|
{ |
|
"best_metric": 0.0752067118883133, |
|
"best_model_checkpoint": "./beans_outputs/checkpoint-130", |
|
"epoch": 10.0, |
|
"eval_steps": 500, |
|
"global_step": 1300, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.7692307692307693, |
|
"grad_norm": 4.380956172943115, |
|
"learning_rate": 0.005, |
|
"loss": 0.4846, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.9624060150375939, |
|
"eval_loss": 0.0752067118883133, |
|
"eval_runtime": 0.7531, |
|
"eval_samples_per_second": 176.6, |
|
"eval_steps_per_second": 22.573, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 1.5384615384615383, |
|
"grad_norm": 0.7832342982292175, |
|
"learning_rate": 0.005, |
|
"loss": 0.3473, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.9548872180451128, |
|
"eval_loss": 0.15986023843288422, |
|
"eval_runtime": 0.7821, |
|
"eval_samples_per_second": 170.05, |
|
"eval_steps_per_second": 21.736, |
|
"step": 260 |
|
}, |
|
{ |
|
"epoch": 2.3076923076923075, |
|
"grad_norm": 1.9846404790878296, |
|
"learning_rate": 0.005, |
|
"loss": 0.2953, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.9548872180451128, |
|
"eval_loss": 0.1191997155547142, |
|
"eval_runtime": 0.8019, |
|
"eval_samples_per_second": 165.864, |
|
"eval_steps_per_second": 21.201, |
|
"step": 390 |
|
}, |
|
{ |
|
"epoch": 3.076923076923077, |
|
"grad_norm": 1.4397724866867065, |
|
"learning_rate": 0.005, |
|
"loss": 0.2891, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 3.8461538461538463, |
|
"grad_norm": 3.0908043384552, |
|
"learning_rate": 0.005, |
|
"loss": 0.2653, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.9398496240601504, |
|
"eval_loss": 0.13930319249629974, |
|
"eval_runtime": 0.7871, |
|
"eval_samples_per_second": 168.974, |
|
"eval_steps_per_second": 21.598, |
|
"step": 520 |
|
}, |
|
{ |
|
"epoch": 4.615384615384615, |
|
"grad_norm": 2.720701217651367, |
|
"learning_rate": 0.005, |
|
"loss": 0.2344, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.9624060150375939, |
|
"eval_loss": 0.10011590272188187, |
|
"eval_runtime": 0.8095, |
|
"eval_samples_per_second": 164.306, |
|
"eval_steps_per_second": 21.001, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 5.384615384615385, |
|
"grad_norm": 0.17460593581199646, |
|
"learning_rate": 0.005, |
|
"loss": 0.21, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.9624060150375939, |
|
"eval_loss": 0.0893179401755333, |
|
"eval_runtime": 0.7824, |
|
"eval_samples_per_second": 169.993, |
|
"eval_steps_per_second": 21.728, |
|
"step": 780 |
|
}, |
|
{ |
|
"epoch": 6.153846153846154, |
|
"grad_norm": 0.6050938963890076, |
|
"learning_rate": 0.005, |
|
"loss": 0.2397, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 6.923076923076923, |
|
"grad_norm": 1.7308460474014282, |
|
"learning_rate": 0.005, |
|
"loss": 0.3117, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.924812030075188, |
|
"eval_loss": 0.1933087259531021, |
|
"eval_runtime": 0.7988, |
|
"eval_samples_per_second": 166.494, |
|
"eval_steps_per_second": 21.281, |
|
"step": 910 |
|
}, |
|
{ |
|
"epoch": 7.6923076923076925, |
|
"grad_norm": 1.327934741973877, |
|
"learning_rate": 0.005, |
|
"loss": 0.2459, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.924812030075188, |
|
"eval_loss": 0.19011461734771729, |
|
"eval_runtime": 0.7881, |
|
"eval_samples_per_second": 168.75, |
|
"eval_steps_per_second": 21.57, |
|
"step": 1040 |
|
}, |
|
{ |
|
"epoch": 8.461538461538462, |
|
"grad_norm": 0.8504745364189148, |
|
"learning_rate": 0.005, |
|
"loss": 0.25, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.9699248120300752, |
|
"eval_loss": 0.08675242960453033, |
|
"eval_runtime": 0.7998, |
|
"eval_samples_per_second": 166.287, |
|
"eval_steps_per_second": 21.255, |
|
"step": 1170 |
|
}, |
|
{ |
|
"epoch": 9.23076923076923, |
|
"grad_norm": 3.838754177093506, |
|
"learning_rate": 0.005, |
|
"loss": 0.1834, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"grad_norm": 0.03880231827497482, |
|
"learning_rate": 0.005, |
|
"loss": 0.2038, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.9473684210526315, |
|
"eval_loss": 0.15278470516204834, |
|
"eval_runtime": 0.8065, |
|
"eval_samples_per_second": 164.915, |
|
"eval_steps_per_second": 21.079, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"step": 1300, |
|
"total_flos": 8.508699063477043e+17, |
|
"train_loss": 0.2738876885634202, |
|
"train_runtime": 131.5423, |
|
"train_samples_per_second": 78.606, |
|
"train_steps_per_second": 9.883 |
|
} |
|
], |
|
"logging_steps": 100, |
|
"max_steps": 1300, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 10, |
|
"save_steps": 500, |
|
"stateful_callbacks": { |
|
"TrainerControl": { |
|
"args": { |
|
"should_epoch_stop": false, |
|
"should_evaluate": false, |
|
"should_log": false, |
|
"should_save": true, |
|
"should_training_stop": true |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 8.508699063477043e+17, |
|
"train_batch_size": 8, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|