Replies: 1 comment 1 reply
-
Hi @schell, thanks for your interest in the project. Right now wonnx is unmaintained. I do not have a use case for it to justify working on it, and @haixuanTao seems to have moved on to other projects as well. Since the last release, the ecosystem around ML and GPU acceleration has greatly evolved and there are now many more and better options available, such as WebLLM and tract. Wonnx does not have any specific advantage over the current alternatives and is probably a lot slower. That said it still is a very educative example of how to run large models on a GPU in a crossplatform way, and on how to process ONNX models into usable form. We should probably consider archiving the GitHub repository though. |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Hello, I was just wondering about the status of this project. Is it being worked on? Will we see another release? What are/were the hurdles to getting to 1.0?
Beta Was this translation helpful? Give feedback.
All reactions