mind / wheels

Performance-optimized wheels for TensorFlow (SSE, AVX, FMA, XLA, MPI)
886 stars 108 forks source link

tensorflow-gpu 1.8.0-rc1 wheel for cuda9.1 cudnn7.1 nccl2.1 py36 linux #40

Closed kirk86 closed 6 years ago

kirk86 commented 6 years ago

Hi everyone I have just built a wheel for llinux with cuda 9.1 cudnn 7.1, nccl2.1 and python 3.6. You can find it in the following link! Maybe you wanna add it to the repository with the rest of the community wheels?

Cheers!

danqing commented 6 years ago

Thanks for providing the wheel. You may want to add it here: https://github.com/yaroslavvb/tensorflow-community-wheels.

kirk86 commented 6 years ago

@danqing Thanks for the suggestion, I keep getting the following error

gsutil cp "tensorflow-1.8.0-cp36-cp36m-linux_x86_64.whl" "gs://tensorflow-community-wheels"
Copying file://tensorflow-1.8.0-cp36-cp36m-linux_x86_64.whl [Content-Type=application/octet-stream]...
ResumableUploadAbortException: 401 Anonymous caller does not have storage.objects.create access to tensorflow-community-wheels.

Any ideas what I'm missing. I'm using gsutil from pip. If I just issue gsutil config then it asks to pinpoint to a project id and nowhere there is tensorflow-community-wheels listed. I believe it must be sth really silly that I'm missing here?

danqing commented 6 years ago

I'm not quite sure what you are trying to achieve - are you looking for a GCS bucket? I'm not sure what that bucket is, but apparently it doesn't support anonymous writes.

kirk86 commented 6 years ago

@danqing Thanks for getting back to me. As per the instructions here, it is shown how to build and in the end of those instructions it shows how to copy your newly built wheel into the tensorflow-community-wheels google storage.

I'm not quite sure what you are trying to achieve

I'm trying to copy my wheel into the tensorflow-community-wheels google storage as you suggested in your first thread as well.