-
Notifications
You must be signed in to change notification settings - Fork 807
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
Add int16 support to MINIMUM and MAXIMUM #2678
Conversation
Hi Andres, thanks for the PR. Seems like TfLite already has this so it would be great to support in TFLM as well. Could you add a unit test in the Thanks! |
Sure, will do. Can you point me to some documentation about how to run the unit tests in this project? |
To test with x86 use: |
@ddavis-2015 is this documented somewhere? I took a look at CONTRIBUTING.md but I didn't see anything. I'd open a PR with testing instructions if there isn't anything yet. |
Step 6. |
|
@andresovela ok looks good. Now for the procedural aspects of this process (we ALL have to do this): The code must pass Google code style checks, and there must be an issue # in your initial description. For the code style issues, find the code style issue test in the For the issue # please create an issue in this repo (it can just be a very short description of the problem). Then at the end of your initial description of this PR add: Finally, you can check that the Continuous Integration tests pass by going to |
@ddavis-2015 I fixed the code style issue but it looks like I don't have permissions to either add labels nor add reviewers to PRs |
@andresovela Sorry, didnt know about the permissions. All tests pass, reviewer added. |
Thanks again Andres, and thank you David for the help! |
This PR adds support for int16 to the MINIMUM and MAXIMUM kernels. This is necessary to work with some audio models using int16x8 quantization.
This is my first contribution to TFLM, let me know if there's anything else that needs to be done :)
bug=#2679