Files
lr-1e-05-epochs-1.0-summ-c3…/train_results.json

13 lines
376 B
JSON
Raw Normal View History

{
"entropy": 1.6328125,
"epoch": 1.0,
"mean_token_accuracy": 0.7037371794382731,
"num_tokens": 785524633.0,
"total_flos": 1254725330468864.0,
"train_loss": 1.377876113716714,
"train_runtime": 31760.1021,
"train_samples": 1871956,
"train_samples_per_second": 24.347,
"train_steps_per_second": 0.19,
"weight_norm": 259.0048262098604
}