deepseek-8b-orpo-full / train_results.json
zfz1's picture
Model save
9adce4c verified
raw
history blame contribute delete
218 Bytes
{
"epoch": 2.0,
"total_flos": 0.0,
"train_loss": 0.2130171504020691,
"train_runtime": 11716.0982,
"train_samples": 20000,
"train_samples_per_second": 3.414,
"train_steps_per_second": 0.107
}