search

Xilinx / Vitis-AI

Vitis AI is Xilinx’s development stack for AI inference on Xilinx hardware platforms, including both edge devices and Alveo cards.
https://www.xilinx.com/ai
Apache License 2.0
1.48k stars 628 forks source link

Can not find PyTorch and Tensorflow2 Optimizer Tools in VAI 2.0 Docker #653

Closed jia0511 closed 2 years ago

jia0511 commented 2 years ago

Hello xilinx,

Judging from the vitis-ai2.0 information you released as follow, it can be seen that TF2 and pytorch pruning tools are currently supported, but after I entered, there is only the vaitrace tool, not the vai_p tool. Did you miss it or do you need other configurations?

Optimizer TensorFlow 1.15 Support tf.keras.Optimizer for model training TensorFlow 2.x Support TensorFlow 2.3-2.6 Add iterative pruning PyTorch Support PyTorch 1.4-1.9.1 Support shared parameters in pruning Add one-step pruning Add once-for-all(OFA) Unified APIs for iterative and one-step pruning Enable pruned model to be used by quantizer Support nn.Conv3d and nn.ConvTranspose3d

gau-nernst commented 2 years ago

I think the pre-built CPU image does not include optimizer tools. You need to build your on GPU docker image for those.

jia0511 commented 2 years ago

My problem occurred in the GPU version, and the GPU version built by other colleagues has this problem, and the vai_p_tensorflow2 or vai_p_pytorch tools cannot be found.

white0dew commented 2 years ago

My problem occurred in the GPU version, and the GPU version built by other colleagues has this problem, and the vai_p_tensorflow2 or vai_p_pytorch tools cannot be found.

hello,would pls teach me how to download the GPU version built by other colleagues? Thank you!

jia0511 commented 2 years ago

Dear everyone,

I found that this interface does not exist in the GPU version in binary form. pytorch is encapsulated through the nndct python interface, and the same is true for tf2. The example is in the tools/Vitis-AI-Optimizer directory, so I close this ticket. Thank you!