tdpeuter/open-webui
tdpeuter/open-webui
1
Fork 0
forked from open-webui/open-webui
Code Issues Pull requests 1 Projects 1 Releases 2 Packages Activity Actions
open-webui/Dockerfile-cuda

86 lines
3.1 KiB
Text
Raw Normal View History

Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
# syntax=docker/dockerfile:1
Parametrize CUDA_VERSION in Dockerfile Standardized CUDA_VERSION as a global ARG to ensure consistency and facilitate version updates across the Dockerfile. This change allows the CUDA version to be defined once at the beginning and reused, reducing the chance of mismatched versions and easing maintenance when changing CUDA versions. It further streamlines the build process for potential multi-stage builds with varying CUDA dependencies. Refs #nvidia-update
2024-03-17 07:27:06 +01:00
ARG CUDA_VERSION=12.3.2
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
######## WebUI frontend ########
FROM node:21-alpine3.19 as build
WORKDIR /app
COPY package.json package-lock.json ./
RUN npm ci
COPY . .
RUN npm run build
Optimize Dockerfile for CUDA support Refactored the Dockerfile to better organize and streamline environment variable settings, emphasizing support for a CUDA-based WebUI backend while retaining the ability to build a CPU-only image. Consolidated ENV commands to reduce layers, improving build efficiency, and set a default PORT environment to enhance container usability. Enabled exposure of the backend service on port 8080 and leveraged combined RUN directives to minimize the image footprint. These changes facilitate a more robust deployment process, catering to both CPU and CUDA environments.
2024-03-17 06:55:37 +01:00
######## CPU-only WebUI backend ########
# To support both CPU and GPU backend, we need to keep the ability to build the CPU-only image.
#FROM python:3.11-slim-bookworm as base
#FROM --platform=linux/amd64 ubuntu:22.04 AS cpu-builder-amd64
#FROM --platform=linux/amd64 cpu-builder-amd64 AS cpu-build-amd64
#RUN OPENWEBUI_CPU_TARGET="cpu" sh gen_linux.sh
#FROM --platform=linux/amd64 cpu-builder-amd64 AS cpu_avx-build-amd64
#RUN OPENWEBUI_CPU_TARGET="cpu_avx" sh gen_linux.sh
#FROM --platform=linux/amd64 cpu-builder-amd64 AS cpu_avx2-build-amd64
#RUN OPENWEBUI_CPU_TARGET="cpu_avx2" sh gen_linux.sh
######## CUDA WebUI backend ########
Parametrize CUDA_VERSION in Dockerfile Standardized CUDA_VERSION as a global ARG to ensure consistency and facilitate version updates across the Dockerfile. This change allows the CUDA version to be defined once at the beginning and reused, reducing the chance of mismatched versions and easing maintenance when changing CUDA versions. It further streamlines the build process for potential multi-stage builds with varying CUDA dependencies. Refs #nvidia-update
2024-03-17 07:27:06 +01:00
FROM --platform=linux/amd64 nvidia/cuda:"$CUDA_VERSION"-devel-ubuntu22.04 AS cuda-build-amd64
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
# Set environment variables for NVIDIA Container Toolkit
Optimize Dockerfile for CUDA support Refactored the Dockerfile to better organize and streamline environment variable settings, emphasizing support for a CUDA-based WebUI backend while retaining the ability to build a CPU-only image. Consolidated ENV commands to reduce layers, improving build efficiency, and set a default PORT environment to enhance container usability. Enabled exposure of the backend service on port 8080 and leveraged combined RUN directives to minimize the image footprint. These changes facilitate a more robust deployment process, catering to both CPU and CUDA environments.
2024-03-17 06:55:37 +01:00
ENV LD_LIBRARY_PATH=/usr/local/nvidia/lib:/usr/local/nvidia/lib64 \
NVIDIA_DRIVER_CAPABILITIES=all \
NVIDIA_VISIBLE_DEVICES=all
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
Optimize Dockerfile for CUDA support Refactored the Dockerfile to better organize and streamline environment variable settings, emphasizing support for a CUDA-based WebUI backend while retaining the ability to build a CPU-only image. Consolidated ENV commands to reduce layers, improving build efficiency, and set a default PORT environment to enhance container usability. Enabled exposure of the backend service on port 8080 and leveraged combined RUN directives to minimize the image footprint. These changes facilitate a more robust deployment process, catering to both CPU and CUDA environments.
2024-03-17 06:55:37 +01:00
ENV ENV=prod \
PORT=8080
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
Optimize Dockerfile for CUDA support Refactored the Dockerfile to better organize and streamline environment variable settings, emphasizing support for a CUDA-based WebUI backend while retaining the ability to build a CPU-only image. Consolidated ENV commands to reduce layers, improving build efficiency, and set a default PORT environment to enhance container usability. Enabled exposure of the backend service on port 8080 and leveraged combined RUN directives to minimize the image footprint. These changes facilitate a more robust deployment process, catering to both CPU and CUDA environments.
