| { |
| "best_global_step": 4084, |
| "best_metric": 0.9286365509033203, |
| "best_model_checkpoint": "siglip2-finetune-full/checkpoint-4084", |
| "epoch": 4.0, |
| "eval_steps": 500, |
| "global_step": 4084, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.48971596474045054, |
| "grad_norm": 59.27558135986328, |
| "learning_rate": 2.665344571145265e-06, |
| "loss": 4.268, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.9794319294809011, |
| "grad_norm": 61.66608810424805, |
| "learning_rate": 2.2935052057511156e-06, |
| "loss": 2.4217, |
| "step": 1000 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.666574839302112, |
| "eval_loss": 1.7632426023483276, |
| "eval_model_preparation_time": 0.0024, |
| "eval_runtime": 476.9192, |
| "eval_samples_per_second": 68.502, |
| "eval_steps_per_second": 8.563, |
| "step": 1021 |
| }, |
| { |
| "epoch": 1.4691478942213516, |
| "grad_norm": 71.97106170654297, |
| "learning_rate": 1.921665840356966e-06, |
| "loss": 1.709, |
| "step": 1500 |
| }, |
| { |
| "epoch": 1.9588638589618022, |
| "grad_norm": 63.00515365600586, |
| "learning_rate": 1.5498264749628163e-06, |
| "loss": 1.4697, |
| "step": 2000 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.7615243342516069, |
| "eval_loss": 1.171113133430481, |
| "eval_model_preparation_time": 0.0024, |
| "eval_runtime": 475.7496, |
| "eval_samples_per_second": 68.671, |
| "eval_steps_per_second": 8.584, |
| "step": 2042 |
| }, |
| { |
| "epoch": 2.4485798237022527, |
| "grad_norm": 51.789241790771484, |
| "learning_rate": 1.1779871095686663e-06, |
| "loss": 1.2603, |
| "step": 2500 |
| }, |
| { |
| "epoch": 2.9382957884427032, |
| "grad_norm": 42.61498260498047, |
| "learning_rate": 8.061477441745166e-07, |
| "loss": 1.2054, |
| "step": 3000 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.7891031527395164, |
| "eval_loss": 0.9870542287826538, |
| "eval_model_preparation_time": 0.0024, |
| "eval_runtime": 475.3188, |
| "eval_samples_per_second": 68.733, |
| "eval_steps_per_second": 8.592, |
| "step": 3063 |
| }, |
| { |
| "epoch": 3.4280117531831538, |
| "grad_norm": 50.29034423828125, |
| "learning_rate": 4.343083787803669e-07, |
| "loss": 1.1072, |
| "step": 3500 |
| }, |
| { |
| "epoch": 3.9177277179236043, |
| "grad_norm": 67.87676239013672, |
| "learning_rate": 6.246901338621716e-08, |
| "loss": 1.0737, |
| "step": 4000 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.7995102540557086, |
| "eval_loss": 0.9286365509033203, |
| "eval_model_preparation_time": 0.0024, |
| "eval_runtime": 478.0209, |
| "eval_samples_per_second": 68.344, |
| "eval_steps_per_second": 8.544, |
| "step": 4084 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 4084, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 4, |
| "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": 1.0955968870687949e+19, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
