zephyr-7b-sft-math_code / train_results.json
dlibf's picture
Model save
ffb2306 verified
raw
history blame contribute delete
197 Bytes
{
"epoch": 1.0,
"train_loss": 1.0355008869995306,
"train_runtime": 11406.2159,
"train_samples": 241210,
"train_samples_per_second": 12.721,
"train_steps_per_second": 0.099
}