DUAL-GPO-2
/

phi-2-sft-lora-chat

alignment-handbook

Generated from Trainer

Model card Files Files and versions Metrics Training metrics Community

phi-2-sft-lora-chat / train_results.json

lole25's picture

Model save

b0895e0 verified 6 months ago

196 Bytes

	{
	"epoch": 0.02,
	"train_loss": 1.2791897563246992,
	"train_runtime": 2444.0756,
	"train_samples": 207865,
	"train_samples_per_second": 2.168,
	"train_steps_per_second": 0.045
	}