baseline-gemma-1.1-7b-it-sft / train_results.json
ZhangShenao's picture
Model save
10e371b verified
raw
history blame
252 Bytes
{
"epoch": 2.9959072305593453,
"total_flos": 1.338644246774258e+19,
"train_loss": 0.3336014450614244,
"train_runtime": 19884.0933,
"train_samples": 395000,
"train_samples_per_second": 7.076,
"train_steps_per_second": 0.055
}