tinyllama-1.1b-sum-dpo-qlora / train_results.json
martimfasantos's picture
Model save
ea4ba97 verified
raw
history blame
200 Bytes
{
"epoch": 1.0,
"train_loss": 0.0003277428618961422,
"train_runtime": 17.7068,
"train_samples": 92858,
"train_samples_per_second": 5244.214,
"train_steps_per_second": 327.728
}