2024-03-17 06:55:37 +01:00
## Base URL Config ##
ENV OLLAMA_BASE_URL="/ollama" \
OPENAI_API_BASE_URL=""
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
Optimize Dockerfile for CUDA support Refactored the Dockerfile to better organize and streamline environment variable settings, emphasizing support for a CUDA-based WebUI backend while retaining the ability to build a CPU-only image. Consolidated ENV commands to reduce layers, improving build efficiency, and set a default PORT environment to enhance container usability. Enabled exposure of the backend service on port 8080 and leveraged combined RUN directives to minimize the image footprint. These changes facilitate a more robust deployment process, catering to both CPU and CUDA environments.
2024-03-17 06:55:37 +01:00
## API Key and Security Config ##
ENV OPENAI_API_KEY="" \
WEBUI_SECRET_KEY="" \
SCARF_NO_ANALYTICS=true \
DO_NOT_TRACK=true
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
######## Preloaded models ########
# whisper TTS Settings
Optimize Dockerfile for CUDA support Refactored the Dockerfile to better organize and streamline environment variable settings, emphasizing support for a CUDA-based WebUI backend while retaining the ability to build a CPU-only image. Consolidated ENV commands to reduce layers, improving build efficiency, and set a default PORT environment to enhance container usability. Enabled exposure of the backend service on port 8080 and leveraged combined RUN directives to minimize the image footprint. These changes facilitate a more robust deployment process, catering to both CPU and CUDA environments.
2024-03-17 06:55:37 +01:00
ENV WHISPER_MODEL="base" \
WHISPER_MODEL_DIR="/app/backend/data/cache/whisper/models"
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
# RAG Embedding Model Settings
# any sentence transformer model; models to use can be found at https://huggingface.co/models?library=sentence-transformers
# Leaderboard: https://huggingface.co/spaces/mteb/leaderboard
# for better performance and multilangauge support use "intfloat/multilingual-e5-large" (~2.5GB) or "intfloat/multilingual-e5-base" (~1.5GB)
# IMPORTANT: If you change the default model (all-MiniLM-L6-v2) and vice versa, you aren't able to use RAG Chat with your previous documents loaded in the WebUI! You need to re-embed them.
Optimize Dockerfile for CUDA support Refactored the Dockerfile to better organize and streamline environment variable settings, emphasizing support for a CUDA-based WebUI backend while retaining the ability to build a CPU-only image. Consolidated ENV commands to reduce layers, improving build efficiency, and set a default PORT environment to enhance container usability. Enabled exposure of the backend service on port 8080 and leveraged combined RUN directives to minimize the image footprint. These changes facilitate a more robust deployment process, catering to both CPU and CUDA environments.
2024-03-17 06:55:37 +01:00
ENV RAG_EMBEDDING_MODEL="all-MiniLM-L6-v2" \
# device type for whisper tts and embedding models - "cpu" (default), "cuda" (NVIDIA GPU and CUDA required), or "mps" (apple silicon) - choosing this right can lead to better performance
RAG_EMBEDDING_MODEL_DEVICE_TYPE="cuda" \
RAG_EMBEDDING_MODEL_DIR="/app/backend/data/cache/embedding/models" \
SENTENCE_TRANSFORMERS_HOME=$RAG_EMBEDDING_MODEL_DIR
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
######## Preloaded models ########
WORKDIR /app/backend
# Install Python & dependencies in the container
RUN apt-get update && \
apt-get install -y --no-install-recommends python3.11 python3-pip ffmpeg libsm6 libxext6 pandoc netcat-openbsd && \
rm -rf /var/lib/apt/lists/*
COPY ./backend/requirements.txt ./requirements.txt
Optimize Dockerfile for CUDA support Refactored the Dockerfile to better organize and streamline environment variable settings, emphasizing support for a CUDA-based WebUI backend while retaining the ability to build a CPU-only image. Consolidated ENV commands to reduce layers, improving build efficiency, and set a default PORT environment to enhance container usability. Enabled exposure of the backend service on port 8080 and leveraged combined RUN directives to minimize the image footprint. These changes facilitate a more robust deployment process, catering to both CPU and CUDA environments.
2024-03-17 06:55:37 +01:00
RUN pip3 install torch torchvision torchaudio --no-cache-dir && \
pip3 install -r requirements.txt --no-cache-dir
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
# copy built frontend files
COPY --from=build /app/build /app/build
COPY --from=build /app/CHANGELOG.md /app/CHANGELOG.md
COPY --from=build /app/package.json /app/package.json
# copy backend files
COPY ./backend .
Optimize Dockerfile for CUDA support Refactored the Dockerfile to better organize and streamline environment variable settings, emphasizing support for a CUDA-based WebUI backend while retaining the ability to build a CPU-only image. Consolidated ENV commands to reduce layers, improving build efficiency, and set a default PORT environment to enhance container usability. Enabled exposure of the backend service on port 8080 and leveraged combined RUN directives to minimize the image footprint. These changes facilitate a more robust deployment process, catering to both CPU and CUDA environments.
2024-03-17 06:55:37 +01:00
EXPOSE 8080
Create Dockerfile-cuda +Dockerfile-cuda I created this file to help add CUDA support to open-webui for access to a GPU during embedding operations.
2024-03-17 00:26:21 +01:00
CMD [ "bash", "start.sh"]
Copy permalink