zephyr-7b-dpo-qlora / train_results.json
jiangjun0105woven's picture
Model save
aa9c5c8 verified
raw
history blame contribute delete
216 Bytes
{
"epoch": 1.0,
"total_flos": 0.0,
"train_loss": 0.5397420258873861,
"train_runtime": 14909.9591,
"train_samples": 61134,
"train_samples_per_second": 4.1,
"train_steps_per_second": 0.043
}