Qwen2.5-7B-Instruct-method3 / all_results.json
luowenyang's picture
Upload folder using huggingface_hub
ca93f20 verified
raw
history blame contribute delete
203 Bytes
{
"epoch": 3.0,
"total_flos": 150195282509824.0,
"train_loss": 0.3441459048878063,
"train_runtime": 5442.3628,
"train_samples_per_second": 1.937,
"train_steps_per_second": 0.243
}