Files
ComfyUI-ARM64-Docker/comfyui-rocm-dockerfile
T
derek e3f2375e4c
Docker Build and Release Upload / build-rocm (push) Failing after 28s
Docker Build and Release Upload / build-uvicorn (push) Has been cancelled
Docker Build and Release Upload / build-cuda (push) Has been cancelled
build with no libraries and pass in the device via compose
2026-04-16 19:15:17 -07:00

41 lines
1.0 KiB
Plaintext

# Use the official ROCm base image
FROM ubuntu:24.04
# Set environment variables for 7900 XTX (gfx1100)
ENV HSA_OVERRIDE_GFX_VERSION=11.0.0
ENV TORCH_ROCM_AOTRITON_ENABLE_EXPERIMENTAL=1
ENV PYTHONUNBUFFERED=1
# Install basic Python and Graphics libs
RUN apt update -y && apt install -y \
python3-pip python3-dev git libgl1-mesa-glx libglib2.0-0 \
&& rm -rf /var/lib/apt/lists/*
# Set working directory
WORKDIR /app
# Clone ComfyUI
RUN git clone https://github.com/comfyanonymous/ComfyUI.git .
# Install PyTorch with ROCm 6.2 support (matching the base image)
RUN pip3 install --no-cache-dir torch torchvision torchaudio \
--index-url https://download.pytorch.org/whl/rocm6.2
# Install ComfyUI dependencies
RUN pip3 install --no-cache-dir -r requirements.txt
# Set custom Nodes
WORKDIR /app/custom_nodes
# Install ComfyUi Manager
RUN git clone https://github.com/Comfy-Org/ComfyUI-Manager.git
# Go back to root
WORKDIR /app
# Expose the default ComfyUI port
EXPOSE 8188
# Launch ComfyUI
CMD ["python3", "main.py", "--listen", "0.0.0.0"]