c1_math_0d_4s_1k / train_results.json
ryanmarten's picture
Upload model
737c19f verified
{
"epoch": 6.666666666666667,
"total_flos": 2.3990446016220365e+17,
"train_loss": 0.4939488410949707,
"train_runtime": 2825.0166,
"train_samples_per_second": 2.478,
"train_steps_per_second": 0.025
}