c1_math_0d_16s_10k / train_results.json
ryanmarten's picture
Upload model
55521bf verified
raw
history blame contribute delete
210 Bytes
{
"epoch": 4.992,
"total_flos": 1.8658329858534277e+18,
"train_loss": 0.3731630460956158,
"train_runtime": 19785.2158,
"train_samples_per_second": 2.527,
"train_steps_per_second": 0.02
}