-
Notifications
You must be signed in to change notification settings - Fork 2.8k
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
Model editing support #275
Comments
Is it possible to edit names of input and output layers of onnx model? |
There's one more scenario happened frequently is, do some manual constant folding optimization. Say, it's known that a sub-graph can be constant-folded, and user would like to directly remove the sub-graph and replace it with an initializer. |
@linkerzhang could this be implemented as an ONNX Optimizer? |
@lutzroeder Editing would be great. I really would like to use it for training with yolo and darknet. In fact I found your tool googleing exactly for a way to edit the files with a gui |
does netron support editing now? |
Editing feature in Netron itself would be very cool!! I love the Netron UI for ONNX model inspection and all information is presented in one integrated fashion! |
Even just basic editing support would be amazing:
|
Since models can be edited through the onnx python API e.g here, it would be a possible way of resolving this to introduce a way to run python script, and then the community could add their own scripts into the editor. |
This could be helpful for anyone looking for something similar: https://github.com/ZhangGe6/onnx-modifier |
Interested in this feature too and like to contribute.
The current protoc.js and protobuf.js only have decoder/reader. |
Yes, extend the existing libraries instead of adding dependencies. |
I have an initial implementation for protobuf encoding (#995) |
Bump! Any news? |
bump |
This work is still active and could be help |
Is there any plan to have Netron support model editing please? This will be a very useful feature
The text was updated successfully, but these errors were encountered: