DUAL-GPO
/

zephyr-7b-gpo-iter0

alignment-handbook

Generated from Trainer

Model card Files Files and versions Metrics Training metrics Community

zephyr-7b-gpo-iter0 / train_results.json

lole25's picture

Model save

407a137 verified 3 months ago

raw history blame contribute delete

No virus

197 Bytes

	{
	"epoch": 2.0,
	"train_loss": 0.019690872263745406,
	"train_runtime": 18916.4989,
	"train_samples": 61135,
	"train_samples_per_second": 0.106,
	"train_steps_per_second": 0.053
	}