Dockerfile 3.54 KB
Newer Older
Timothy J. Baek's avatar
Timothy J. Baek committed
1
2
# syntax=docker/dockerfile:1

Jannik Streidl's avatar
Jannik Streidl committed
3
######## WebUI frontend ########
Jannik S's avatar
Jannik S committed
4
FROM node:21-alpine3.19 as build
5

Timothy J. Baek's avatar
Timothy J. Baek committed
6
7
WORKDIR /app

8
9
10
11
#RUN apt-get update \ 
#    && apt-get install -y --no-install-recommends wget \ 
#    # cleanup
#    && rm -rf /var/lib/apt/lists/*
Jannik Streidl's avatar
Jannik Streidl committed
12

Timothy J. Baek's avatar
Timothy J. Baek committed
13
# wget embedding model weight from alpine (does not exist from slim-buster)
Jannik Streidl's avatar
Jannik Streidl committed
14
15
#RUN wget "https://chroma-onnx-models.s3.amazonaws.com/all-MiniLM-L6-v2/onnx.tar.gz" -O - | \
#    tar -xzf - -C /app
Timothy J. Baek's avatar
Timothy J. Baek committed
16

Xiaodong Ye's avatar
Xiaodong Ye committed
17
COPY package.json package-lock.json ./
18
RUN npm ci
19

20
21
COPY . .
RUN npm run build
Timothy J. Baek's avatar
Timothy J. Baek committed
22

Jannik Streidl's avatar
Jannik Streidl committed
23
######## WebUI backend ########
Timothy J. Baek's avatar
Timothy J. Baek committed
24
FROM python:3.11-slim-bookworm as base
Timothy J. Baek's avatar
Timothy J. Baek committed
25

Jannik Streidl's avatar
Jannik Streidl committed
26
27
28
## Basis ##
ENV ENV=prod \
    PORT=8080
Timothy J. Baek's avatar
Timothy J. Baek committed
29

Jannik Streidl's avatar
Jannik Streidl committed
30
31
32
## Basis URL Config ##
ENV OLLAMA_BASE_URL="/ollama" \
    OPENAI_API_BASE_URL=""
Timothy J. Baek's avatar
Timothy J. Baek committed
33

Jannik Streidl's avatar
Jannik Streidl committed
34
35
36
37
38
## API Key and Security Config ##
ENV OPENAI_API_KEY="" \
    WEBUI_SECRET_KEY="" \
    SCARF_NO_ANALYTICS=true \
    DO_NOT_TRACK=true
Timothy J. Baek's avatar
Timothy J. Baek committed
39

Jannik Streidl's avatar
Jannik Streidl committed
40
41
42
43
#### Preloaded models ##########################################################
## whisper TTS Settings ##
ENV WHISPER_MODEL="base" \
    WHISPER_MODEL_DIR="/app/backend/data/cache/whisper/models"
44

Jannik Streidl's avatar
Jannik Streidl committed
45
## RAG Embedding Model Settings ##
46
47
# 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 
48
# for better persormance and multilangauge support use "intfloat/multilingual-e5-large" (~2.5GB) or "intfloat/multilingual-e5-base" (~1.5GB)
49
# 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.
Jannik Streidl's avatar
Jannik Streidl committed
50
51
52
53
54
55
ENV RAG_EMBEDDING_MODEL="all-MiniLM-L6-v2" \
    # 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
    RAG_EMBEDDING_MODEL_DEVICE_TYPE="cpu" \
    RAG_EMBEDDING_MODEL_DIR="/app/backend/data/cache/embedding/models" \
    SENTENCE_TRANSFORMERS_HOME="/app/backend/data/cache/embedding/models"
#### Preloaded models ##########################################################
56

Timothy J. Baek's avatar
Timothy J. Baek committed
57
58
WORKDIR /app/backend

59
# install python dependencies
Timothy J. Baek's avatar
Timothy J. Baek committed
60
COPY ./backend/requirements.txt ./requirements.txt
Timothy J. Baek's avatar
Timothy J. Baek committed
61

Jannik Streidl's avatar
Jannik Streidl committed
62
63
RUN pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu --no-cache-dir \
    && pip3 install -r requirements.txt --no-cache-dir
64

Jannik Streidl's avatar
Jannik Streidl committed
65
#  install required packages
Timothy J. Baek's avatar
Timothy J. Baek committed
66
RUN apt-get update \
Jannik Streidl's avatar
Jannik Streidl committed
67
68
69
70
71
    # 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
Timothy J. Baek's avatar
Timothy J. Baek committed
72
73
    && rm -rf /var/lib/apt/lists/*

74
# preload embedding model
75
RUN python -c "import os; from chromadb.utils import embedding_functions; sentence_transformer_ef = embedding_functions.SentenceTransformerEmbeddingFunction(model_name=os.environ['RAG_EMBEDDING_MODEL'], device=os.environ['RAG_EMBEDDING_MODEL_DEVICE_TYPE'])"
76
# preload tts model
77
RUN python -c "import os; from faster_whisper import WhisperModel; WhisperModel(os.environ['WHISPER_MODEL'], device='auto', compute_type='int8', download_root=os.environ['WHISPER_MODEL_DIR'])"
78

79
# copy embedding weight from build
Jannik Streidl's avatar
Jannik Streidl committed
80
81
# 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
82
83
84

# copy built frontend files
COPY --from=build /app/build /app/build
85
86
COPY --from=build /app/CHANGELOG.md /app/CHANGELOG.md
COPY --from=build /app/package.json /app/package.json
87
88

# copy backend files
Timothy J. Baek's avatar
Timothy J. Baek committed
89
90
COPY ./backend .

Jannik S's avatar
Jannik S committed
91
92
EXPOSE 8080

pandego's avatar
pandego committed
93
CMD [ "bash", "start.sh"]