ap-normistral-7b-sft-qlora / all_results.json
hugodk-sch's picture
End of training
bee5574 verified
raw
history blame
359 Bytes
{
"epoch": 5.0,
"eval_loss": 1.6054625511169434,
"eval_runtime": 67.8587,
"eval_samples": 468,
"eval_samples_per_second": 3.905,
"eval_steps_per_second": 0.501,
"train_loss": 1.0955926315350966,
"train_runtime": 10263.5013,
"train_samples": 4208,
"train_samples_per_second": 1.028,
"train_steps_per_second": 0.129
}