Rano23
/

sft-16k

Generated from Trainer

Model card Files Files and versions

sft-16k / train_results.json

Rano23's picture

End of training

8b13783 verified about 1 year ago

225 Bytes

	{
	"epoch": 2.9993856739157145,
	"total_flos": 1.9337325951414436e+18,
	"train_loss": 0.40084019302589413,
	"train_runtime": 90458.8938,
	"train_samples_per_second": 2.159,
	"train_steps_per_second": 0.135
	}