List your company

LLM Alignment on AlpacaEval Length-Controlled (test)

8.78LC Win RateUNA-score (MSE)
-0.09122.21194.5156.8181Aug 27, 2024

Evaluation results

MethodLinks
UNA-score (MSE)Base Model=Mistral, Fine-tuning data=HelpSteer22024.088.78
UNA-score (MSE)Base Model=Llama, Fine-tuning data=HelpSteer22024.087.87
UNA-binary (BCE)Base Model=Mistral, Fine-tuning data=HelpSteer22024.087.41
KTOBase Model=Mistral, Fine-tuning data=HelpSteer22024.084.46
KTOBase Model=Llama, Fine-tuning data=HelpSteer22024.084.17
UNA-binary (BCE)Base Model=Llama, Fine-tuning data=HelpSteer22024.083.96
DPOBase Model=Mistral, Fine-tuning data=HelpSteer22024.083.67
UNA-pairwiseBase Model=Mistral, Fine-tuning data=HelpSteer22024.083.67
DPOBase Model=Llama, Fine-tuning data=HelpSteer22024.082.09
UNA-pairwiseBase Model=Llama, Fine-tuning data=HelpSteer22024.082.09
MistralBase Model=Mistral, Fine-tuning data=HelpSteer22024.080.31
LlamaBase Model=Llama, Fine-tuning data=HelpSteer22024.080.25