groderg's picture
Evaluation on the test set completed on 2024_09_24.
52e23c4 verified
raw
history blame contribute delete
239 Bytes
{
"epoch": 28.0,
"learning_rate": 0.0001,
"total_flos": 2.778404267780425e+19,
"train_loss": 0.2165746406882549,
"train_runtime": 45987.1682,
"train_samples_per_second": 75.812,
"train_steps_per_second": 2.375
}