2023-10-08 22:38:42 +00:00
# syntax=docker/dockerfile:1
2024-03-20 07:44:09 +00:00
# Initialize device type args
2024-03-22 08:55:46 +00:00
# use build args in the docker build commmand with --build-arg="BUILDARG=true"
2024-03-20 07:44:09 +00:00
ARG USE_CUDA = false
ARG USE_MPS = false
2024-03-22 08:31:35 +00:00
ARG INCLUDE_OLLAMA = false
2023-10-08 22:38:42 +00:00
2024-03-14 10:18:27 +00:00
######## WebUI frontend ########
2024-03-16 11:43:48 +00:00
FROM node:21-alpine3.19 as build
2023-10-21 23:14:12 +00:00
2023-11-15 00:28:51 +00:00
WORKDIR /app
2024-03-14 10:33:54 +00:00
#RUN apt-get update \
# && apt-get install -y --no-install-recommends wget \
# # cleanup
# && rm -rf /var/lib/apt/lists/*
2024-03-14 10:18:27 +00:00
2024-01-08 04:50:09 +00:00
# wget embedding model weight from alpine (does not exist from slim-buster)
2024-03-14 10:18:27 +00:00
#RUN wget "https://chroma-onnx-models.s3.amazonaws.com/all-MiniLM-L6-v2/onnx.tar.gz" -O - | \
# tar -xzf - -C /app
2024-01-08 04:50:09 +00:00
2024-01-25 10:08:35 +00:00
COPY package.json package-lock.json ./
2024-01-05 05:15:13 +00:00
RUN npm ci
2024-01-05 04:32:51 +00:00
2024-01-05 05:15:13 +00:00
COPY . .
RUN npm run build
2023-10-08 22:38:42 +00:00
2024-03-14 10:18:27 +00:00
######## WebUI backend ########
2024-01-07 16:28:35 +00:00
FROM python:3.11-slim-bookworm as base
2023-11-15 00:28:51 +00:00
2024-03-20 07:44:09 +00:00
# Use args
ARG USE_CUDA
ARG USE_MPS
2024-03-22 08:31:35 +00:00
ARG INCLUDE_OLLAMA
2024-03-20 07:44:09 +00:00
2024-03-14 10:18:27 +00:00
## Basis ##
ENV ENV = prod \
2024-03-22 08:31:35 +00:00
PORT = 8080 \
2024-03-22 11:48:48 +00:00
# pass build args to the build
INCLUDE_OLLAMA_DOCKER = ${ INCLUDE_OLLAMA } \
USE_MPS_DOCKER = ${ USE_MPS } \
USE_CUDA_DOCKER = ${ USE_CUDA }
2024-01-05 02:55:15 +00:00
2024-03-14 10:18:27 +00:00
## Basis URL Config ##
ENV OLLAMA_BASE_URL = "/ollama" \
OPENAI_API_BASE_URL = ""
2024-01-05 02:55:15 +00:00
2024-03-14 10:18:27 +00:00
## API Key and Security Config ##
ENV OPENAI_API_KEY = "" \
WEBUI_SECRET_KEY = "" \
SCARF_NO_ANALYTICS = true \
DO_NOT_TRACK = true
2023-11-15 00:28:51 +00:00
2024-03-18 16:08:34 +00:00
#### Preloaded models #########################################################
2024-03-14 10:18:27 +00:00
## whisper TTS Settings ##
ENV WHISPER_MODEL = "base" \
WHISPER_MODEL_DIR = "/app/backend/data/cache/whisper/models"
2024-02-10 02:11:01 +00:00
2024-03-14 10:18:27 +00:00
## RAG Embedding Model Settings ##
2024-02-17 18:38:29 +00:00
# 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
2024-03-20 22:28:57 +00:00
# for better performance and multilangauge support use "intfloat/multilingual-e5-large" (~2.5GB) or "intfloat/multilingual-e5-base" (~1.5GB)
