trainer_state.json

{
 "best_metric": 0.04597444459795952,
 "best_model_checkpoint": "./vit-base-beans-demo-v5/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": 0.051635466516017914,
 "learning_rate": 0.00019230769230769233,
 "loss": 0.0112,
 "step": 10
 },
 {
 "epoch": 0.3076923076923077,
 "grad_norm": 0.04402761533856392,
 "learning_rate": 0.00018461538461538463,
 "loss": 0.0096,
 "step": 20
 },
 {
 "epoch": 0.46153846153846156,
 "grad_norm": 0.04412761703133583,
 "learning_rate": 0.00017692307692307693,
 "loss": 0.0084,
 "step": 30
 },
 {
 "epoch": 0.6153846153846154,
 "grad_norm": 0.057089973241090775,
 "learning_rate": 0.00016923076923076923,
 "loss": 0.0728,
 "step": 40
 },
 {
 "epoch": 0.7692307692307693,
 "grad_norm": 0.07369471341371536,
 "learning_rate": 0.00016153846153846155,
 "loss": 0.0737,
 "step": 50
 },
 {
 "epoch": 0.9230769230769231,
 "grad_norm": 3.7364368438720703,
 "learning_rate": 0.00015384615384615385,
 "loss": 0.0938,
 "step": 60
 },
 {
 "epoch": 1.0769230769230769,
 "grad_norm": 4.054567337036133,
 "learning_rate": 0.00014615384615384615,
 "loss": 0.0521,
 "step": 70
 },
 {
 "epoch": 1.2307692307692308,
 "grad_norm": 0.03869302198290825,
 "learning_rate": 0.00013846153846153847,
 "loss": 0.1791,
 "step": 80
 },
 {
 "epoch": 1.3846153846153846,
 "grad_norm": 2.1249592304229736,
 "learning_rate": 0.00013076923076923077,
 "loss": 0.1318,
 "step": 90
 },
 {
 "epoch": 1.5384615384615383,
 "grad_norm": 0.046364787966012955,
 "learning_rate": 0.0001230769230769231,
 "loss": 0.0604,
 "step": 100
 },
 {
 "epoch": 1.5384615384615383,
 "eval_accuracy": 0.9548872180451128,
 "eval_loss": 0.20907698571681976,
 "eval_runtime": 1.2497,
 "eval_samples_per_second": 106.427,
 "eval_steps_per_second": 13.603,
 "step": 100
 },
 {
 "epoch": 1.6923076923076923,
 "grad_norm": 0.2539267838001251,
 "learning_rate": 0.00011538461538461538,
 "loss": 0.0395,
 "step": 110
 },
 {
 "epoch": 1.8461538461538463,
 "grad_norm": 0.07288012653589249,
 "learning_rate": 0.0001076923076923077,
 "loss": 0.0084,
 "step": 120
 },
 {
 "epoch": 2.0,
 "grad_norm": 0.10698893666267395,
 "learning_rate": 0.0001,
 "loss": 0.0072,
 "step": 130
 },
 {
 "epoch": 2.1538461538461537,
 "grad_norm": 0.02628379687666893,
 "learning_rate": 9.230769230769232e-05,
 "loss": 0.0079,
 "step": 140
 },
 {
 "epoch": 2.3076923076923075,
 "grad_norm": 0.03258649259805679,
 "learning_rate": 8.461538461538461e-05,
 "loss": 0.0126,
 "step": 150
 },
 {
 "epoch": 2.4615384615384617,
 "grad_norm": 0.024210650473833084,
 "learning_rate": 7.692307692307693e-05,
 "loss": 0.0047,
 "step": 160
 },
 {
 "epoch": 2.6153846153846154,
 "grad_norm": 0.03985947370529175,
 "learning_rate": 6.923076923076924e-05,
 "loss": 0.0046,
 "step": 170
 },
 {
 "epoch": 2.769230769230769,
 "grad_norm": 0.6222206354141235,
 "learning_rate": 6.153846153846155e-05,
 "loss": 0.0136,
 "step": 180
 },
 {
 "epoch": 2.9230769230769234,
 "grad_norm": 0.0636032223701477,
 "learning_rate": 5.384615384615385e-05,
 "loss": 0.0286,
 "step": 190
 },
 {
 "epoch": 3.076923076923077,
 "grad_norm": 0.022785609588027,
 "learning_rate": 4.615384615384616e-05,
 "loss": 0.0042,
 "step": 200
 },
 {
 "epoch": 3.076923076923077,
 "eval_accuracy": 0.9849624060150376,
 "eval_loss": 0.04597444459795952,
 "eval_runtime": 1.2861,
 "eval_samples_per_second": 103.415,
 "eval_steps_per_second": 13.218,
 "step": 200
 },
 {
 "epoch": 3.230769230769231,
 "grad_norm": 0.021587267518043518,
 "learning_rate": 3.846153846153846e-05,
 "loss": 0.004,
 "step": 210
 },
 {
 "epoch": 3.3846153846153846,
 "grad_norm": 0.02168501354753971,
 "learning_rate": 3.0769230769230774e-05,
 "loss": 0.0066,
 "step": 220
 },
 {
 "epoch": 3.5384615384615383,
 "grad_norm": 0.021682456135749817,
 "learning_rate": 2.307692307692308e-05,
 "loss": 0.0039,
 "step": 230
 },
 {
 "epoch": 3.6923076923076925,
 "grad_norm": 0.024300260469317436,
 "learning_rate": 1.5384615384615387e-05,
 "loss": 0.0039,
 "step": 240
 },
 {
 "epoch": 3.8461538461538463,
 "grad_norm": 0.03438858315348625,
 "learning_rate": 7.692307692307694e-06,
 "loss": 0.0043,
 "step": 250
 },
 {
 "epoch": 4.0,
 "grad_norm": 0.020293371751904488,
 "learning_rate": 0.0,
 "loss": 0.004,
 "step": 260
 },
 {
 "epoch": 4.0,
 "step": 260,
 "total_flos": 3.205097416476426e+17,
 "train_loss": 0.03272535361063022,
 "train_runtime": 82.4869,
 "train_samples_per_second": 50.141,
 "train_steps_per_second": 3.152
 }
 ],
 "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": false
 },
 "attributes": {}
 }
 },
 "total_flos": 3.205097416476426e+17,
 "train_batch_size": 16,
 "trial_name": null,
 "trial_params": null
}