 39.666015625, | |
| "completions/min_length": 26.8, | |
| "completions/min_terminated_length": 26.8, | |
| "epoch": 0.3657142857142857, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 37627098.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "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": 55.4, | |
| "completions/max_terminated_length": 55.4, | |
| "completions/mean_length": 39.70166015625, | |
| "completions/mean_terminated_length": 39.70166015625, | |
| "completions/min_length": 25.0, | |
| "completions/min_terminated_length": 25.0, | |
| "epoch": 0.3961904761904762, | |
| "grad_norm": 0.004820019006729126, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0002, | |
| "num_tokens": 40622155.0, | |
| "reward": 10.498974609375, | |
| "reward_std": 0.005800485238432884, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.99990234375, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 0.99990234375, | |
| "rewards/log_2_reward": 9.9990234375, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "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": 56.2, | |
| "completions/max_terminated_length": 56.2, | |
| "completions/mean_length": 39.75986328125, | |
| "completions/mean_terminated_length": 39.75986328125, | |
| "completions/min_length": 25.2, | |
| "completions/min_terminated_length": 25.2, | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 43618960.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "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": 57.2, | |
| "completions/max_terminated_length": 57.2, | |
| "completions/mean_length": 39.5822265625, | |
| "completions/mean_terminated_length": 39.5822265625, | |
| "completions/min_length": 25.2, | |
| "completions/min_terminated_length": 25.2, | |
| "epoch": 0.45714285714285713, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": -0.0002, | |
| "num_tokens": 46609594.0, | |
| "reward": 10.498974609375, | |
| "reward_std": 0.005800485238432884, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.99990234375, | |
| "rewards/confidence_one_or_zero": 0.99990234375, | |
| "rewards/format_reward": 0.99990234375, | |
| "rewards/log_2_reward": 9.9990234375, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "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": 55.2, | |
| "completions/max_terminated_length": 55.2, | |
| "completions/mean_length": 39.87529296875, | |
| "completions/mean_terminated_length": 39.87529296875, | |
| "completions/min_length": 27.0, | |
| "completions/min_terminated_length": 27.0, | |
| "epoch": 0.4876190476190476, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0001, | |
| "num_tokens": 49602237.0, | |
| "reward": 10.498974609375, | |
| "reward_std": 0.005800485238432884, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.99990234375, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 0.99990234375, | |
| "rewards/log_2_reward": 9.9990234375, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "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": 59.0, | |
| "completions/max_terminated_length": 59.0, | |
| "completions/mean_length": 40.1205078125, | |
| "completions/mean_terminated_length": 40.1205078125, | |
| "completions/min_length": 27.2, | |
| "completions/min_terminated_length": 27.2, | |
| "epoch": 0.518095238095238, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 52598607.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "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": 59.2, | |
| "completions/max_terminated_length": 59.2, | |
| "completions/mean_length": 40.13505859375, | |
| "completions/mean_terminated_length": 40.13505859375, | |
| "completions/min_length": 25.8, | |
| "completions/min_terminated_length": 25.8, | |
| "epoch": 0.5485714285714286, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 55596438.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "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": 56.8, | |
| "completions/max_terminated_length": 56.8, | |
| "completions/mean_length": 40.24052734375, | |
| "completions/mean_terminated_length": 40.24052734375, | |
| "completions/min_length": 27.8, | |
| "completions/min_terminated_length": 27.8, | |
| "epoch": 0.579047619047619, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 58594133.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "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": 56.6, | |
| "completions/max_terminated_length": 56.6, | |
| "completions/mean_length": 40.23056640625, | |
| "completions/mean_terminated_length": 40.23056640625, | |
| "completions/min_length": 26.0, | |
| "completions/min_terminated_length": 26.0, | |
| "epoch": 0.6095238095238096, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 61594062.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6095238095238096, | |
