View on GitHub

docker-rocm-xtra

ROCm docker images with fixes/support for extra architectures, such as gfx803/gfx1010.

Intro

llama.cpp + ROCm Support

Usage

Create drun alias (if it doesn’t exist yet)


alias  drun='sudo docker run -it --network=host --device=/dev/kfd --device=/dev/dri --ipc=host --shm-size 16G --group-add video --cap-add=SYS_PTRACE --security-opt seccomp=unconfined -v $HOME/dockerx:/dockerx -v $(pwd):/current'

Example Usage


drun --pull=always --rm ulyssesrr/rocm-xtra-llamacpp:latest \
    -m /current/models/nous-hermes-llama2-13b.ggmlv3.q5_K_M.bin \
    -t 12 \
    -c 2048 \
    --mirostat 2 \
    --mirostat-lr 0.1 \
    --mirostat-ent 5.0 \
    --color \
    --interactive-first \
    --keep -1 \
    --in-suffix "### Response: " \
    --reverse-prompt "### Instruction:" \
    -p "### Instruction: " \
    --gpu-layers 24

Note: /current maps to the current directory in the host from where drun was invoked.