| { | |
| "best_metric": 0.040931012481451035, | |
| "best_model_checkpoint": "./vit-base-beans-demo-v18Set/checkpoint-200", | |
| "epoch": 4.0, | |
| "eval_steps": 100, | |
| "global_step": 260, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 1.6004313230514526, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.8087, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 2.299769639968872, | |
| "learning_rate": 0.00018461538461538463, | |
| "loss": 0.3525, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 1.0842756032943726, | |
| "learning_rate": 0.00017692307692307693, | |
| "loss": 0.2053, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 6.000638961791992, | |
| "learning_rate": 0.00016923076923076923, | |
| "loss": 0.1943, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.6018532514572144, | |
| "learning_rate": 0.00016153846153846155, | |
| "loss": 0.1727, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 1.624453067779541, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.2282, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0769230769230769, | |
| "grad_norm": 0.2731366753578186, | |
| "learning_rate": 0.00014615384615384615, | |
| "loss": 0.1134, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2307692307692308, | |
| "grad_norm": 0.6613112688064575, | |
| "learning_rate": 0.00013846153846153847, | |
| "loss": 0.0936, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3846153846153846, | |
| "grad_norm": 0.14583992958068848, | |
| "learning_rate": 0.00013076923076923077, | |
| "loss": 0.0824, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 0.4384416937828064, | |
| "learning_rate": 0.0001230769230769231, | |
| "loss": 0.0851, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "eval_accuracy": 0.9548872180451128, | |
| "eval_loss": 0.13281743228435516, | |
| "eval_runtime": 2.2563, | |
| "eval_samples_per_second": 58.947, | |
| "eval_steps_per_second": 7.535, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.6923076923076923, | |
| "grad_norm": 1.8822920322418213, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 0.142, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8461538461538463, | |
| "grad_norm": 0.11390437930822372, | |
| "learning_rate": 0.0001076923076923077, | |
| "loss": 0.0943, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.1020607054233551, | |
| "learning_rate": 0.0001, | |
| "loss": 0.0639, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1538461538461537, | |
| "grad_norm": 0.09703055769205093, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.1224, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 2.91983962059021, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.0243, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 0.09748347103595734, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0213, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6153846153846154, | |
| "grad_norm": 0.0757695659995079, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.0179, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.769230769230769, | |
| "grad_norm": 0.06924037635326385, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.0293, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9230769230769234, | |
| "grad_norm": 0.0656639039516449, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.0136, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 0.06468123197555542, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0131, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "eval_accuracy": 0.9849624060150376, | |
| "eval_loss": 0.040931012481451035, | |
| "eval_runtime": 1.3479, | |
| "eval_samples_per_second": 98.671, | |
| "eval_steps_per_second": 12.612, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.230769230769231, | |
| "grad_norm": 0.060022223740816116, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.012, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 3.3846153846153846, | |
| "grad_norm": 0.06361185014247894, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.0162, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 3.5384615384615383, | |
| "grad_norm": 0.06728459149599075, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 0.0115, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 3.6923076923076925, | |
| "grad_norm": 0.06117574870586395, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.0115, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.05541615188121796, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0113, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.05663846433162689, | |
| "learning_rate": 0.0, | |
| "loss": 0.0131, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "step": 260, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_loss": 0.11360841490901434, | |
| "train_runtime": 111.3779, | |
| "train_samples_per_second": 37.135, | |
| "train_steps_per_second": 2.334 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 260, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.205097416476426e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |