llama-3.2-3b-dpo-2 / train_results.json
tanliboy's picture
Model save
daad045 verified
raw
history blame
233 Bytes
{
"epoch": 2.998693948628646,
"total_flos": 0.0,
"train_loss": 0.5855818307081304,
"train_runtime": 16735.6732,
"train_samples": 73493,
"train_samples_per_second": 13.174,
"train_steps_per_second": 0.103
}