Can the quantized yolov5 model with Brevitas be deployed on the DPU (Deep Learning Processing Unit)?

Xilinx / brevitas

Brevitas: neural network quantization in PyTorch

https://xilinx.github.io/brevitas/

Other

1.19k stars 196 forks source link

Can the quantized yolov5 model with Brevitas be deployed on the DPU (Deep Learning Processing Unit)? #800

Open IkrameBeggar opened 9 months ago

IkrameBeggar commented 9 months ago

Hello, I want to quantize a yolov5 model and deploy it on ZCU104 using the DPU. I want to know if the quantized model with Brevitas can be compiled and deployed on the DPU.

ihsanalhafiz commented 7 months ago

I think that is a different thing. Since for using DPU, you need to compile the trained model with Vitis AI, the Vitis AI has a quantization feature. So, I don't see any need to use brevitas.