joboffer's picture
Training in progress, step 30, checkpoint
d6f62ac verified
{
"best_metric": 2.0486373901367188,
"best_model_checkpoint": "miner_id_24/checkpoint-30",
"epoch": 0.11461318051575932,
"eval_steps": 5,
"global_step": 30,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0038204393505253103,
"eval_loss": 2.355255603790283,
"eval_runtime": 86.7091,
"eval_samples_per_second": 2.549,
"eval_steps_per_second": 0.646,
"step": 1
},
{
"epoch": 0.011461318051575931,
"grad_norm": 0.22935272753238678,
"learning_rate": 0.00012,
"loss": 2.368,
"step": 3
},
{
"epoch": 0.019102196752626553,
"eval_loss": 2.2753238677978516,
"eval_runtime": 87.5683,
"eval_samples_per_second": 2.524,
"eval_steps_per_second": 0.64,
"step": 5
},
{
"epoch": 0.022922636103151862,
"grad_norm": 0.3813602030277252,
"learning_rate": 0.0001992114701314478,
"loss": 2.2968,
"step": 6
},
{
"epoch": 0.034383954154727794,
"grad_norm": 0.19919824600219727,
"learning_rate": 0.00018763066800438636,
"loss": 2.2231,
"step": 9
},
{
"epoch": 0.038204393505253106,
"eval_loss": 2.1532111167907715,
"eval_runtime": 87.6284,
"eval_samples_per_second": 2.522,
"eval_steps_per_second": 0.639,
"step": 10
},
{
"epoch": 0.045845272206303724,
"grad_norm": 0.15824028849601746,
"learning_rate": 0.000163742398974869,
"loss": 2.1492,
"step": 12
},
{
"epoch": 0.05730659025787966,
"grad_norm": 0.16715355217456818,
"learning_rate": 0.00013090169943749476,
"loss": 2.1101,
"step": 15
},
{
"epoch": 0.05730659025787966,
"eval_loss": 2.094489097595215,
"eval_runtime": 87.493,
"eval_samples_per_second": 2.526,
"eval_steps_per_second": 0.64,
"step": 15
},
{
"epoch": 0.06876790830945559,
"grad_norm": 0.14857010543346405,
"learning_rate": 9.372094804706867e-05,
"loss": 2.1153,
"step": 18
},
{
"epoch": 0.07640878701050621,
"eval_loss": 2.063730001449585,
"eval_runtime": 87.4237,
"eval_samples_per_second": 2.528,
"eval_steps_per_second": 0.641,
"step": 20
},
{
"epoch": 0.08022922636103152,
"grad_norm": 0.13450752198696136,
"learning_rate": 5.7422070843492734e-05,
"loss": 2.0838,
"step": 21
},
{
"epoch": 0.09169054441260745,
"grad_norm": 0.13830527663230896,
"learning_rate": 2.7103137257858868e-05,
"loss": 2.08,
"step": 24
},
{
"epoch": 0.09551098376313276,
"eval_loss": 2.0513598918914795,
"eval_runtime": 87.4326,
"eval_samples_per_second": 2.528,
"eval_steps_per_second": 0.64,
"step": 25
},
{
"epoch": 0.10315186246418338,
"grad_norm": 0.14177772402763367,
"learning_rate": 7.022351411174866e-06,
"loss": 2.1142,
"step": 27
},
{
"epoch": 0.11461318051575932,
"grad_norm": 0.15173637866973877,
"learning_rate": 0.0,
"loss": 2.0229,
"step": 30
},
{
"epoch": 0.11461318051575932,
"eval_loss": 2.0486373901367188,
"eval_runtime": 87.4288,
"eval_samples_per_second": 2.528,
"eval_steps_per_second": 0.641,
"step": 30
}
],
"logging_steps": 3,
"max_steps": 30,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 10,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 1,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7.6173476364288e+16,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}