problem with the ggml files when using the docker images

[dnalexen]

I have a problem with the ggml images when I use the docker image ghcr.io/ggerganov/llama.cpp:full. I am asked to regenerate your ggml files but I have already done that. See attached screenshot.

[athrael-soju]

If your model is already quantized, then you don't need to use --all-in-one anymore.

something like this should work:

docker run -v /home/ndjossou/Documents/llama_for_docker/models:/models ghcr.io/ggerganov/llama.cpp:light -m models/7B/your_model_name.bin -p "hello" -n 128

[github-actions]

This issue was closed because it has been inactive for 14 days since being marked as stale.