Closed acoj1993 closed 4 years ago
Hi @acoj1993, I'm glad that the files helped you! We had to invest a lot of time to make it work.
Yes, we have been using this Docker image for benchmarking he-transformer. The docker image includes the installation of the Python package (see here). Afterward you should be able to run source external/venv-tf-py3/bin/activate
and within that environment there is the package installed. As an example, you can try running the examples from the he-transformer repository.
No, we have been using the Docker image from the beginning on. Are you facing any issues trying to install it locally?
@acoj1993 As building the docker image took a very long time, you can also find a ready-to-go image at https://hub.docker.com/repository/docker/marblehe/base_ngraph-he.
Hey @pjattke.
Thank you! You have saved me some research time, again. I ran the examples successfully. Great! Also thanks for the docker image, building the image is indeed a pain in the ass. I have had like 5 or 6 unsuccessful builds that failed after some 2 hours of building, I can only assume that you guys did as well.
Cheers :)
Hi,
thanks a lot for this repository, you have saved me a lot of time!
I have a question regarding the interaction with the he-transformer. Have you been using the docker image for your own tests? If yes, how exactly? Have you also tried building the library directly on the Linux machine?
Thanks in advance.