c1_math_0d_16s_1k / train_results.json
ryanmarten's picture
Upload model
dacc396 verified
raw
history blame contribute delete
222 Bytes
{
"epoch": 6.666666666666667,
"total_flos": 2.3541271679205376e+17,
"train_loss": 0.4919714012316295,
"train_runtime": 2750.5888,
"train_samples_per_second": 2.545,
"train_steps_per_second": 0.025
}