YYYYYYibo
/

nash_dpo_rank4_on_vanilla_iter_3

alignment-handbook

Generated from Trainer

Model card Files Files and versions Community

nash_dpo_rank4_on_vanilla_iter_3 / train_results.json

YYYYYYibo's picture

Model save

5ffbfd9 verified 6 months ago

193 Bytes

	{
	"epoch": 1.0,
	"train_loss": 0.5940336465835572,
	"train_runtime": 8165.5058,
	"train_samples": 21135,
	"train_samples_per_second": 2.588,
	"train_steps_per_second": 0.02
	}