fil_b64_le3_s4000 / checkpoint-1000 /trainer_state.json
mikhail-panzo's picture
Training in progress, step 1000, checkpoint
bf55226 verified
{
"best_metric": 0.43401163816452026,
"best_model_checkpoint": "mikhail-panzo/fil_b64_le3_s4000/checkpoint-500",
"epoch": 44.44444444444444,
"eval_steps": 500,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 2.2222222222222223,
"grad_norm": 2.2609777450561523,
"learning_rate": 2.5e-05,
"loss": 0.7231,
"step": 50
},
{
"epoch": 4.444444444444445,
"grad_norm": 1.939948320388794,
"learning_rate": 5e-05,
"loss": 0.575,
"step": 100
},
{
"epoch": 6.666666666666667,
"grad_norm": 2.9677317142486572,
"learning_rate": 7.5e-05,
"loss": 0.5207,
"step": 150
},
{
"epoch": 8.88888888888889,
"grad_norm": 4.588547706604004,
"learning_rate": 0.0001,
"loss": 0.4996,
"step": 200
},
{
"epoch": 11.11111111111111,
"grad_norm": 1.8626277446746826,
"learning_rate": 0.000125,
"loss": 0.4881,
"step": 250
},
{
"epoch": 13.333333333333334,
"grad_norm": 3.4871737957000732,
"learning_rate": 0.00015,
"loss": 0.4729,
"step": 300
},
{
"epoch": 15.555555555555555,
"grad_norm": 3.5084946155548096,
"learning_rate": 0.000175,
"loss": 0.4661,
"step": 350
},
{
"epoch": 17.77777777777778,
"grad_norm": 7.88467264175415,
"learning_rate": 0.0002,
"loss": 0.4925,
"step": 400
},
{
"epoch": 20.0,
"grad_norm": 4.796986103057861,
"learning_rate": 0.00022500000000000002,
"loss": 0.4899,
"step": 450
},
{
"epoch": 22.22222222222222,
"grad_norm": 6.419567584991455,
"learning_rate": 0.00025,
"loss": 0.4644,
"step": 500
},
{
"epoch": 22.22222222222222,
"eval_loss": 0.43401163816452026,
"eval_runtime": 17.3691,
"eval_samples_per_second": 9.154,
"eval_steps_per_second": 1.151,
"step": 500
},
{
"epoch": 24.444444444444443,
"grad_norm": 2.6020901203155518,
"learning_rate": 0.000275,
"loss": 0.4492,
"step": 550
},
{
"epoch": 26.666666666666668,
"grad_norm": 7.353488445281982,
"learning_rate": 0.0003,
"loss": 0.4556,
"step": 600
},
{
"epoch": 28.88888888888889,
"grad_norm": 4.736776351928711,
"learning_rate": 0.00032500000000000004,
"loss": 0.4669,
"step": 650
},
{
"epoch": 31.11111111111111,
"grad_norm": 7.411691665649414,
"learning_rate": 0.00035,
"loss": 0.4892,
"step": 700
},
{
"epoch": 33.333333333333336,
"grad_norm": 5.571273326873779,
"learning_rate": 0.000375,
"loss": 0.4858,
"step": 750
},
{
"epoch": 35.55555555555556,
"grad_norm": 7.498672008514404,
"learning_rate": 0.0004,
"loss": 0.4929,
"step": 800
},
{
"epoch": 37.77777777777778,
"grad_norm": 6.763402462005615,
"learning_rate": 0.000425,
"loss": 0.4874,
"step": 850
},
{
"epoch": 40.0,
"grad_norm": 3.335564136505127,
"learning_rate": 0.00045000000000000004,
"loss": 0.513,
"step": 900
},
{
"epoch": 42.22222222222222,
"grad_norm": 3.1004624366760254,
"learning_rate": 0.000475,
"loss": 0.4882,
"step": 950
},
{
"epoch": 44.44444444444444,
"grad_norm": 10.330741882324219,
"learning_rate": 0.0005,
"loss": 0.6468,
"step": 1000
},
{
"epoch": 44.44444444444444,
"eval_loss": 0.753709614276886,
"eval_runtime": 16.7842,
"eval_samples_per_second": 9.473,
"eval_steps_per_second": 1.192,
"step": 1000
}
],
"logging_steps": 50,
"max_steps": 4000,
"num_input_tokens_seen": 0,
"num_train_epochs": 182,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.3761741337261632e+16,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}