deepseek-coder-1.3b-base-JUnit / train_results.json
iHateNLP's picture
End of training
d4fa5be verified
raw
history blame contribute delete
286 Bytes
{
"epoch": 1.9985964912280703,
"raw_train_examples": 78434,
"total_flos": 8.025299762167153e+17,
"train_examples": 39900,
"train_loss": 0.37913749282088366,
"train_runtime": 28381.1385,
"train_samples_per_second": 2.812,
"train_steps_per_second": 0.088
}