Qwen2.5-7B-sft-hhrlhf-gen-dpo / train_results.json
AmberYifan's picture
Model save
bca2cfc verified
raw
history blame contribute delete
216 Bytes
{
"epoch": 3.0,
"total_flos": 0.0,
"train_loss": 0.1578371215044373,
"train_runtime": 5401.3477,
"train_samples": 9999,
"train_samples_per_second": 5.554,
"train_steps_per_second": 0.174
}