test4-2.5-72B / all_results.json
raphgg's picture
add weights
d69da8b
raw
history blame contribute delete
220 Bytes
{
"epoch": 0.9968944099378882,
"total_flos": 1412529456676864.0,
"train_loss": 1.4701230704227342,
"train_runtime": 47403.0675,
"train_samples_per_second": 4.347,
"train_steps_per_second": 0.002
}