Open msj121 opened 1 year ago
ATM the tensorflow ports in a bad shape with regard to windows suport, patched but not building in CI. I'm currently looking at fixing at least tensorflow
for windows CI, and would try an update after that.
(Past port updates without CI coverage contributed to the problems.)
I could use this on mac m2/intel as well.
This is an automated message. Per our repo policy, stale issues get closed if there has been no activity in the past 180 days. The issue will be automatically closed in 14 days. If you wish to keep this issue open, please add a new comment.
This is an automated message. Per our repo policy, stale issues get closed if there has been no activity in the past 180 days. The issue will be automatically closed in 14 days. If you wish to keep this issue open, please add a new comment.
Tensorflow is very much out of date, with 2.18.0 coming soon.
Maybe most of us would be happy with tensorflow lite (for shipping software), and could do the machine learning using the distribution's packaging.
With tensorflow and bazel progress, it seems to get even harder to form a good port: bazel is very much designed to make hard what vcpkg wants: inject (vcpkg) toolchain settings and (vcpkg) system libs.
There's a cmake build system for tensorflow lite.
https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/CMakeLists.txt
Library name
tensorflow
New version number
2.12.0
Other information that may be useful (release notes, etc...)
It is important to be able to access models, older versions will fail on new models.
Thank you so much!