Minbyul
/

mistral-7b-dpo-full-wo-live_qa-ep3

Text Generation

alignment-handbook

Generated from Trainer

Inference Endpoints

text-generation-inference

Model card Files Files and versions Community

mistral-7b-dpo-full-wo-live_qa-ep3 / eval_results.json

Minbyul's picture

End of training

8c24609 verified 2 months ago

raw history blame contribute delete

No virus

560 Bytes

	{
	"epoch": 1.0,
	"eval_logits/chosen": -3.3306896686553955,
	"eval_logits/rejected": -3.137249708175659,
	"eval_logps/chosen": -1001.12890625,
	"eval_logps/rejected": -1768.898193359375,
	"eval_loss": 0.2304462045431137,
	"eval_rewards/accuracies": 0.8472222089767456,
	"eval_rewards/chosen": -1.4963774681091309,
	"eval_rewards/margins": 7.075459003448486,
	"eval_rewards/rejected": -8.5718355178833,
	"eval_runtime": 34.9747,
	"eval_samples": 284,
	"eval_samples_per_second": 8.12,
	"eval_steps_per_second": 0.257
	}