| { |
| "best_global_step": 2253, |
| "best_metric": 0.7667109370231628, |
| "best_model_checkpoint": "siglip2-finetune-full/checkpoint-2253", |
| "epoch": 3.0, |
| "eval_steps": 500, |
| "global_step": 2253, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.6657789613848203, |
| "grad_norm": 84.61146545410156, |
| "learning_rate": 2.3871992737176576e-06, |
| "loss": 3.5666, |
| "step": 500 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.7465462716378163, |
| "eval_loss": 1.435652256011963, |
| "eval_model_preparation_time": 0.0023, |
| "eval_runtime": 375.3834, |
| "eval_samples_per_second": 64.02, |
| "eval_steps_per_second": 8.002, |
| "step": 751 |
| }, |
| { |
| "epoch": 1.3315579227696404, |
| "grad_norm": 65.66386413574219, |
| "learning_rate": 1.7063095778483886e-06, |
| "loss": 1.6147, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.9973368841544608, |
| "grad_norm": 44.739898681640625, |
| "learning_rate": 1.0254198819791194e-06, |
| "loss": 1.1571, |
| "step": 1500 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.8380908788282291, |
| "eval_loss": 0.8885560631752014, |
| "eval_model_preparation_time": 0.0023, |
| "eval_runtime": 372.6212, |
| "eval_samples_per_second": 64.494, |
| "eval_steps_per_second": 8.062, |
| "step": 1502 |
| }, |
| { |
| "epoch": 2.6631158455392807, |
| "grad_norm": 80.52566528320312, |
| "learning_rate": 3.4453018610985024e-07, |
| "loss": 0.9224, |
| "step": 2000 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.8532789613848203, |
| "eval_loss": 0.7667109370231628, |
| "eval_model_preparation_time": 0.0023, |
| "eval_runtime": 366.5531, |
| "eval_samples_per_second": 65.562, |
| "eval_steps_per_second": 8.195, |
| "step": 2253 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 2253, |
| "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": true |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 6.044645313614316e+18, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
