[Feature Request] Extend quantization tool to support blocked quantization #20981
Labels
feature request
request for unsupported feature or enhancement
quantization
issues related to quantization
Describe the feature request
Onnx has recently introduced layers to support blocked quantization. It would be useful to extend the current quantization tool to support this new feature.
Describe scenario use case
This would allow us to quantize fp32 models in blockwise style
The text was updated successfully, but these errors were encountered: