Qwen2.5-7B-Multi-Parallel / all_results.json
PumeTu's picture
Upload folder using huggingface_hub
3ebeae9 verified
raw
history blame
221 Bytes
{
"epoch": 2.9835189309576835,
"total_flos": 234603226333184.0,
"train_loss": 0.20078422256878445,
"train_runtime": 13460.8563,
"train_samples_per_second": 16.011,
"train_steps_per_second": 0.031
}