Qwen-sft-ls-v0.1 / all_results.json
sci-m-wang's picture
Upload 16 files
eabb226 verified
raw
history blame contribute delete
222 Bytes
{
"epoch": 9.987995198079231,
"total_flos": 8.185270082173747e+17,
"train_loss": 0.7521515275423344,
"train_runtime": 19400.1091,
"train_samples_per_second": 1.717,
"train_steps_per_second": 0.107
}