b2_math_askllm_1k / train_results.json
ryanmarten's picture
End of training
7f6db8b verified
raw
history blame contribute delete
208 Bytes
{
"epoch": 6.72,
"total_flos": 1.3266552166862029e+17,
"train_loss": 0.3581548386386463,
"train_runtime": 5222.7845,
"train_samples_per_second": 1.34,
"train_steps_per_second": 0.013
}