| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 8.16326530612245, | |
| "eval_steps": 20, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 2.9999999999999997e-05, | |
| "loss": 1.7509, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 5.9999999999999995e-05, | |
| "loss": 1.6967, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "eval_loss": 1.6115258932113647, | |
| "eval_runtime": 30.6688, | |
| "eval_samples_per_second": 12.619, | |
| "eval_steps_per_second": 1.598, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 8.999999999999999e-05, | |
| "loss": 1.5138, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "learning_rate": 0.00011999999999999999, | |
| "loss": 1.0417, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.27, | |
| "eval_loss": 0.5990618467330933, | |
| "eval_runtime": 45.282, | |
| "eval_samples_per_second": 8.546, | |
| "eval_steps_per_second": 1.082, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "learning_rate": 0.00015, | |
| "loss": 0.4159, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "learning_rate": 0.00017999999999999998, | |
| "loss": 0.3112, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "eval_loss": 0.277026891708374, | |
| "eval_runtime": 30.7302, | |
| "eval_samples_per_second": 12.593, | |
| "eval_steps_per_second": 1.595, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.71, | |
| "learning_rate": 0.00020999999999999998, | |
| "loss": 0.285, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 6.53, | |
| "learning_rate": 0.00023999999999999998, | |
| "loss": 0.27, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 6.53, | |
| "eval_loss": 0.24872899055480957, | |
| "eval_runtime": 30.6498, | |
| "eval_samples_per_second": 12.627, | |
| "eval_steps_per_second": 1.599, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 7.35, | |
| "learning_rate": 0.00027, | |
| "loss": 0.2464, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 8.16, | |
| "learning_rate": 0.0003, | |
| "loss": 0.2276, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 8.16, | |
| "eval_loss": 0.22304466366767883, | |
| "eval_runtime": 30.6362, | |
| "eval_samples_per_second": 12.632, | |
| "eval_steps_per_second": 1.599, | |
| "step": 100 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 34, | |
| "save_steps": 20, | |
| "total_flos": 2.5467399903903744e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |