SP Mohanty
SP Mohanty
@eostermueller : How did you solve this issue ?
@jzf2101 : Good to seem some momentum around this. repo2docker with GPU support is something we have been actively using on [crowdAI](https://www.crowdai.org). We managed to collect about 800 repositories (and...
@betatim : I quickly skimmed through it, but I think I understand whats hapenning with it. In case of GPU support for the k8 service on GKE, which I think...
@jzf2101 : Ahh ! well `nvidia-docker` will install the basic nvidia drivers depending on the gpu on the host machine, but not the full cuda toolkit. The nvidia drivers enable...
@jzf2101 : No, when we use crowdai-repo2docker, then we do not even need `nvidia-docker`, the goal there is to build a CUDA capable docker image. When we run the docker...
whoever is trying to work on this issue, can refer to my comments on https://github.com/ffledgling/dtella/commit/57f5514a973a3c7b2ed444d07b3fc3e95612b70a#diff-294f88bdf22b65e7d12deb35ed329d35R175 for the why and hows to fix this.
@kimchicanh : Did you add your collaborator to the gitlab.aicrowd.com repository ? And was his SSH keys etc configured to work with gitlab.aicrowd.com ?
@Sakari1994 : How much RAM do you have on your machine ? The script there basically tries to build a docker image, so as long as you have enough ram...
@Sakari1994 : Can you try adding the flag "--build-memory-limit=4g" inside the building.sh script ?
Weird !! Anyone else facing this same issue ? @waleedgondal : Can you try this on your setup, and monitor the memory usage ? Maybe use a low build-memory-limit to...