DUAL-GPO
/

phi-2-gpo-test-longest-iter-random-0

alignment-handbook

Generated from Trainer

Model card Files Files and versions Community

phi-2-gpo-test-longest-iter-random-0 / train_results.json

BraylonDash's picture

Model save

4b97d3f verified 3 months ago

raw history blame contribute delete

No virus

197 Bytes

	{
	"epoch": 3.97,
	"train_loss": 0.0003399105843263603,
	"train_runtime": 3335.538,
	"train_samples": 61135,
	"train_samples_per_second": 1.199,
	"train_steps_per_second": 0.074
	}