[CoreML] Cache the quantizer's pattern graphs instead of rebuilding them per model - #23323

[CoreML] Cache the quantizer's pattern graphs instead of rebuilding them per model#23323

vijay-kapse wants to merge 5 commits into

pytorch:mainpytorch/executorch:mainfrom

vijay-kapse:perf/cache-coreml-quantizer-patternsvijay-kapse/executorch:perf/cache-coreml-quantizer-patternsCopy head branch name to clipboard