Andrew Murray
Andrew Murray
@sla-te to explain a bit, we never intentionally supported the conversion to TensorFlow that you're performing. We have a mechanism to allow conversion to NumPy. For a while there, we...
@sla-te are you happy for this to be closed? You could instead monitor the TensorFlow issue.
The TensorFlow issue has been resolved, addressed in https://github.com/tensorflow/tensorflow/commit/8b3369b3cb4c3ecc3e0059a550c2d7cad136d371. This should be fixed in a future release of TensorFlow.
Does the image size limit imposed by #2583 mean that this scenario is no longer viable?
In Pillow, we translate input data into a few standard modes. So, for example, we don't have a mode for float64. Instead, we unpack a variety of floating point types...
@kmilos are you saying that all "I" modes should go back to NumPy as uint32? So in the context of this issue, that would also mean int8, int16 and int32...
If there are negative values in the result, that sounds to me like Pillow has created an invalid image.
When I try running your Python code, the following dialog appears every time. That doesn't seem like the automatic behaviour that we otherwise have in ImageGrab. ![Screen Shot 2022-06-25 at...
I don't think that opening an issue about it would lead to success. https://github.com/flatpak/xdg-desktop-portal/issues/649#issuecomment-974830861 > Bypassing the screenshot portal is unacceptable, as it defeats the point of having the portal...