| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.6923076923076925, | |
| "eval_steps": 5, | |
| "global_step": 50, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.6938247793422839, | |
| "learning_rate": 4e-05, | |
| "loss": 1.4136, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.8631641326116998, | |
| "learning_rate": 0.0002, | |
| "loss": 1.4131, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "eval_loss": 1.1345069408416748, | |
| "eval_runtime": 2.6731, | |
| "eval_samples_per_second": 7.482, | |
| "eval_steps_per_second": 1.122, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.43474097166898595, | |
| "learning_rate": 0.00019396926207859084, | |
| "loss": 0.923, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_loss": 0.8694825172424316, | |
| "eval_runtime": 1.3826, | |
| "eval_samples_per_second": 14.466, | |
| "eval_steps_per_second": 2.17, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 0.2409966406944847, | |
| "learning_rate": 0.0001766044443118978, | |
| "loss": 0.743, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "eval_loss": 0.7064261436462402, | |
| "eval_runtime": 1.3848, | |
| "eval_samples_per_second": 14.443, | |
| "eval_steps_per_second": 2.166, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.3817272155364255, | |
| "learning_rate": 0.00015000000000000001, | |
| "loss": 0.6198, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_loss": 0.6521738767623901, | |
| "eval_runtime": 1.3817, | |
| "eval_samples_per_second": 14.475, | |
| "eval_steps_per_second": 2.171, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.2535200861748062, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 0.5551, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "eval_loss": 0.6249467134475708, | |
| "eval_runtime": 1.3886, | |
| "eval_samples_per_second": 14.403, | |
| "eval_steps_per_second": 2.16, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.13673734506176724, | |
| "learning_rate": 8.263518223330697e-05, | |
| "loss": 0.532, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "eval_loss": 0.6073147654533386, | |
| "eval_runtime": 1.3824, | |
| "eval_samples_per_second": 14.468, | |
| "eval_steps_per_second": 2.17, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "grad_norm": 0.14094467505755442, | |
| "learning_rate": 5.000000000000002e-05, | |
| "loss": 0.4771, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "eval_loss": 0.6010892391204834, | |
| "eval_runtime": 1.3823, | |
| "eval_samples_per_second": 14.468, | |
| "eval_steps_per_second": 2.17, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 0.1354251186197571, | |
| "learning_rate": 2.339555568810221e-05, | |
| "loss": 0.4848, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "eval_loss": 0.5987650156021118, | |
| "eval_runtime": 1.382, | |
| "eval_samples_per_second": 14.472, | |
| "eval_steps_per_second": 2.171, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "grad_norm": 0.13940840189652645, | |
| "learning_rate": 6.030737921409169e-06, | |
| "loss": 0.4541, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "eval_loss": 0.5984574556350708, | |
| "eval_runtime": 1.3847, | |
| "eval_samples_per_second": 14.444, | |
| "eval_steps_per_second": 2.167, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 0.14114542040303035, | |
| "learning_rate": 0.0, | |
| "loss": 0.4496, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "eval_loss": 0.5981998443603516, | |
| "eval_runtime": 1.3793, | |
| "eval_samples_per_second": 14.5, | |
| "eval_steps_per_second": 2.175, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "step": 50, | |
| "total_flos": 1335726637056.0, | |
| "train_loss": 0.6651756596565247, | |
| "train_runtime": 177.5954, | |
| "train_samples_per_second": 4.505, | |
| "train_steps_per_second": 0.282 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 50, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1335726637056.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |