14 lines
697 B
Docker
14 lines
697 B
Docker
# apps/embedding-runtime/Dockerfile
|
|
# SmartMLOps Tier-2 embedding-runtime — OpenAI-compatible embeddings/rerank server.
|
|
# KServe owns lifecycle (ClusterServingRuntime container); we own only this serving code.
|
|
FROM python:3.11-slim
|
|
ENV PIP_DISABLE_PIP_VERSION_CHECK=1 \
|
|
HF_HUB_DISABLE_TELEMETRY=1 \
|
|
PYTHONUNBUFFERED=1
|
|
WORKDIR /app
|
|
COPY requirements.txt /app/requirements.txt
|
|
RUN --mount=type=cache,target=/root/.cache/pip pip install -r /app/requirements.txt
|
|
COPY server.py /app/server.py
|
|
EXPOSE 8080
|
|
# KServe mounts the model artifact at /mnt/models; server.py reads EMBEDDING_MODEL_DIR (default /mnt/models).
|
|
ENTRYPOINT ["uvicorn", "server:app", "--host", "0.0.0.0", "--port", "8080"]
|