2024-02-17 18:38:29 +00:00
# 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.
2024-03-14 10:18:27 +00:00
ENV RAG_EMBEDDING_MODEL = "all-MiniLM-L6-v2" \
RAG_EMBEDDING_MODEL_DIR = "/app/backend/data/cache/embedding/models" \
2024-03-18 16:08:34 +00:00
SENTENCE_TRANSFORMERS_HOME = "/app/backend/data/cache/embedding/models" \
# device type for whisper tts and embbeding models - "cpu" (default) or "mps" (apple silicon) - choosing this right can lead to better performance
# Important:
# If you want to use CUDA you need to install the nvidia-container-toolkit (https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html)
# you can set this to "cuda" but its recomended to use --build-arg CUDA_ENABLED=true flag when building the image
2024-03-22 11:48:48 +00:00
# RAG_EMBEDDING_MODEL_DEVICE_TYPE="cpu" \
2024-03-20 07:44:09 +00:00
DEVICE_COMPUTE_TYPE = "int8"
2024-03-18 16:08:34 +00:00
# device type for whisper tts and embbeding models - "cpu" (default), "cuda" (nvidia gpu and CUDA required) or "mps" (apple silicon) - choosing this right can lead to better performance
2024-03-14 10:18:27 +00:00
#### Preloaded models ##########################################################
2024-02-17 18:38:29 +00:00
2023-11-15 00:28:51 +00:00
WORKDIR /app/backend
2024-01-23 12:48:27 +00:00
# install python dependencies
2023-11-15 00:28:51 +00:00
COPY ./backend/requirements.txt ./requirements.txt
2024-01-08 05:22:37 +00:00
2024-03-20 07:44:09 +00:00
RUN if [ " $USE_CUDA " = "true" ] ; then \
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu117 --no-cache-dir && \
pip3 install -r requirements.txt --no-cache-dir; \
elif [ " $USE_MPS " = "true" ] ; then \
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu --no-cache-dir && \
pip3 install -r requirements.txt --no-cache-dir && \
python -c "import os; from faster_whisper import WhisperModel; WhisperModel(os.environ['WHISPER_MODEL'], device='cpu', compute_type='int8', download_root=os.environ['WHISPER_MODEL_DIR'])" && \
2024-03-22 11:48:48 +00:00
python -c "import os; from chromadb.utils import embedding_functions; sentence_transformer_ef = embedding_functions.SentenceTransformerEmbeddingFunction(model_name=os.environ['RAG_EMBEDDING_MODEL'], device='mps')" ; \
2024-03-18 16:08:34 +00:00
else \
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu --no-cache-dir && \
2024-03-20 07:44:09 +00:00
pip3 install -r requirements.txt --no-cache-dir && \
python -c "import os; from faster_whisper import WhisperModel; WhisperModel(os.environ['WHISPER_MODEL'], device='cpu', compute_type='int8', download_root=os.environ['WHISPER_MODEL_DIR'])" && \
2024-03-22 11:48:48 +00:00
python -c "import os; from chromadb.utils import embedding_functions; sentence_transformer_ef = embedding_functions.SentenceTransformerEmbeddingFunction(model_name=os.environ['RAG_EMBEDDING_MODEL'], device='cpu')" ; \
2024-03-18 16:08:34 +00:00
fi
2024-03-22 08:31:35 +00:00
RUN if [ " $INCLUDE_OLLAMA " = "true" ] ; then \
apt-get update && \
# Install pandoc and netcat
apt-get install -y --no-install-recommends pandoc netcat-openbsd && \
# for RAG OCR
apt-get install -y --no-install-recommends ffmpeg libsm6 libxext6 && \
# install helper tools
apt-get install -y --no-install-recommends curl && \
# install ollama
curl -fsSL https://ollama.com/install.sh | sh && \
# cleanup
rm -rf /var/lib/apt/lists/*; \
else \
apt-get update && \
# Install pandoc and netcat
apt-get install -y --no-install-recommends pandoc netcat-openbsd && \
# for RAG OCR
apt-get install -y --no-install-recommends ffmpeg libsm6 libxext6 && \
# cleanup
rm -rf /var/lib/apt/lists/*; \
fi
2024-01-23 07:11:50 +00:00
2024-03-18 16:08:34 +00:00
2024-02-13 14:11:53 +00:00
2024-01-23 12:48:27 +00:00
# copy embedding weight from build
2024-03-14 10:18:27 +00:00
# RUN mkdir -p /root/.cache/chroma/onnx_models/all-MiniLM-L6-v2
# COPY --from=build /app/onnx /root/.cache/chroma/onnx_models/all-MiniLM-L6-v2/onnx
2024-01-23 12:48:27 +00:00
# copy built frontend files
COPY --from= build /app/build /app/build
2024-02-23 08:54:22 +00:00
COPY --from= build /app/CHANGELOG.md /app/CHANGELOG.md
COPY --from= build /app/package.json /app/package.json
2024-01-23 12:48:27 +00:00
# copy backend files
2023-11-15 00:28:51 +00:00
COPY ./backend .
2024-03-16 19:11:09 +00:00
EXPOSE 8080
2024-03-20 07:44:09 +00:00
CMD [ "bash" , "start.sh" ]