vit-base-beans-demo-v5 / trainer_state.json
SeungYon's picture
SeungYon
1c69359 verified
{
"best_metric": 0.027972575277090073,
"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": 1.7359464168548584,
"learning_rate": 0.00019230769230769233,
"loss": 0.9068,
"step": 10
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.7515268325805664,
"learning_rate": 0.00018461538461538463,
"loss": 0.3442,
"step": 20
},
{
"epoch": 0.46153846153846156,
"grad_norm": 3.3439853191375732,
"learning_rate": 0.00017692307692307693,
"loss": 0.1684,
"step": 30
},
{
"epoch": 0.6153846153846154,
"grad_norm": 2.0178005695343018,
"learning_rate": 0.00016923076923076923,
"loss": 0.3005,
"step": 40
},
{
"epoch": 0.7692307692307693,
"grad_norm": 1.4199612140655518,
"learning_rate": 0.00016153846153846155,
"loss": 0.1432,
"step": 50
},
{
"epoch": 0.9230769230769231,
"grad_norm": 3.5320632457733154,
"learning_rate": 0.00015384615384615385,
"loss": 0.3094,
"step": 60
},
{
"epoch": 1.0769230769230769,
"grad_norm": 3.943850040435791,
"learning_rate": 0.00014615384615384615,
"loss": 0.164,
"step": 70
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.42932555079460144,
"learning_rate": 0.00013846153846153847,
"loss": 0.0936,
"step": 80
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.2571287155151367,
"learning_rate": 0.00013076923076923077,
"loss": 0.0563,
"step": 90
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.2423134446144104,
"learning_rate": 0.0001230769230769231,
"loss": 0.0721,
"step": 100
},
{
"epoch": 1.5384615384615383,
"eval_accuracy": 0.9699248120300752,
"eval_loss": 0.08323061466217041,
"eval_runtime": 1.2671,
"eval_samples_per_second": 104.964,
"eval_steps_per_second": 13.416,
"step": 100
},
{
"epoch": 1.6923076923076923,
"grad_norm": 3.0321273803710938,
"learning_rate": 0.00011538461538461538,
"loss": 0.1154,
"step": 110
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.13413630425930023,
"learning_rate": 0.0001076923076923077,
"loss": 0.0794,
"step": 120
},
{
"epoch": 2.0,
"grad_norm": 0.14623944461345673,
"learning_rate": 0.0001,
"loss": 0.0543,
"step": 130
},
{
"epoch": 2.1538461538461537,
"grad_norm": 0.12642432749271393,
"learning_rate": 9.230769230769232e-05,
"loss": 0.0683,
"step": 140
},
{
"epoch": 2.3076923076923075,
"grad_norm": 0.09180299937725067,
"learning_rate": 8.461538461538461e-05,
"loss": 0.0452,
"step": 150
},
{
"epoch": 2.4615384615384617,
"grad_norm": 3.4162721633911133,
"learning_rate": 7.692307692307693e-05,
"loss": 0.0202,
"step": 160
},
{
"epoch": 2.6153846153846154,
"grad_norm": 0.06922739744186401,
"learning_rate": 6.923076923076924e-05,
"loss": 0.0191,
"step": 170
},
{
"epoch": 2.769230769230769,
"grad_norm": 0.26483821868896484,
"learning_rate": 6.153846153846155e-05,
"loss": 0.0165,
"step": 180
},
{
"epoch": 2.9230769230769234,
"grad_norm": 0.0765213593840599,
"learning_rate": 5.384615384615385e-05,
"loss": 0.0188,
"step": 190
},
{
"epoch": 3.076923076923077,
"grad_norm": 0.06417355686426163,
"learning_rate": 4.615384615384616e-05,
"loss": 0.0117,
"step": 200
},
{
"epoch": 3.076923076923077,
"eval_accuracy": 0.9924812030075187,
"eval_loss": 0.027972575277090073,
"eval_runtime": 1.8447,
"eval_samples_per_second": 72.097,
"eval_steps_per_second": 9.215,
"step": 200
},
{
"epoch": 3.230769230769231,
"grad_norm": 0.05795031413435936,
"learning_rate": 3.846153846153846e-05,
"loss": 0.011,
"step": 210
},
{
"epoch": 3.3846153846153846,
"grad_norm": 0.055242665112018585,
"learning_rate": 3.0769230769230774e-05,
"loss": 0.0108,
"step": 220
},
{
"epoch": 3.5384615384615383,
"grad_norm": 0.057339392602443695,
"learning_rate": 2.307692307692308e-05,
"loss": 0.0118,
"step": 230
},
{
"epoch": 3.6923076923076925,
"grad_norm": 0.05471523106098175,
"learning_rate": 1.5384615384615387e-05,
"loss": 0.0101,
"step": 240
},
{
"epoch": 3.8461538461538463,
"grad_norm": 0.05334159731864929,
"learning_rate": 7.692307692307694e-06,
"loss": 0.0101,
"step": 250
},
{
"epoch": 4.0,
"grad_norm": 0.06180350109934807,
"learning_rate": 0.0,
"loss": 0.0103,
"step": 260
},
{
"epoch": 4.0,
"step": 260,
"total_flos": 3.205097416476426e+17,
"train_loss": 0.11813208366242739,
"train_runtime": 92.1542,
"train_samples_per_second": 44.881,
"train_steps_per_second": 2.821
}
],
"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
}