llama3-8b-coding-gpt4o-100k / train_results.json
chansung's picture
Model save
a7bcb23 verified
raw
history blame contribute delete
239 Bytes
{
"epoch": 10.0,
"total_flos": 7.982333171800736e+18,
"train_loss": 0.3983345487382677,
"train_runtime": 11606.8402,
"train_samples": 116368,
"train_samples_per_second": 14.865,
"train_steps_per_second": 0.116
}