llama3-8b-ja / train_results.json
minhhien0811's picture
Add fine-tuned model
d878be5
{
"epoch": 2.0,
"total_flos": 192624989306880.0,
"train_loss": 1.1565231682577926,
"train_runtime": 4692.4705,
"train_samples_per_second": 38.067,
"train_steps_per_second": 0.297
}