ggerganov / llama.cpp

LLM inference in C/C++
MIT License
66.05k stars 9.49k forks source link

Feature Request: Could you please add the latest version of llama.cpp to dockerhub? #9223

Open vamshigvk opened 1 month ago

vamshigvk commented 1 month ago

Prerequisites

Feature Description

We are in need of a docker tag with some value on the dockerhub instead of tag:"latest", could you please push atleast once with the latest or any of the versions to your dockerhub with name "https://hub.docker.com/r/ggerganov/llamacpp-server"

Motivation

this will help us to pull the docker images from dockerhub instead of ghcr.io

Possible Implementation

No response

ngxson commented 1 month ago

We are in need of a docker tag with some value on the dockerhub instead of tag:"latest", could you please push atleast once with the latest or any of the versions to your dockerhub with name "https://hub.docker.com/r/ggerganov/llamacpp-server"

This is an experimental version, please do not use it.

ghcr.io registry works well, is there any reason why you don't want to use it?

vamshigvk commented 1 month ago

My team is not accepting the ghcr.io registry since the release names on github.com/ggerganov/llama.cpp repo and the tag names on ghcr.io registry are not matching. ( I noticed that you are using commit id as the tag id)

boxley-anaconda commented 1 month ago

Hi! I'm was looking for a pre-built image for llama-bench, and searching DockerHub found a match by @ggerganov: https://hub.docker.com/r/ggerganov/llama-bench However, that page has nothing (a browser-friendly 404 page).

Are there plans to upload an image somewhere? Maybe I'm looking in the wrong place?