Implementation of image size agnostic model in FastPathology #173
-
We are attempting to integrate a model in FastPathology where the image size can be arbitrary. We export the ONNX model with the following dynamic axis (the pytorch lightning to_onnx() is used) :
And then using the following attributes in the pipeline:
We however get the following error message:
If we change the order of the input-nodes attribute values to
Is this a bug or is there a way to solve this? Full pipeline:
|
Beta Was this translation helpful? Give feedback.
Answered by
smistad
Sep 11, 2023
Replies: 1 comment 9 replies
-
The support for dynamic input shapes is a bit buggy. It has been improved in FAST version 4.7 which is not released yet. You could download one of the latest artifacts and test |
Beta Was this translation helpful? Give feedback.
9 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Yes, but TensorRT requires you to set the min, mean and max shapes which is not possible to do with text pipelines yet. With OpenVINO you are recommended to set these limits, but you don't have to.