This repository has been archived by the owner on Jan 7, 2023. It is now read-only.
Releases: intel/optimized-models
Releases · intel/optimized-models
provide intel optimized models for pytorch and mxnet
pytorch:
a. Support resnet50 model inference
b. Support calibration
c. Support both int8 and fp32
mxnet:
a. Support wide&deep model training and inference
b. Support accuracy and performance evaluation in inference benchmark
c. Support calibration
d. Support int8 and fp32