| { | |
| "best_metric": 0.9606256742179072, | |
| "best_model_checkpoint": "./vit-xray-tumor/checkpoint-125", | |
| "epoch": 20.0, | |
| "eval_steps": 125, | |
| "global_step": 680, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.7352941176470589, | |
| "grad_norm": 0.32342442870140076, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.7033, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.4705882352941178, | |
| "grad_norm": 0.5360887050628662, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6638, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.2058823529411766, | |
| "grad_norm": 0.32269367575645447, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.5639, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "grad_norm": 0.3643856644630432, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5276, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.6764705882352944, | |
| "grad_norm": 0.3383520543575287, | |
| "learning_rate": 9.954227982894034e-06, | |
| "loss": 0.5283, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.6764705882352944, | |
| "eval_accuracy": 0.9606256742179072, | |
| "eval_loss": 0.29475122690200806, | |
| "eval_runtime": 3.6867, | |
| "eval_samples_per_second": 502.885, | |
| "eval_steps_per_second": 2.17, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.411764705882353, | |
| "grad_norm": 0.3172575831413269, | |
| "learning_rate": 9.817749962596115e-06, | |
| "loss": 0.5181, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 5.147058823529412, | |
| "grad_norm": 0.2688044309616089, | |
| "learning_rate": 9.59306468881811e-06, | |
| "loss": 0.5133, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 0.47131991386413574, | |
| "learning_rate": 9.284285880837947e-06, | |
| "loss": 0.5142, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 6.617647058823529, | |
| "grad_norm": 0.437072217464447, | |
| "learning_rate": 8.897066910207958e-06, | |
| "loss": 0.5043, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 7.352941176470588, | |
| "grad_norm": 0.9283384084701538, | |
| "learning_rate": 8.438497294267117e-06, | |
| "loss": 0.516, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 7.352941176470588, | |
| "eval_accuracy": 0.9600862998921251, | |
| "eval_loss": 0.28430742025375366, | |
| "eval_runtime": 3.3555, | |
| "eval_samples_per_second": 552.534, | |
| "eval_steps_per_second": 2.384, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 8.088235294117647, | |
| "grad_norm": 0.4921472370624542, | |
| "learning_rate": 7.916972895537471e-06, | |
| "loss": 0.5053, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 8.823529411764707, | |
| "grad_norm": 0.6513373851776123, | |
| "learning_rate": 7.342042203498952e-06, | |
| "loss": 0.5023, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 9.558823529411764, | |
| "grad_norm": 0.5955798625946045, | |
| "learning_rate": 6.724231513139853e-06, | |
| "loss": 0.5046, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 10.294117647058824, | |
| "grad_norm": 0.5370674729347229, | |
| "learning_rate": 6.074852201055121e-06, | |
| "loss": 0.4918, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 11.029411764705882, | |
| "grad_norm": 0.6328652501106262, | |
| "learning_rate": 5.405793627637157e-06, | |
| "loss": 0.4878, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 11.029411764705882, | |
| "eval_accuracy": 0.9600862998921251, | |
| "eval_loss": 0.275637149810791, | |
| "eval_runtime": 3.4137, | |
| "eval_samples_per_second": 543.107, | |
| "eval_steps_per_second": 2.344, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 1.029469609260559, | |
| "learning_rate": 4.729305457072913e-06, | |
| "loss": 0.4806, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.7929940223693848, | |
| "learning_rate": 4.057773380608411e-06, | |
| "loss": 0.4819, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 13.235294117647058, | |
| "grad_norm": 0.8358489871025085, | |
| "learning_rate": 3.403492349320101e-06, | |
| "loss": 0.4693, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 13.970588235294118, | |
| "grad_norm": 1.030651569366455, | |
| "learning_rate": 2.778441468230483e-06, | |
| "loss": 0.4632, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 14.705882352941176, | |
| "grad_norm": 1.082851529121399, | |
| "learning_rate": 2.1940646731880887e-06, | |
| "loss": 0.459, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 14.705882352941176, | |
| "eval_accuracy": 0.9600862998921251, | |
| "eval_loss": 0.2800777554512024, | |
| "eval_runtime": 3.4811, | |
| "eval_samples_per_second": 532.584, | |
| "eval_steps_per_second": 2.298, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 15.441176470588236, | |
| "grad_norm": 0.9598743319511414, | |
| "learning_rate": 1.6610612060565235e-06, | |
| "loss": 0.4518, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 16.176470588235293, | |
| "grad_norm": 0.8586243987083435, | |
| "learning_rate": 1.1891897243618184e-06, | |
| "loss": 0.4604, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 16.91176470588235, | |
| "grad_norm": 1.1040153503417969, | |
| "learning_rate": 7.870896319167548e-07, | |
| "loss": 0.4542, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 17.647058823529413, | |
| "grad_norm": 0.8632078766822815, | |
| "learning_rate": 4.6212290164521554e-07, | |
| "loss": 0.4473, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 18.38235294117647, | |
| "grad_norm": 0.9933310151100159, | |
| "learning_rate": 2.2023928664194229e-07, | |
| "loss": 0.4462, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 18.38235294117647, | |
| "eval_accuracy": 0.959546925566343, | |
| "eval_loss": 0.2760601043701172, | |
| "eval_runtime": 3.319, | |
| "eval_samples_per_second": 558.602, | |
| "eval_steps_per_second": 2.41, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 19.11764705882353, | |
| "grad_norm": 0.9852302074432373, | |
| "learning_rate": 6.58673872923693e-08, | |
| "loss": 0.4523, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 19.852941176470587, | |
| "grad_norm": 0.8414192795753479, | |
| "learning_rate": 1.8335688835802169e-09, | |
| "loss": 0.4441, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 680, | |
| "total_flos": 1.3409213882909e+19, | |
| "train_loss": 0.5017087179071763, | |
| "train_runtime": 256.3708, | |
| "train_samples_per_second": 674.96, | |
| "train_steps_per_second": 2.652 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 680, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 125, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3409213882909e+19, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |