| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9927007299270073, | |
| "eval_steps": 100, | |
| "global_step": 17, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "completion_length": 31.543402910232544, | |
| "epoch": 0.058394160583941604, | |
| "grad_norm": 0.6463247537612915, | |
| "kl": 0.0, | |
| "learning_rate": 1e-06, | |
| "loss": 0.0, | |
| "reward": 0.4809027835726738, | |
| "reward_std": 0.4057047627866268, | |
| "rewards/semantic_entropy": 0.4809027835726738, | |
| "step": 1 | |
| }, | |
| { | |
| "completion_length": 27.753472566604614, | |
| "epoch": 0.11678832116788321, | |
| "grad_norm": 0.31535470485687256, | |
| "kl": 0.0, | |
| "learning_rate": 2e-06, | |
| "loss": 0.0, | |
| "reward": 0.5850694514811039, | |
| "reward_std": 0.3451368249952793, | |
| "rewards/semantic_entropy": 0.5850694514811039, | |
| "step": 2 | |
| }, | |
| { | |
| "completion_length": 32.34895849227905, | |
| "epoch": 0.17518248175182483, | |
| "grad_norm": 0.38025209307670593, | |
| "kl": 0.0008878707885742188, | |
| "learning_rate": 1.9781476007338054e-06, | |
| "loss": 0.0, | |
| "reward": 0.4791666679084301, | |
| "reward_std": 0.38540373742580414, | |
| "rewards/semantic_entropy": 0.4791666679084301, | |
| "step": 3 | |
| }, | |
| { | |
| "completion_length": 27.75607681274414, | |
| "epoch": 0.23357664233576642, | |
| "grad_norm": 0.7206792235374451, | |
| "kl": 0.0020771026611328125, | |
| "learning_rate": 1.9135454576426007e-06, | |
| "loss": 0.0, | |
| "reward": 0.5998263880610466, | |
| "reward_std": 0.348458593711257, | |
| "rewards/semantic_entropy": 0.5998263880610466, | |
| "step": 4 | |
| }, | |
| { | |
| "completion_length": 29.860243558883667, | |
| "epoch": 0.291970802919708, | |
| "grad_norm": 0.5087810754776001, | |
| "kl": 0.0018243789672851562, | |
| "learning_rate": 1.8090169943749474e-06, | |
| "loss": 0.0, | |
| "reward": 0.5034722238779068, | |
| "reward_std": 0.39532990381121635, | |
| "rewards/semantic_entropy": 0.5034722238779068, | |
| "step": 5 | |
| }, | |
| { | |
| "completion_length": 28.811631679534912, | |
| "epoch": 0.35036496350364965, | |
| "grad_norm": 0.4511905610561371, | |
| "kl": 0.0026292800903320312, | |
| "learning_rate": 1.669130606358858e-06, | |
| "loss": 0.0, | |
| "reward": 0.5295138992369175, | |
| "reward_std": 0.3658079691231251, | |
| "rewards/semantic_entropy": 0.5295138992369175, | |
| "step": 6 | |
| }, | |
| { | |
| "completion_length": 31.047743320465088, | |
| "epoch": 0.40875912408759124, | |
| "grad_norm": 0.46206873655319214, | |
| "kl": 0.003326416015625, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.0, | |
| "reward": 0.4973958395421505, | |
| "reward_std": 0.39973679929971695, | |
| "rewards/semantic_entropy": 0.4973958395421505, | |
| "step": 7 | |
| }, | |
| { | |
| "completion_length": 30.878472328186035, | |
| "epoch": 0.46715328467153283, | |
| "grad_norm": 0.4796462655067444, | |
| "kl": 0.004016876220703125, | |
| "learning_rate": 1.3090169943749473e-06, | |
| "loss": 0.0, | |
| "reward": 0.5217013955116272, | |
| "reward_std": 0.36933426558971405, | |
| "rewards/semantic_entropy": 0.5217013955116272, | |
| "step": 8 | |
| }, | |
| { | |
| "completion_length": 34.5590283870697, | |
| "epoch": 0.5255474452554745, | |
| "grad_norm": 0.43365350365638733, | |
| "kl": 0.0037174224853515625, | |
| "learning_rate": 1.1045284632676535e-06, | |
| "loss": 0.0, | |
| "reward": 0.4444444589316845, | |
| "reward_std": 0.4142540544271469, | |
| "rewards/semantic_entropy": 0.4444444589316845, | |
| "step": 9 | |
| }, | |
| { | |
| "completion_length": 31.54600763320923, | |
| "epoch": 0.583941605839416, | |
| "grad_norm": 0.3461940586566925, | |
| "kl": 0.0054531097412109375, | |
| "learning_rate": 8.954715367323466e-07, | |
| "loss": 0.0, | |
| "reward": 0.5312499962747097, | |
| "reward_std": 0.33972141705453396, | |
| "rewards/semantic_entropy": 0.5312499962747097, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 25.934895992279053, | |
| "epoch": 0.6423357664233577, | |
| "grad_norm": 0.7275694012641907, | |
| "kl": 0.0059986114501953125, | |
| "learning_rate": 6.909830056250526e-07, | |
| "loss": 0.0, | |
| "reward": 0.5425347350537777, | |
| "reward_std": 0.39345845952630043, | |
| "rewards/semantic_entropy": 0.5425347350537777, | |
| "step": 11 | |
| }, | |
| { | |
| "completion_length": 27.58420157432556, | |
| "epoch": 0.7007299270072993, | |
| "grad_norm": 0.578926146030426, | |
| "kl": 0.010272979736328125, | |
| "learning_rate": 5.000000000000002e-07, | |
| "loss": 0.0, | |
| "reward": 0.5564236119389534, | |
| "reward_std": 0.38245424441993237, | |
| "rewards/semantic_entropy": 0.5564236119389534, | |
| "step": 12 | |
| }, | |
| { | |
| "completion_length": 32.76302146911621, | |
| "epoch": 0.7591240875912408, | |
| "grad_norm": 0.3081968128681183, | |
| "kl": 0.0044574737548828125, | |
| "learning_rate": 3.308693936411421e-07, | |
| "loss": 0.0, | |
| "reward": 0.4583333395421505, | |
| "reward_std": 0.39592672139406204, | |
| "rewards/semantic_entropy": 0.4583333395421505, | |
| "step": 13 | |
| }, | |
| { | |
| "completion_length": 28.50086808204651, | |
| "epoch": 0.8175182481751825, | |
| "grad_norm": 0.4480704367160797, | |
| "kl": 0.006603240966796875, | |
| "learning_rate": 1.9098300562505264e-07, | |
| "loss": 0.0, | |
| "reward": 0.4782986231148243, | |
| "reward_std": 0.3709658682346344, | |
| "rewards/semantic_entropy": 0.4782986231148243, | |
| "step": 14 | |
| }, | |
| { | |
| "completion_length": 28.36545181274414, | |
| "epoch": 0.8759124087591241, | |
| "grad_norm": 0.2726985514163971, | |
| "kl": 0.00714111328125, | |
| "learning_rate": 8.645454235739902e-08, | |
| "loss": 0.0, | |
| "reward": 0.572048619389534, | |
| "reward_std": 0.3576664440333843, | |
| "rewards/semantic_entropy": 0.572048619389534, | |
| "step": 15 | |
| }, | |
| { | |
| "completion_length": 29.555555820465088, | |
| "epoch": 0.9343065693430657, | |
| "grad_norm": 0.5542200207710266, | |
| "kl": 0.011875152587890625, | |
| "learning_rate": 2.185239926619431e-08, | |
| "loss": 0.0, | |
| "reward": 0.5442708358168602, | |
| "reward_std": 0.4052053317427635, | |
| "rewards/semantic_entropy": 0.5442708358168602, | |
| "step": 16 | |
| }, | |
| { | |
| "completion_length": 28.62326431274414, | |
| "epoch": 0.9927007299270073, | |
| "grad_norm": 0.4658753275871277, | |
| "kl": 0.010679244995117188, | |
| "learning_rate": 0.0, | |
| "loss": 0.0, | |
| "reward": 0.564236119389534, | |
| "reward_std": 0.3762592002749443, | |
| "rewards/semantic_entropy": 0.564236119389534, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.9927007299270073, | |
| "step": 17, | |
| "total_flos": 0.0, | |
| "train_loss": 2.8922741848572622e-08, | |
| "train_runtime": 6158.5867, | |
| "train_samples_per_second": 0.133, | |
| "train_steps_per_second": 0.003 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 17, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 10, | |
| "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 | |
| } | |