Keep the attention sink and ring caches when the KV cache is quantized - #24473

Keep the attention sink and ring caches when the KV cache is quantized#24473

pssrawat wants to merge 1 commit into

pytorch:mainpytorch/executorch:mainfrom

pssrawat:export-D126929701pssrawat/executorch:export-D126929701Copy head branch name to clipboard