lole25
/

phi-2-gpo-ultrachat-lora-2

alignment-handbook

Generated from Trainer

Model card Files Files and versions Metrics Training metrics Community

phi-2-gpo-ultrachat-lora-2 / train_results.json

lole25's picture

Model save

51b1f39 verified 8 months ago

195 Bytes

	{
	"epoch": 1.98,
	"train_loss": 0.00995884225458691,
	"train_runtime": 2568.9627,
	"train_samples": 6113,
	"train_samples_per_second": 4.759,
	"train_steps_per_second": 0.074
	}