trainer_state.json

{
 "best_metric": 0.9196428571428571,
 "best_model_checkpoint": "swinv2-tiny-patch4-window8-256-Lego-v2-3ep/checkpoint-147",
 "epoch": 3.0,
 "eval_steps": 500,
 "global_step": 147,
 "is_hyper_param_search": false,
 "is_local_process_zero": true,
 "is_world_process_zero": true,
 "log_history": [
 {
 "epoch": 0.20408163265306123,
 "grad_norm": 4.521535396575928,
 "learning_rate": 3.3333333333333335e-05,
 "loss": 2.7556,
 "step": 10
 },
 {
 "epoch": 0.40816326530612246,
 "grad_norm": 7.520199775695801,
 "learning_rate": 4.810606060606061e-05,
 "loss": 2.4445,
 "step": 20
 },
 {
 "epoch": 0.6122448979591837,
 "grad_norm": 13.214489936828613,
 "learning_rate": 4.431818181818182e-05,
 "loss": 1.8022,
 "step": 30
 },
 {
 "epoch": 0.8163265306122449,
 "grad_norm": 14.265619277954102,
 "learning_rate": 4.053030303030303e-05,
 "loss": 1.3415,
 "step": 40
 },
 {
 "epoch": 1.0,
 "eval_accuracy": 0.7857142857142857,
 "eval_loss": 0.6598730683326721,
 "eval_runtime": 0.7369,
 "eval_samples_per_second": 151.997,
 "eval_steps_per_second": 5.428,
 "step": 49
 },
 {
 "epoch": 1.0204081632653061,
 "grad_norm": 15.687291145324707,
 "learning_rate": 3.6742424242424246e-05,
 "loss": 1.138,
 "step": 50
 },
 {
 "epoch": 1.2244897959183674,
 "grad_norm": 11.222309112548828,
 "learning_rate": 3.295454545454545e-05,
 "loss": 0.8949,
 "step": 60
 },
 {
 "epoch": 1.4285714285714286,
 "grad_norm": 13.59701919555664,
 "learning_rate": 2.916666666666667e-05,
 "loss": 0.8637,
 "step": 70
 },
 {
 "epoch": 1.6326530612244898,
 "grad_norm": 8.102276802062988,
 "learning_rate": 2.537878787878788e-05,
 "loss": 0.7632,
 "step": 80
 },
 {
 "epoch": 1.836734693877551,
 "grad_norm": 10.037010192871094,
 "learning_rate": 2.1590909090909093e-05,
 "loss": 0.7005,
 "step": 90
 },
 {
 "epoch": 2.0,
 "eval_accuracy": 0.8928571428571429,
 "eval_loss": 0.3036225736141205,
 "eval_runtime": 0.7855,
 "eval_samples_per_second": 142.578,
 "eval_steps_per_second": 5.092,
 "step": 98
 },
 {
 "epoch": 2.0408163265306123,
 "grad_norm": 9.7354154586792,
 "learning_rate": 1.7803030303030303e-05,
 "loss": 0.6343,
 "step": 100
 },
 {
 "epoch": 2.2448979591836733,
 "grad_norm": 8.97431468963623,
 "learning_rate": 1.4015151515151515e-05,
 "loss": 0.5505,
 "step": 110
 },
 {
 "epoch": 2.4489795918367347,
 "grad_norm": 8.927824974060059,
 "learning_rate": 1.0227272727272729e-05,
 "loss": 0.5527,
 "step": 120
 },
 {
 "epoch": 2.6530612244897958,
 "grad_norm": 7.921907424926758,
 "learning_rate": 6.43939393939394e-06,
 "loss": 0.4982,
 "step": 130
 },
 {
 "epoch": 2.857142857142857,
 "grad_norm": 9.745354652404785,
 "learning_rate": 2.651515151515152e-06,
 "loss": 0.5615,
 "step": 140
 },
 {
 "epoch": 3.0,
 "eval_accuracy": 0.9196428571428571,
 "eval_loss": 0.2128717005252838,
 "eval_runtime": 0.7718,
 "eval_samples_per_second": 145.113,
 "eval_steps_per_second": 5.183,
 "step": 147
 },
 {
 "epoch": 3.0,
 "step": 147,
 "total_flos": 4.6849871171251814e+17,
 "train_loss": 1.081290222349621,
 "train_runtime": 252.638,
 "train_samples_per_second": 74.419,
 "train_steps_per_second": 0.582
 }
 ],
 "logging_steps": 10,
 "max_steps": 147,
 "num_input_tokens_seen": 0,
 "num_train_epochs": 3,
 "save_steps": 500,
 "total_flos": 4.6849871171251814e+17,
 "train_batch_size": 32,
 "trial_name": null,
 "trial_params": null
}