-
Notifications
You must be signed in to change notification settings - Fork 75
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
realDivide missing in tfjs-node #12
Comments
This is a great find (and sorry for the delay in response, @danwexler ). Would you be open to making this a PR? |
Yep. Will do. FYI, I have found a number of memory leaks in the code as well, which are particularly visible when running in Node on a GPU. I'll try to make a separate PR for those fixes as well. |
Awesome, looking forward to it! If you haven't started yet, make sure to pull master. I pushed a refactor that moves code into |
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions. |
Whoops, I didn't realize this was already fixed by https://github.com/thekevinscott/UpscalerJS/pull/60/files |
The tf.div in models/idealo-gans.js fails when running using tfjs-node on Ubuntu. I get the following error:
'realDivide' not yet implemented or not found in the registry. This kernel may not be supported by the tfjs backend you have chosen
I do NOT see this error running using the CPU backend on OSX. Everything works fine. On Ubuntu it fails. I've tested this using the 1.15.5 tensorflow gpu installation with both the latest tfjs-node-gpu@2.8.5 and with the earlier version that matches the upscaler package.json, tfjs-node-gpu@2.1.0. I have not yet dug into the TFJS or TF side of things to find the bug.
A simple workaround is:
The text was updated successfully, but these errors were encountered: