Files
onyx/backend/Dockerfile.model_server
T

87 lines
3.5 KiB
Docker

FROM python:3.11.7-slim-bookworm
LABEL com.danswer.maintainer="founders@onyx.app"
LABEL com.danswer.description="This image is for the Onyx model server which runs all of the \
AI models for Onyx. This container and all the code is MIT Licensed and free for all to use. \
You can find it at https://hub.docker.com/r/onyx/onyx-model-server. For more details, \
visit https://github.com/onyx-dot-app/onyx."
# Default ONYX_VERSION, typically overriden during builds by GitHub Actions.
ARG ONYX_VERSION=0.0.0-dev
ENV ONYX_VERSION=${ONYX_VERSION} \
DANSWER_RUNNING_IN_DOCKER="true" \
HF_HOME=/app/.cache/huggingface
COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvx /bin/
# Create non-root user for security best practices
RUN mkdir -p /app && \
groupadd -g 1001 onyx && \
useradd -u 1001 -g onyx -m -s /bin/bash onyx && \
chown -R onyx:onyx /app && \
mkdir -p /var/log/onyx && \
chmod 755 /var/log/onyx && \
chown onyx:onyx /var/log/onyx
# --- add toolchain needed for Rust/Python builds (fastuuid) ---
ENV RUSTUP_HOME=/usr/local/rustup \
CARGO_HOME=/usr/local/cargo \
PATH=/usr/local/cargo/bin:$PATH
RUN set -eux; \
apt-get update && apt-get install -y --no-install-recommends \
build-essential \
pkg-config \
curl \
ca-certificates \
# Install latest stable Rust (supports Cargo.lock v4)
&& curl -sSf https://sh.rustup.rs | sh -s -- -y --profile minimal --default-toolchain stable \
&& rustc --version && cargo --version \
&& apt-get remove -y --allow-remove-essential perl-base \
&& apt-get autoremove -y \
&& rm -rf /var/lib/apt/lists/*
COPY ./requirements/model_server.txt /tmp/requirements.txt
RUN uv pip install --system --no-cache-dir --upgrade \
-r /tmp/requirements.txt && \
rm -rf ~/.cache/uv /tmp/*.txt
# Pre-downloading models for setups with limited egress
# Download tokenizers, distilbert for the Onyx model
# Download model weights
# Run Nomic to pull in the custom architecture and have it cached locally
RUN python -c "from transformers import AutoTokenizer; \
AutoTokenizer.from_pretrained('distilbert-base-uncased'); \
AutoTokenizer.from_pretrained('mixedbread-ai/mxbai-rerank-xsmall-v1'); \
from huggingface_hub import snapshot_download; \
snapshot_download(repo_id='onyx-dot-app/hybrid-intent-token-classifier'); \
snapshot_download(repo_id='onyx-dot-app/information-content-model'); \
snapshot_download('nomic-ai/nomic-embed-text-v1'); \
snapshot_download('mixedbread-ai/mxbai-rerank-xsmall-v1'); \
from sentence_transformers import SentenceTransformer; \
SentenceTransformer(model_name_or_path='nomic-ai/nomic-embed-text-v1', trust_remote_code=True);" && \
# In case the user has volumes mounted to /app/.cache/huggingface that they've downloaded while
# running Onyx, move the current contents of the cache folder to a temporary location to ensure
# it's preserved in order to combine with the user's cache contents
mv /app/.cache/huggingface /app/.cache/temp_huggingface && \
chown -R onyx:onyx /app
WORKDIR /app
# Utils used by model server
COPY ./onyx/utils/logger.py /app/onyx/utils/logger.py
COPY ./onyx/utils/middleware.py /app/onyx/utils/middleware.py
# Place to fetch version information
COPY ./onyx/__init__.py /app/onyx/__init__.py
# Shared between Onyx Backend and Model Server
COPY ./shared_configs /app/shared_configs
# Model Server main code
COPY ./model_server /app/model_server
ENV PYTHONPATH=/app
CMD ["uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "9000"]