| { | |
| "best_metric": 3.0561561584472656, | |
| "best_model_checkpoint": "car_brands_image_detection/checkpoint-9795", | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 9795, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 5.048170566558838, | |
| "learning_rate": 9.538224730631092e-07, | |
| "loss": 3.061, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 6.777750492095947, | |
| "learning_rate": 9.025141097998974e-07, | |
| "loss": 3.0282, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 4.8603620529174805, | |
| "learning_rate": 8.512057465366855e-07, | |
| "loss": 3.0189, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 5.830420017242432, | |
| "learning_rate": 7.998973832734736e-07, | |
| "loss": 3.011, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 5.956003189086914, | |
| "learning_rate": 7.485890200102616e-07, | |
| "loss": 3.0075, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 5.180022239685059, | |
| "learning_rate": 6.972806567470497e-07, | |
| "loss": 3.0021, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5952698918709344, | |
| "eval_loss": 3.1075940132141113, | |
| "eval_runtime": 692.5128, | |
| "eval_samples_per_second": 100.56, | |
| "eval_steps_per_second": 12.57, | |
| "step": 3265 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 6.344422340393066, | |
| "learning_rate": 6.459722934838378e-07, | |
| "loss": 2.9912, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 6.641003608703613, | |
| "learning_rate": 5.946639302206258e-07, | |
| "loss": 2.9868, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 4.404183387756348, | |
| "learning_rate": 5.433555669574141e-07, | |
| "loss": 2.9654, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 6.073057651519775, | |
| "learning_rate": 4.920472036942021e-07, | |
| "loss": 2.9538, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 6.557525634765625, | |
| "learning_rate": 4.4073884043099026e-07, | |
| "loss": 2.9638, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 4.880246639251709, | |
| "learning_rate": 3.894304771677783e-07, | |
| "loss": 2.9471, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 8.338777542114258, | |
| "learning_rate": 3.381221139045664e-07, | |
| "loss": 2.9398, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5991901089906518, | |
| "eval_loss": 3.0709476470947266, | |
| "eval_runtime": 688.3993, | |
| "eval_samples_per_second": 101.161, | |
| "eval_steps_per_second": 12.645, | |
| "step": 6530 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 6.223824977874756, | |
| "learning_rate": 2.8681375064135456e-07, | |
| "loss": 2.9497, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 6.486847877502441, | |
| "learning_rate": 2.3550538737814263e-07, | |
| "loss": 2.9271, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 6.441537857055664, | |
| "learning_rate": 1.8419702411493075e-07, | |
| "loss": 2.9339, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 5.405475616455078, | |
| "learning_rate": 1.3288866085171882e-07, | |
| "loss": 2.9417, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 7.3839545249938965, | |
| "learning_rate": 8.158029758850691e-08, | |
| "loss": 2.9116, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "grad_norm": 7.343071937561035, | |
| "learning_rate": 3.027193432529502e-08, | |
| "loss": 2.9239, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6021482215425265, | |
| "eval_loss": 3.0561561584472656, | |
| "eval_runtime": 700.4906, | |
| "eval_samples_per_second": 99.415, | |
| "eval_steps_per_second": 12.427, | |
| "step": 9795 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 9795, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 2.4353840270149964e+19, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |