zephyr-7b-dpo-lora-pairrm / train_results.json
shenxq's picture
Model save
a1c8240 verified
raw
history blame contribute delete
195 Bytes
{
"epoch": 1.0,
"train_loss": 0.6566078573155155,
"train_runtime": 38377.8314,
"train_samples": 19996,
"train_samples_per_second": 0.521,
"train_steps_per_second": 0.033
}