01_moedl_dense-0119 / train_results.json
vuiseng9's picture
Initial Commit
7690a97
{
"epoch": 2.0,
"num_input_tokens_seen": 1027671040,
"total_flos": 6.947227234861056e+17,
"train_loss": 1.2691249875690298,
"train_runtime": 6762.0302,
"train_samples": 1003585,
"train_samples_per_second": 296.829,
"train_steps_per_second": 2.319
}