Regarding the batchsize support issue for NPU 能否支持大于1的batch_size?

[postgraduate0]

I saw in the documentation that Ryzen AI only supports batch size 1, which cannot be modified. I tried changing dummy_input to dummy_input (8,3，224，224) when converting the ResNet model format, and both conversion and quantization were fine. However, when reasoning on NPU, an error still occurred:

So the current Ryzen AI architecture only supports Batch Size=1, is this a hard constraint? If it weren't how to modify it to support multiple batches

[dwithchenna]

yes, we currently support batch size = 1