-
Notifications
You must be signed in to change notification settings - Fork 10k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Server: Unix Socket Support #6413
Conversation
This should be tested in all platforms |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Cool, please add a simple test scenario with unix socket family.
@FSSRepo @phymbert |
I suggest adding a simple dedicated scenario in a new |
Regarding server tests, @phymbert has provided quite good documentation over here: https://github.com/ggerganov/llama.cpp/tree/master/examples/server/tests One way to improve this even further and help new contributors to implement tests, is to reference a very small PR that introduces a basic server test, without any extra changes. I'm not sure if we have one yet - if not, we can create, and we can point people to that PR as a starting point for implementing new tests. |
Yes, a good example is: |
# Conflicts: # examples/server/server.cpp
The idea of this pull request is to ease integration of llama.cpp server using unix sockets instead tcp.
cpp-httplib has support for unix sockets built in: yhirose/cpp-httplib#1346
my idea was to not add an additional parameter, but use a --host prefix: unix:// (similar to docker's client/server pattern).
a very first attempt is here, mainly to understand if this is something you could imagine in the code.
(the file should not exist before) ./server --host unix:///tmp/llama.sock --model ~/Projects/models/mistral-7b-instruct-v0.2.Q4_K_M.gguf
connect using socat
connect using curl:
open points: