llama3-3b-summarize-gpt4o-128k / train_results.json
chansung's picture
Model save
19985e7 verified
{
"epoch": 9.956521739130435,
"total_flos": 2.4191218891641324e+18,
"train_loss": 1.418152675716155,
"train_runtime": 3547.1741,
"train_samples": 129221,
"train_samples_per_second": 39.395,
"train_steps_per_second": 0.615
}