tianzhihui-isc's picture
End of training
fd13d59
raw
history blame
389 Bytes
{
"epoch": 5.96,
"eval_accuracy": 0.8028747433264887,
"eval_loss": 3.8734209537506104,
"eval_runtime": 5.8108,
"eval_samples_per_second": 83.809,
"eval_steps_per_second": 2.753,
"total_flos": 2.0253789729052508e+18,
"train_loss": 4.31497132544424,
"train_runtime": 1061.9918,
"train_samples_per_second": 24.757,
"train_steps_per_second": 0.192
}