LLaMA-O1-Base-1127 / train_results.json
qq8933's picture
Upload folder using huggingface_hub
8c9cb64 verified
raw
history blame contribute delete
208 Bytes
{
"epoch": 4.0,
"total_flos": 1.048251868267099e+19,
"train_loss": 0.24899632067771982,
"train_runtime": 10196.9918,
"train_samples_per_second": 2.78,
"train_steps_per_second": 0.116
}