Custom model deployment on voxl-tflite-server
-
I have trained a custom model on Tensorflow and would like to deploy the same on Voxel 2. In the Voxel 2 documentation, it is mentioned that custom models can be deployed using v2.8.0 opsets. However, my model is dependent on opsets from the latest version of Tensorflow (v2.11.0). Can the voxl-tflite-server support opsets from the latest version of Tensorflow/tflite? If not, are there any other alternatives?
-
You can find the code for that TensorFlow Lite build here https://gitlab.com/voxl-public/voxl-sdk/third-party/qrb5165-tflite
There are just few patches applied to get it to build correctly. You could try rebasing it to 2.11