Sergio López
Sergio López
@jakecorrenti if you have a minute, please let me know if I have addressed all your comments. I'd like to get this one and #288 for the next release.
At this very moment I'm refactoring the network configuration API to support multiple network interfaces. I'm going to incorporate the ability to specify the offloading flags too.
This is the PR https://github.com/containers/libkrun/pull/367
@sroecker is llama.cpp working properly with you with a container generated from this Containerfile? Which models have you tested? I'm asking because the Vulkan backend hasn't worked for me since...
Tested with Mistral-7B and Wizard-Vicuna-13B and got random answers with both of them. Sadly, the Vulkan backend is still broken for Apple Silicon GPUs upstream. I think we're going to...
This one got superseded by #212
This should be fixed with libkrun `1.12.0` and later.
I'm setting up a new repo to simplify testing. I'll put a reference here once the repo is ready.
> Do we plan to have a crun package for cs9 that supports libkrun? . Since "krun" is really just a different name for crun, telling crun to behave in...
The root cause on this and #1688 is the same. In libkrun we don't support relying the exit code from the guest to the host. But, for container use case,...