d1_code_shortest_1k / train_results.json
ryanmarten's picture
End of training
76b0a52 verified
{
"epoch": 6.72,
"total_flos": 2.5890632312553472e+17,
"train_loss": 0.6343953520059585,
"train_runtime": 9236.5188,
"train_samples_per_second": 0.758,
"train_steps_per_second": 0.008
}