| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.4946695095948828, | |
| "eval_steps": 500, | |
| "global_step": 87, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "completion_length": 383.43863677978516, | |
| "epoch": 0.017057569296375266, | |
| "grad_norm": 0.7056748867034912, | |
| "kl": 0.0, | |
| "learning_rate": 3.333333333333333e-07, | |
| "loss": 0.1106, | |
| "reward": 0.712053619325161, | |
| "reward_std": 0.39283478632569313, | |
| "rewards/accuracy_reward": 0.7087053954601288, | |
| "rewards/format_reward": 0.003348214435391128, | |
| "step": 1 | |
| }, | |
| { | |
| "completion_length": 417.2608995437622, | |
| "epoch": 0.08528784648187633, | |
| "grad_norm": 10.000102996826172, | |
| "kl": 0.0006268918514251709, | |
| "learning_rate": 1.6666666666666669e-06, | |
| "loss": 0.0585, | |
| "reward": 0.7357701230794191, | |
| "reward_std": 0.37447383999824524, | |
| "rewards/accuracy_reward": 0.7313058394938707, | |
| "rewards/format_reward": 0.004464285942958668, | |
| "step": 5 | |
| }, | |
| { | |
| "completion_length": 497.51051940917966, | |
| "epoch": 0.17057569296375266, | |
| "grad_norm": 0.3063008487224579, | |
| "kl": 0.0017390489578247071, | |
| "learning_rate": 2.9987834972573546e-06, | |
| "loss": 0.0784, | |
| "reward": 0.7703125357627869, | |
| "reward_std": 0.3321108963340521, | |
| "rewards/accuracy_reward": 0.7689732491970063, | |
| "rewards/format_reward": 0.0013392857974395156, | |
| "step": 10 | |
| }, | |
| { | |
| "completion_length": 500.5500213623047, | |
| "epoch": 0.255863539445629, | |
| "grad_norm": 2.702301263809204, | |
| "kl": 0.010482597351074218, | |
| "learning_rate": 2.956412726139078e-06, | |
| "loss": 0.1039, | |
| "reward": 0.8607143312692642, | |
| "reward_std": 0.2081299727782607, | |
| "rewards/accuracy_reward": 0.8604911163449287, | |
| "rewards/format_reward": 0.00022321429569274187, | |
| "step": 15 | |
| }, | |
| { | |
| "completion_length": 579.4114112854004, | |
| "epoch": 0.3411513859275053, | |
| "grad_norm": 0.1238444522023201, | |
| "kl": 0.002919578552246094, | |
| "learning_rate": 2.8551756519155732e-06, | |
| "loss": 0.0479, | |
| "reward": 0.8312500357627869, | |
| "reward_std": 0.18435155414044857, | |
| "rewards/accuracy_reward": 0.8312500357627869, | |
| "rewards/format_reward": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "completion_length": 549.7174324035644, | |
| "epoch": 0.42643923240938164, | |
| "grad_norm": 0.8831402659416199, | |
| "kl": 0.004978370666503906, | |
| "learning_rate": 2.699164145105252e-06, | |
| "loss": 0.049, | |
| "reward": 0.8495536148548126, | |
| "reward_std": 0.16553675523027778, | |
| "rewards/accuracy_reward": 0.8495536148548126, | |
| "rewards/format_reward": 0.0, | |
| "step": 25 | |
| }, | |
| { | |
| "completion_length": 592.7422164916992, | |
| "epoch": 0.511727078891258, | |
| "grad_norm": 0.1488523930311203, | |
| "kl": 0.0038507461547851564, | |
| "learning_rate": 2.4946839873611927e-06, | |
| "loss": 0.0382, | |
| "reward": 0.8165178954601288, | |
| "reward_std": 0.17473283503204584, | |
| "rewards/accuracy_reward": 0.8165178954601288, | |
| "rewards/format_reward": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "completion_length": 622.1555999755859, | |
| "epoch": 0.5970149253731343, | |
| "grad_norm": 0.15904287993907928, | |
| "kl": 0.003652000427246094, | |
| "learning_rate": 2.25e-06, | |
| "loss": 0.0396, | |
| "reward": 0.7752232491970062, | |
| "reward_std": 0.2052750363945961, | |
| "rewards/accuracy_reward": 0.7752232491970062, | |
| "rewards/format_reward": 0.0, | |
| "step": 35 | |
| }, | |
| { | |
| "completion_length": 623.6841766357422, | |
| "epoch": 0.6823027718550106, | |
| "grad_norm": 0.11450017243623734, | |
| "kl": 0.004055023193359375, | |
| "learning_rate": 1.975001990702209e-06, | |
| "loss": 0.0393, | |
| "reward": 0.7602678909897804, | |
| "reward_std": 0.2107524886727333, | |
| "rewards/accuracy_reward": 0.7602678909897804, | |
| "rewards/format_reward": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "completion_length": 709.4147674560547, | |
| "epoch": 0.767590618336887, | |
| "grad_norm": 0.08205878734588623, | |
| "kl": 0.011851119995117187, | |
| "learning_rate": 1.6808050203829845e-06, | |
