fine-tuned / train_results.json
tejp's picture
tejp/custom_dataset
cd5054b
raw
history blame contribute delete
203 Bytes
{
"epoch": 3.0,
"total_flos": 3254926098493440.0,
"train_loss": 2.3485514322916665,
"train_runtime": 131.9053,
"train_samples_per_second": 0.318,
"train_steps_per_second": 0.023
}