search

Xilinx / Vitis-AI-Tutorials

MIT License
380 stars 145 forks source link

Python API execute_async multiple outputs #73

Closed just-an-intern closed 2 years ago

just-an-intern commented 2 years ago

Hello,

I am trying to use my compiled CNN model using the Python API and the "dpu.execute_async(inputData,outputData)" function. However I am not able to recover the output Tensors after using the Runner. My script is based on the MNIST Tutorial script. My subgraph has this form: image And I have multiple output sizes (after the FPN), I am not sure how to initialize the buffer (as a np.empty() array).

How can I fix this ?

Thank you for your help.

bhargavin1872008 commented 2 years ago

when running the requirements.txt of keras-yolov3-modelset -i 'm getting error for coremltools.it is showing like "couldn't find a version that satisfies the requirement tensorflow<=1.14 and tensorflow >=1.5(from tfcoremltools -r requirements.txt).(from version :2.2.0,2.2..1, 2.2.2, ...2.7.0rc0,2.7.0.rc1............) like this .can someone help me regarding this. Also ,i have a doubt .can we use ubuntu 20.04 ,cuda 11.7 ,cudnn 8.4.0 for this project. or have to use ubuntu 18.04,cuda 10.0 only which only works.please help me regarding this,i have less time in my hand.

just-an-intern commented 2 years ago

I found a solution that uses an example design that does exactly this : https://support.xilinx.com/s/article/Multiple-output-model-example-design-from-training-to-application-on-ZCU102?language=en_US