| "loss": 0.0359, | |
| "reward": 0.5767857410013676, | |
| "reward_std": 0.23020651061087846, | |
| "rewards/accuracy_reward": 0.5767857410013676, | |
| "rewards/format_reward": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "completion_length": 716.3203491210937, | |
| "epoch": 0.8528784648187633, | |
| "grad_norm": 0.11587467789649963, | |
| "kl": 0.00437164306640625, | |
| "learning_rate": 1.3793001469249112e-06, | |
| "loss": 0.0316, | |
| "reward": 0.574553594738245, | |
| "reward_std": 0.24477371014654636, | |
| "rewards/accuracy_reward": 0.574553594738245, | |
| "rewards/format_reward": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "completion_length": 724.092886352539, | |
| "epoch": 0.9381663113006397, | |
| "grad_norm": 0.09563813358545303, | |
| "kl": 0.003982162475585938, | |
| "learning_rate": 1.0826738041253211e-06, | |
| "loss": 0.0343, | |
| "reward": 0.586830385029316, | |
| "reward_std": 0.25132143292576076, | |
| "rewards/accuracy_reward": 0.586830385029316, | |
| "rewards/format_reward": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "completion_length": 572.2068037553267, | |
| "epoch": 1.0341151385927505, | |
| "grad_norm": 2.0511133670806885, | |
| "kl": 0.006678494540127841, | |
| "learning_rate": 8.029152419343472e-07, | |
| "loss": 0.0443, | |
| "reward": 0.7159091213887389, | |
| "reward_std": 0.1854671297002245, | |
| "rewards/accuracy_reward": 0.7159091213887389, | |
| "rewards/format_reward": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "completion_length": 398.4535873413086, | |
| "epoch": 1.1194029850746268, | |
| "grad_norm": 0.10762660950422287, | |
| "kl": 0.007398223876953125, | |
| "learning_rate": 5.513319366069343e-07, | |
| "loss": 0.0402, | |
| "reward": 0.9424107536673546, | |
| "reward_std": 0.08303849077783525, | |
| "rewards/accuracy_reward": 0.9424107536673546, | |
| "rewards/format_reward": 0.0, | |
| "step": 65 | |
| }, | |
| { | |
| "completion_length": 435.78461837768555, | |
| "epoch": 1.2046908315565032, | |
| "grad_norm": 0.2290273904800415, | |
| "kl": 0.022534942626953124, | |
| "learning_rate": 3.380925572585183e-07, | |
| "loss": 0.0382, | |
| "reward": 0.9220982596278191, | |
| "reward_std": 0.10309098572470247, | |
| "rewards/accuracy_reward": 0.9220982596278191, | |
| "rewards/format_reward": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "completion_length": 480.5415397644043, | |
| "epoch": 1.2899786780383795, | |
| "grad_norm": 0.15823833644390106, | |
| "kl": 0.005892181396484375, | |
| "learning_rate": 1.718159615201853e-07, | |
| "loss": 0.043, | |
| "reward": 0.8839286133646965, | |
| "reward_std": 0.1372868578415364, | |
| "rewards/accuracy_reward": 0.883705398440361, | |
| "rewards/format_reward": 0.00022321429569274187, | |
| "step": 75 | |
| }, | |
| { | |
| "completion_length": 517.8933265686035, | |
| "epoch": 1.375266524520256, | |
| "grad_norm": 0.12397471815347672, | |
| "kl": 0.005947113037109375, | |
| "learning_rate": 5.922283255294164e-08, | |
| "loss": 0.0319, | |
| "reward": 0.8825893267989159, | |
| "reward_std": 0.14030799297615887, | |
| "rewards/accuracy_reward": 0.8825893267989159, | |
| "rewards/format_reward": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "completion_length": 516.631275177002, | |
| "epoch": 1.4605543710021323, | |
| "grad_norm": 0.13666246831417084, | |
| "kl": 0.006908416748046875, | |
| "learning_rate": 4.864037798685106e-09, | |
| "loss": 0.0268, | |
| "reward": 0.8665178909897804, | |
| "reward_std": 0.14118066830560566, | |
| "rewards/accuracy_reward": 0.8662946775555611, | |
| "rewards/format_reward": 0.00022321429569274187, | |
| "step": 85 | |
| }, | |
| { | |
| "completion_length": 589.9185562133789, | |
| "epoch": 1.4946695095948828, | |
| "kl": 0.005696296691894531, | |
| "reward": 0.7940848618745804, | |
| "reward_std": 0.18597882147878408, | |
| "rewards/accuracy_reward": 0.7940848618745804, | |
| "rewards/format_reward": 0.0, | |
| "step": 87, | |
| "total_flos": 0.0, | |
| "train_loss": 0.04595949243882607, | |
| "train_runtime": 10042.601, | |
| "train_samples_per_second": 1.12, | |
| "train_steps_per_second": 0.009 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 87, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |