DUAL-GPO
/

phi-2-gpo-test-longest-iter-random2-0

alignment-handbook

Generated from Trainer

Model card Files Files and versions Community

phi-2-gpo-test-longest-iter-random2-0 / train_results.json

BraylonDash's picture

Model save

ba1b430 verified 3 months ago

raw history blame contribute delete

No virus

198 Bytes

	{
	"epoch": 3.97,
	"train_loss": 0.0010566485528421077,
	"train_runtime": 4635.3882,
	"train_samples": 61135,
	"train_samples_per_second": 0.863,
	"train_steps_per_second": 0.054
	}