Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

There are differences in the test results of .pth and .onnx #30

Open
lalawayne opened this issue Nov 6, 2024 · 0 comments
Open

There are differences in the test results of .pth and .onnx #30

lalawayne opened this issue Nov 6, 2024 · 0 comments

Comments

@lalawayne
Copy link

Hello, first of all, thank you very much for providing the onnx model framework. When I used your Depth anything V2 vits.onnx model to test my own data set, there was a problem. The test results were different from the pytorch version of V2 vits.pth. The test results of the pytorch version were significantly better than the onnx model.
Below is the code I used for testing:
Pytorch version:
python run.py --img-path ... --outdir ... --encoder vits
ONNX version:
python dynamo.py export --encoder vits --output weights/vits.onnx --opset 17
python dynamo.py infer weights/vits.onnx -i ... -o ...
Other parameters are default parameters.
In addition, I found that the depth map tested with the .onnx model is a grayscale image, why not a color one? Sincerely hope to get your reply.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant