LlamaCorn-1.1B / train_results.json
jan-hq's picture
Model save
1242ba5 verified
raw
history blame
199 Bytes
{
"epoch": 3.0,
"train_loss": 0.9284211427668099,
"train_runtime": 113797.7954,
"train_samples": 1686013,
"train_samples_per_second": 11.145,
"train_steps_per_second": 0.174
}