jonra1993's picture
New model! 🤗
db26aa8
raw
history blame
344 Bytes
{
"epoch": 4.0,
"eval_loss": 0.5664765238761902,
"eval_runtime": 20.1946,
"eval_samples_per_second": 21.095,
"eval_steps_per_second": 2.674,
"total_flos": 2145730166784000.0,
"train_loss": 0.6086930384431831,
"train_runtime": 1288.334,
"train_samples_per_second": 6.374,
"train_steps_per_second": 0.798
}