| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.03461205653302567, | |
| "eval_steps": 5, | |
| "global_step": 30, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0011537352177675222, | |
| "eval_loss": 1.1585837602615356, | |
| "eval_runtime": 16.5538, | |
| "eval_samples_per_second": 22.049, | |
| "eval_steps_per_second": 11.055, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003461205653302567, | |
| "grad_norm": 1.0870860815048218, | |
| "learning_rate": 6e-05, | |
| "loss": 4.7529, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.005768676088837612, | |
| "eval_loss": 1.1217771768569946, | |
| "eval_runtime": 16.7257, | |
| "eval_samples_per_second": 21.823, | |
| "eval_steps_per_second": 10.941, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.006922411306605134, | |
| "grad_norm": 1.0277127027511597, | |
| "learning_rate": 0.00012, | |
| "loss": 4.7248, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.010383616959907702, | |
| "grad_norm": 1.0448460578918457, | |
| "learning_rate": 0.00018, | |
| "loss": 4.2535, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.011537352177675224, | |
| "eval_loss": 0.9416375160217285, | |
| "eval_runtime": 16.8164, | |
| "eval_samples_per_second": 21.705, | |
| "eval_steps_per_second": 10.882, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.013844822613210269, | |
| "grad_norm": 1.648591160774231, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 3.8484, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.017306028266512834, | |
| "grad_norm": 2.317676544189453, | |
| "learning_rate": 0.00017071067811865476, | |
| "loss": 3.1999, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.017306028266512834, | |
| "eval_loss": 0.7792818546295166, | |
| "eval_runtime": 16.8701, | |
| "eval_samples_per_second": 21.636, | |
| "eval_steps_per_second": 10.848, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.020767233919815404, | |
| "grad_norm": 1.5525565147399902, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 3.2798, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.02307470435535045, | |
| "eval_loss": 0.7133286595344543, | |
| "eval_runtime": 16.8737, | |
| "eval_samples_per_second": 21.631, | |
| "eval_steps_per_second": 10.845, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02422843957311797, | |
| "grad_norm": 1.2597628831863403, | |
| "learning_rate": 8.435655349597689e-05, | |
| "loss": 2.8077, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.027689645226420537, | |
| "grad_norm": 1.202157974243164, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 2.8526, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.02884338044418806, | |
| "eval_loss": 0.6947705745697021, | |
| "eval_runtime": 16.8902, | |
| "eval_samples_per_second": 21.61, | |
| "eval_steps_per_second": 10.835, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.031150850879723104, | |
| "grad_norm": 1.0633991956710815, | |
| "learning_rate": 1.0899347581163221e-05, | |
| "loss": 2.864, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.03461205653302567, | |
| "grad_norm": 1.1619864702224731, | |
| "learning_rate": 0.0, | |
| "loss": 2.9061, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03461205653302567, | |
| "eval_loss": 0.6916398406028748, | |
| "eval_runtime": 16.8943, | |
| "eval_samples_per_second": 21.605, | |
| "eval_steps_per_second": 10.832, | |
| "step": 30 | |
| } | |
| ], | |
| "logging_steps": 3, | |
| "max_steps": 30, | |
| "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": 2625889099579392.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |