onnxruntime
Add option to allow quantize_input() use input_qtype for initializers.
#5721
Merged

Loading