ds_no_offload_liger_packing_zero2 / trainer_state.json
sedrickkeh's picture
End of training
425bd53 verified
raw
history blame
3.92 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 18,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.16666666666666666,
"grad_norm": 15.757265090942383,
"learning_rate": 5e-06,
"loss": 1.2885,
"step": 1
},
{
"epoch": 0.3333333333333333,
"grad_norm": 16.411413192749023,
"learning_rate": 1e-05,
"loss": 1.3273,
"step": 2
},
{
"epoch": 0.5,
"grad_norm": 9.334980010986328,
"learning_rate": 9.903926402016153e-06,
"loss": 1.2085,
"step": 3
},
{
"epoch": 0.6666666666666666,
"grad_norm": 7.6427531242370605,
"learning_rate": 9.619397662556434e-06,
"loss": 1.1297,
"step": 4
},
{
"epoch": 0.8333333333333334,
"grad_norm": 8.37617301940918,
"learning_rate": 9.157348061512728e-06,
"loss": 1.1569,
"step": 5
},
{
"epoch": 1.0,
"grad_norm": 6.6993207931518555,
"learning_rate": 8.535533905932739e-06,
"loss": 1.0759,
"step": 6
},
{
"epoch": 1.1666666666666667,
"grad_norm": 5.372480869293213,
"learning_rate": 7.777851165098012e-06,
"loss": 1.037,
"step": 7
},
{
"epoch": 1.3333333333333333,
"grad_norm": 3.0929112434387207,
"learning_rate": 6.913417161825449e-06,
"loss": 1.0009,
"step": 8
},
{
"epoch": 1.5,
"grad_norm": 1.9615176916122437,
"learning_rate": 5.975451610080643e-06,
"loss": 0.9742,
"step": 9
},
{
"epoch": 1.6666666666666665,
"grad_norm": 1.9075912237167358,
"learning_rate": 5e-06,
"loss": 0.9315,
"step": 10
},
{
"epoch": 1.8333333333333335,
"grad_norm": 2.0535888671875,
"learning_rate": 4.02454838991936e-06,
"loss": 0.9095,
"step": 11
},
{
"epoch": 2.0,
"grad_norm": 1.7272127866744995,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.9142,
"step": 12
},
{
"epoch": 2.1666666666666665,
"grad_norm": 1.1511447429656982,
"learning_rate": 2.2221488349019903e-06,
"loss": 0.8895,
"step": 13
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.8150257468223572,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.87,
"step": 14
},
{
"epoch": 2.5,
"grad_norm": 0.8002262115478516,
"learning_rate": 8.426519384872733e-07,
"loss": 0.8943,
"step": 15
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.7705777883529663,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.8858,
"step": 16
},
{
"epoch": 2.8333333333333335,
"grad_norm": 0.7509863376617432,
"learning_rate": 9.607359798384785e-08,
"loss": 0.8833,
"step": 17
},
{
"epoch": 3.0,
"grad_norm": 0.7269038558006287,
"learning_rate": 0.0,
"loss": 0.8752,
"step": 18
},
{
"epoch": 3.0,
"step": 18,
"total_flos": 1.2010811790553252e+18,
"train_loss": 1.01401486992836,
"train_runtime": 1429.1641,
"train_samples_per_second": 1.207,
"train_steps_per_second": 0.013
}
],
"logging_steps": 1.0,
"max_steps": 18,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.2010811790553252e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}