| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 17.77777777777778, | |
| "eval_steps": 500, | |
| "global_step": 80, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.89, | |
| "eval_loss": 11.255721092224121, | |
| "eval_runtime": 3.4824, | |
| "eval_samples_per_second": 1.436, | |
| "eval_steps_per_second": 0.287, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 4.6637139320373535, | |
| "eval_runtime": 1.651, | |
| "eval_samples_per_second": 3.029, | |
| "eval_steps_per_second": 0.606, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 8.75e-05, | |
| "loss": 9.1378, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "eval_loss": 3.9639782905578613, | |
| "eval_runtime": 3.628, | |
| "eval_samples_per_second": 1.378, | |
| "eval_steps_per_second": 0.276, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 3.660576343536377, | |
| "eval_runtime": 1.6465, | |
| "eval_samples_per_second": 3.037, | |
| "eval_steps_per_second": 0.607, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 4.44, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 3.7563, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "eval_loss": 3.774608612060547, | |
| "eval_runtime": 3.6221, | |
| "eval_samples_per_second": 1.38, | |
| "eval_steps_per_second": 0.276, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 4.067192077636719, | |
| "eval_runtime": 1.6587, | |
| "eval_samples_per_second": 3.014, | |
| "eval_steps_per_second": 0.603, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "learning_rate": 6.25e-05, | |
| "loss": 2.8464, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 6.89, | |
| "eval_loss": 4.304504871368408, | |
| "eval_runtime": 3.619, | |
| "eval_samples_per_second": 1.382, | |
| "eval_steps_per_second": 0.276, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 3.801119327545166, | |
| "eval_runtime": 1.8745, | |
| "eval_samples_per_second": 2.667, | |
| "eval_steps_per_second": 0.533, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "learning_rate": 5e-05, | |
| "loss": 2.462, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 8.89, | |
| "eval_loss": 4.001826763153076, | |
| "eval_runtime": 3.9564, | |
| "eval_samples_per_second": 1.264, | |
| "eval_steps_per_second": 0.253, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 4.09613037109375, | |
| "eval_runtime": 1.772, | |
| "eval_samples_per_second": 2.822, | |
| "eval_steps_per_second": 0.564, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 10.89, | |
| "eval_loss": 4.627002239227295, | |
| "eval_runtime": 4.0617, | |
| "eval_samples_per_second": 1.231, | |
| "eval_steps_per_second": 0.246, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 11.11, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 2.1133, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 4.324659824371338, | |
| "eval_runtime": 1.8005, | |
| "eval_samples_per_second": 2.777, | |
| "eval_steps_per_second": 0.555, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 12.89, | |
| "eval_loss": 4.851120471954346, | |
| "eval_runtime": 3.967, | |
| "eval_samples_per_second": 1.26, | |
| "eval_steps_per_second": 0.252, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 13.33, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.7555, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 4.627060413360596, | |
| "eval_runtime": 1.8189, | |
| "eval_samples_per_second": 2.749, | |
| "eval_steps_per_second": 0.55, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 14.89, | |
| "eval_loss": 4.949428081512451, | |
| "eval_runtime": 3.9565, | |
| "eval_samples_per_second": 1.264, | |
| "eval_steps_per_second": 0.253, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 15.56, | |
| "learning_rate": 1.25e-05, | |
| "loss": 1.2005, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 5.440162658691406, | |
| "eval_runtime": 1.803, | |
| "eval_samples_per_second": 2.773, | |
| "eval_steps_per_second": 0.555, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 16.89, | |
| "eval_loss": 5.508076190948486, | |
| "eval_runtime": 4.004, | |
| "eval_samples_per_second": 1.249, | |
| "eval_steps_per_second": 0.25, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 17.78, | |
| "learning_rate": 0.0, | |
| "loss": 0.8324, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 17.78, | |
| "eval_loss": 5.481626510620117, | |
| "eval_runtime": 0.8414, | |
| "eval_samples_per_second": 5.942, | |
| "eval_steps_per_second": 1.188, | |
| "step": 80 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 80, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 2.267304128717783e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |