| { | |
| "best_global_step": 2756, | |
| "best_metric": 0.6427269577980042, | |
| "best_model_checkpoint": "siglip2-finetune-full/checkpoint-2756", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 2756, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.36284470246734396, | |
| "grad_norm": 2.6488075256347656, | |
| "learning_rate": 0.00016674057649667407, | |
| "loss": 1.0035, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7256894049346879, | |
| "grad_norm": 1.355397343635559, | |
| "learning_rate": 0.00012978566149297858, | |
| "loss": 0.898, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6578780177890724, | |
| "eval_loss": 0.8105661869049072, | |
| "eval_model_preparation_time": 0.0038, | |
| "eval_runtime": 542.5801, | |
| "eval_samples_per_second": 81.227, | |
| "eval_steps_per_second": 10.153, | |
| "step": 1378 | |
| }, | |
| { | |
| "epoch": 1.0885341074020318, | |
| "grad_norm": 3.95896577835083, | |
| "learning_rate": 9.283074648928308e-05, | |
| "loss": 0.8135, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.4513788098693758, | |
| "grad_norm": 1.9753128290176392, | |
| "learning_rate": 5.587583148558758e-05, | |
| "loss": 0.7563, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.8142235123367199, | |
| "grad_norm": 3.7470362186431885, | |
| "learning_rate": 1.892091648189209e-05, | |
| "loss": 0.6919, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.740651660918497, | |
| "eval_loss": 0.6427269577980042, | |
| "eval_model_preparation_time": 0.0038, | |
| "eval_runtime": 537.3561, | |
| "eval_samples_per_second": 82.016, | |
| "eval_steps_per_second": 10.252, | |
| "step": 2756 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2756, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "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": 7.382728081268048e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |