OpenELM-1_1B-DPO-full-2 / train_results.json
CharlesLi's picture
Model save
079d30d verified
raw
history blame contribute delete
234 Bytes
{
"epoch": 2.998430141287284,
"total_flos": 0.0,
"train_loss": 0.26257153015070145,
"train_runtime": 12376.5138,
"train_samples": 61134,
"train_samples_per_second": 14.819,
"train_steps_per_second": 0.231
}