76 lines
1.9 KiB
JSON
76 lines
1.9 KiB
JSON
|
{
|
||
|
|
"best_global_step": 1656,
|
||
|
|
"best_metric": 0.33448827266693115,
|
||
|
|
"best_model_checkpoint": "siglip2-finetune-full/checkpoint-1656",
|
||
|
|
"epoch": 2.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 1656,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.6038647342995169,
|
||
|
|
"grad_norm": 5.860389709472656,
|
||
|
|
"learning_rate": 0.00016302382908792113,
|
||
|
|
"loss": 0.823,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"eval_accuracy": 0.7698523581165276,
|
||
|
|
"eval_loss": 0.577109158039093,
|
||
|
|
"eval_model_preparation_time": 0.0023,
|
||
|
|
"eval_runtime": 605.3087,
|
||
|
|
"eval_samples_per_second": 43.751,
|
||
|
|
"eval_steps_per_second": 5.47,
|
||
|
|
"step": 828
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2077294685990339,
|
||
|
|
"grad_norm": 5.253721237182617,
|
||
|
|
"learning_rate": 0.00012193919474116682,
|
||
|
|
"loss": 0.6183,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8115942028985508,
|
||
|
|
"grad_norm": 4.610513210296631,
|
||
|
|
"learning_rate": 8.08545603944125e-05,
|
||
|
|
"loss": 0.5136,
|
||
|
|
"step": 1500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0,
|
||
|
|
"eval_accuracy": 0.8734282369822151,
|
||
|
|
"eval_loss": 0.33448827266693115,
|
||
|
|
"eval_model_preparation_time": 0.0023,
|
||
|
|
"eval_runtime": 598.3057,
|
||
|
|
"eval_samples_per_second": 44.263,
|
||
|
|
"eval_steps_per_second": 5.534,
|
||
|
|
"step": 1656
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 500,
|
||
|
|
"max_steps": 2484,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 3,
|
||
|
|
"save_steps": 500,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": false
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 5.794356182313075e+18,
|
||
|
|
"train_batch_size": 32,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|