CharlesLi
/

mistral_rlhf_safe_llama_3_70B_reflect_1000_full

Text Generation

alignment-handbook

Generated from Trainer

text-generation-inference

Inference Endpoints

Model card Files Files and versions Metrics Training metrics Community

mistral_rlhf_safe_llama_3_70B_reflect_1000_full / eval_results.json

CharlesLi's picture

End of training

baa0438 verified 1 day ago

183 Bytes

	{
	"epoch": 1.0,
	"eval_loss": 0.5957714915275574,
	"eval_runtime": 0.7322,
	"eval_samples": 20,
	"eval_samples_per_second": 4.097,
	"eval_steps_per_second": 1.366
	}