mbart-summarization-mlsum / train_results.json
artelabsuper
trained model
6424b22
raw
history blame
195 Bytes
{
"epoch": 4.0,
"train_loss": 2.0082999333540426,
"train_runtime": 37501.9411,
"train_samples": 31997,
"train_samples_per_second": 3.413,
"train_steps_per_second": 3.413
}