baseline-gemma-2-2b-it-sft / train_results.json
ZhangShenao's picture
Model save
2243e3c verified
{
"epoch": 2.9959072305593453,
"total_flos": 3.496351998494638e+18,
"train_loss": 0.2091554065246834,
"train_runtime": 8168.8424,
"train_samples": 395000,
"train_samples_per_second": 17.224,
"train_steps_per_second": 0.134
}