| "eval_completions/clipped_ratio": 0.0, | |
| "eval_completions/max_length": 50.75, | |
| "eval_completions/max_terminated_length": 50.75, | |
| "eval_completions/mean_length": 40.36058712005615, | |
| "eval_completions/mean_terminated_length": 40.36058712005615, | |
| "eval_completions/min_length": 29.5, | |
| "eval_completions/min_terminated_length": 29.5, | |
| "eval_loss": 0.0, | |
| "eval_num_tokens": 61594062.0, | |
| "eval_reward": 10.5, | |
| "eval_reward_std": 0.0, | |
| "eval_rewards/accuracy_reward": 0.0, | |
| "eval_rewards/brier_reward": 1.0, | |
| "eval_rewards/confidence_one_or_zero": 1.0, | |
| "eval_rewards/format_reward": 1.0, | |
| "eval_rewards/log_2_reward": 10.0, | |
| "eval_rewards/mean_confidence_reward": 0.0, | |
| "eval_runtime": 7.5851, | |
| "eval_samples_per_second": 65.919, | |
| "eval_steps_per_second": 0.527, | |
| "step": 100 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 58.0, | |
| "completions/max_terminated_length": 58.0, | |
| "completions/mean_length": 40.31025390625, | |
| "completions/mean_terminated_length": 40.31025390625, | |
| "completions/min_length": 27.4, | |
| "completions/min_terminated_length": 27.4, | |
| "epoch": 0.64, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 64598391.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 105 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 57.2, | |
| "completions/max_terminated_length": 57.2, | |
| "completions/mean_length": 40.2923828125, | |
| "completions/mean_terminated_length": 40.2923828125, | |
| "completions/min_length": 27.4, | |
| "completions/min_terminated_length": 27.4, | |
| "epoch": 0.6704761904761904, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 67599081.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 58.2, | |
| "completions/max_terminated_length": 58.2, | |
| "completions/mean_length": 40.179296875, | |
| "completions/mean_terminated_length": 40.179296875, | |
| "completions/min_length": 27.4, | |
| "completions/min_terminated_length": 27.4, | |
| "epoch": 0.700952380952381, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 70597301.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 115 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 57.2, | |
| "completions/max_terminated_length": 57.2, | |
| "completions/mean_length": 40.3193359375, | |
| "completions/mean_terminated_length": 40.3193359375, | |
| "completions/min_length": 26.8, | |
| "completions/min_terminated_length": 26.8, | |
| "epoch": 0.7314285714285714, | |
| "grad_norm": 0.024884896352887154, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 73598459.0, | |
| "reward": 10.498974609375, | |
| "reward_std": 0.005800485238432884, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.99990234375, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 0.99990234375, | |
| "rewards/log_2_reward": 9.9990234375, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 120 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 55.2, | |
| "completions/max_terminated_length": 55.2, | |
| "completions/mean_length": 40.234375, | |
| "completions/mean_terminated_length": 40.234375, | |
| "completions/min_length": 26.0, | |
| "completions/min_terminated_length": 26.0, | |
| "epoch": 0.7619047619047619, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 76599483.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 125 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 53.8, | |
| "completions/max_terminated_length": 53.8, | |
| "completions/mean_length": 40.09111328125, | |
| "completions/mean_terminated_length": 40.09111328125, | |
| "completions/min_length": 27.2, | |
| "completions/min_terminated_length": 27.2, | |
| "epoch": 0.7923809523809524, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 79599104.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 130 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 54.8, | |
| "completions/max_terminated_length": 54.8, | |
| "completions/mean_length": 39.84951171875, | |
| "completions/mean_terminated_length": 39.84951171875, | |
| "completions/min_length": 25.8, | |
| "completions/min_terminated_length": 25.8, | |
| "epoch": 0.8228571428571428, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 82590299.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 135 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 56.2, | |
| "completions/max_terminated_length": 56.2, | |
| "completions/mean_length": 39.871875, | |
| "completions/mean_terminated_length": 39.871875, | |
| "completions/min_length": 26.8, | |
| "completions/min_terminated_length": 26.8, | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 85586203.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 55.6, | |
| "completions/max_terminated_length": 55.6, | |
| "completions/mean_length": 40.103125, | |
| "completions/mean_terminated_length": 40.103125, | |
