plan of deep learning team for next stable release?
Hi fellow developers,
I've just uploaded pytorch 1.7.0-2 to unstable, which should compile on
at least amd64 arm64 and ppc64el. There is no any non-cpu hardware
acceleration enabled yet. I think my next step is to get the whole
dependency tree of pytorch into testing, so they will be shipped by the
next stable release.
Apart from that, I think my remaining works for the next release mainly
involves the upload of a new upstream release of opencv, and the ensuing
transition.
I tried to build src:tensorflow from the master branch of the salsa git
repository, but it ended up failing due to bazel's complain on some
unsupported flags. (which does not surprise me at all as I still have
an impression on bazel's volatile cmdline interface).
SYCL is not yet upstreamed to the llvm project. So this is not something
to consider for short term.
ROCm gave us a bad impression through its documentations. And I think
we're out of man power to deal with it.
Reply to: