opt-350m / last-checkpoint /trainer_state.json
SystemAdmin123's picture
Training in progress, step 100, checkpoint
89543b3 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 14.285714285714286,
"eval_steps": 200,
"global_step": 100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.14285714285714285,
"eval_loss": 3.078927516937256,
"eval_runtime": 4.841,
"eval_samples_per_second": 310.059,
"eval_steps_per_second": 3.305,
"step": 1
},
{
"epoch": 1.4285714285714286,
"grad_norm": 9.25,
"learning_rate": 0.00019863613034027224,
"loss": 6.1788,
"step": 10
},
{
"epoch": 2.857142857142857,
"grad_norm": 5.25,
"learning_rate": 0.0001879473751206489,
"loss": 5.2684,
"step": 20
},
{
"epoch": 4.285714285714286,
"grad_norm": 5.28125,
"learning_rate": 0.00016772815716257412,
"loss": 4.6887,
"step": 30
},
{
"epoch": 5.714285714285714,
"grad_norm": 2.3125,
"learning_rate": 0.00014016954246529696,
"loss": 4.3802,
"step": 40
},
{
"epoch": 7.142857142857143,
"grad_norm": 3.484375,
"learning_rate": 0.00010825793454723325,
"loss": 4.1083,
"step": 50
},
{
"epoch": 8.571428571428571,
"grad_norm": 2.5625,
"learning_rate": 7.54514512859201e-05,
"loss": 3.8961,
"step": 60
},
{
"epoch": 10.0,
"grad_norm": 2.96875,
"learning_rate": 4.530518418775733e-05,
"loss": 3.8097,
"step": 70
},
{
"epoch": 11.428571428571429,
"grad_norm": 1.546875,
"learning_rate": 2.1085949060360654e-05,
"loss": 3.7435,
"step": 80
},
{
"epoch": 12.857142857142858,
"grad_norm": 2.328125,
"learning_rate": 5.418275829936537e-06,
"loss": 3.721,
"step": 90
},
{
"epoch": 14.285714285714286,
"grad_norm": 1.453125,
"learning_rate": 0.0,
"loss": 3.7058,
"step": 100
}
],
"logging_steps": 10,
"max_steps": 100,
"num_input_tokens_seen": 0,
"num_train_epochs": 15,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.57855601360896e+16,
"train_batch_size": 24,
"trial_name": null,
"trial_params": null
}