| "completions/min_length": 27.4, | |
| "completions/min_terminated_length": 27.4, | |
| "epoch": 0.8838095238095238, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 88590747.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 145 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 55.2, | |
| "completions/max_terminated_length": 55.2, | |
| "completions/mean_length": 39.84501953125, | |
| "completions/mean_terminated_length": 39.84501953125, | |
| "completions/min_length": 27.2, | |
| "completions/min_terminated_length": 27.2, | |
| "epoch": 0.9142857142857143, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "num_tokens": 91584520.0, | |
| "reward": 10.5, | |
| "reward_std": 0.0, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 1.0, | |
| "rewards/confidence_one_or_zero": 1.0, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 10.0, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9142857142857143, | |
| "eval_completions/clipped_ratio": 0.0, | |
| "eval_completions/max_length": 49.75, | |
| "eval_completions/max_terminated_length": 49.75, | |
| "eval_completions/mean_length": 39.660088539123535, | |
| "eval_completions/mean_terminated_length": 39.660088539123535, | |
| "eval_completions/min_length": 30.25, | |
| "eval_completions/min_terminated_length": 30.25, | |
| "eval_loss": 0.0, | |
| "eval_num_tokens": 91584520.0, | |
| "eval_reward": 10.5, | |
| "eval_reward_std": 0.0, | |
| "eval_rewards/accuracy_reward": 0.0, | |
| "eval_rewards/brier_reward": 1.0, | |
| "eval_rewards/confidence_one_or_zero": 1.0, | |
| "eval_rewards/format_reward": 1.0, | |
| "eval_rewards/log_2_reward": 10.0, | |
| "eval_rewards/mean_confidence_reward": 0.0, | |
| "eval_runtime": 7.0368, | |
| "eval_samples_per_second": 71.055, | |
| "eval_steps_per_second": 0.568, | |
| "step": 150 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 54.4, | |
| "completions/max_terminated_length": 54.4, | |
| "completions/mean_length": 39.9212890625, | |
| "completions/mean_terminated_length": 39.9212890625, | |
| "completions/min_length": 26.8, | |
| "completions/min_terminated_length": 26.8, | |
| "epoch": 0.9447619047619048, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": -0.0, | |
| "num_tokens": 94581058.0, | |
| "reward": 10.49995002746582, | |
| "reward_std": 0.0002821940928697586, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.9999843716621399, | |
| "rewards/confidence_one_or_zero": 0.99990234375, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 9.99995002746582, | |
| "rewards/mean_confidence_reward": 3.906250058207661e-05, | |
| "step": 155 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 55.4, | |
| "completions/max_terminated_length": 55.4, | |
| "completions/mean_length": 39.9310546875, | |
| "completions/mean_terminated_length": 39.9310546875, | |
| "completions/min_length": 26.6, | |
| "completions/min_terminated_length": 26.6, | |
| "epoch": 0.9752380952380952, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": -0.0, | |
| "num_tokens": 97578656.0, | |
| "reward": 10.499971961975097, | |
| "reward_std": 0.00015892352676019072, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.999993896484375, | |
| "rewards/confidence_one_or_zero": 0.99990234375, | |
| "rewards/format_reward": 1.0, | |
| "rewards/log_2_reward": 9.999971961975097, | |
| "rewards/mean_confidence_reward": 2.44140625e-05, | |
| "step": 160 | |
| }, | |
| { | |
| "clip_ratio/high_max": 0.0, | |
| "clip_ratio/high_mean": 0.0, | |
| "clip_ratio/low_mean": 0.0, | |
| "clip_ratio/low_min": 0.0, | |
| "clip_ratio/region_mean": 0.0, | |
| "completions/clipped_ratio": 0.0, | |
| "completions/max_length": 61.25, | |
| "completions/max_terminated_length": 61.25, | |
| "completions/mean_length": 39.9080171585083, | |
| "completions/mean_terminated_length": 39.9080171585083, | |
| "completions/min_length": 26.75, | |
| "completions/min_terminated_length": 26.75, | |
| "epoch": 0.9996190476190476, | |
| "num_tokens": 99976491.0, | |
| "reward": 10.49871826171875, | |
| "reward_std": 0.007250606548041105, | |
| "rewards/accuracy_reward": 0.0, | |
| "rewards/brier_reward": 0.9998779296875, | |
| "rewards/confidence_one_or_zero": 0.9998779296875, | |
| "rewards/format_reward": 0.9998779296875, | |
| "rewards/log_2_reward": 9.998779296875, | |
| "rewards/mean_confidence_reward": 0.0, | |
| "step": 164, | |
| "total_flos": 0.0, | |
| "train_loss": 0.03933636305515481, | |
| "train_runtime": 14176.1116, | |
| "train_samples_per_second": 0.741, | |
| "train_steps_per_second": 0.012 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 164, | |
| "num_input_tokens_seen": 99976491, | |
| "num_train_epochs": 1, | |
| "save_steps": 60, | |
| "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": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |