sie-server 0.1.10__tar.gz

sie_server-0.1.10/.gitignore

@@ -0,0 +1,259 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[codz]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py.cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# UV
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#uv.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+#poetry.toml
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#   pdm recommends including project-wide configuration in pdm.toml, but excluding .pdm-python.
+#   https://pdm-project.org/en/latest/usage/project/#working-with-version-control
+#pdm.lock
+#pdm.toml
+.pdm-python
+.pdm-build/
+
+# pixi
+#   Similar to Pipfile.lock, it is generally recommended to include pixi.lock in version control.
+#pixi.lock
+#   Pixi creates a virtual environment in the .pixi directory, just like venv module creates one
+#   in the .venv directory. It is recommended not to include this directory in version control.
+.pixi
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
+
+# Abstra
+# Abstra is an AI-powered process automation framework.
+# Ignore directories containing user credentials, local state, and settings.
+# Learn more at https://abstra.io/docs
+.abstra/
+
+# Visual Studio Code
+#  Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore
+#  that can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
+#  and can be added to the global gitignore or merged into this file. However, if you prefer,
+#  you could uncomment the following to ignore the entire vscode folder
+# .vscode/
+
+# Ruff stuff:
+.ruff_cache/
+
+# PyPI configuration file
+.pypirc
+
+# Cursor
+#  Cursor is an AI-powered code editor. `.cursorignore` specifies files/directories to
+#  exclude from AI features like autocomplete and code analysis. Recommended for sensitive data
+#  refer to https://docs.cursor.com/context/ignore-files
+.cursorignore
+.cursorindexingignore
+
+# vcscode
+.vscode
+
+# Marimo
+marimo/_static/
+marimo/_lsp/
+__marimo__/
+
+# SIE specific
+# Model weights cache
+.cache/
+*.safetensors
+*.bin
+
+# Secrets (never commit)
+*.pem
+*.key
+credentials.json
+*-key.json
+
+# Terraform
+# Local .terraform directories (cached providers/modules)
+**/.terraform/*
+.terraform.lock*
+# State files (contain sensitive data)
+*.tfstate
+*.tfstate.*
+# Crash log files
+crash.log
+crash.*.log
+# Override files (local developer overrides)
+override.tf
+override.tf.json
+*_override.tf
+*_override.tf.json
+# tfvars files may contain secrets
+*.tfvars
+*.tfvars.json
+# Keep .terraform.lock.hcl for reproducibility (provider versions)
+
+# Node.js
+node_modules/
+
+# OS
+.DS_Store
+Thumbs.db
+
+# VIM
+*.swp
+
+# kilocode
+.kilo/
+
+# Worktree metadata
+.base-branch
+
+# Temporary files
+tmp/
+.tmp/
+.local/
+
+.requirements-modal.txt

sie_server-0.1.10/CONTRIBUTING.md

@@ -0,0 +1,34 @@
+# Contributing to SIE Server
+
+## Development Setup
+
+```bash
+# Clone the repo
+git clone https://github.com/superlinked/sie.git
+cd sie
+
+# Install mise and dependencies
+mise trust && mise install
+```
+
+## Running Tests
+
+```bash
+mise run test packages/sie_server
+```
+
+## Code Style
+
+- Format with `mise run lint -f`
+- Type check with `mise run typecheck`
+
+## Pull Requests
+
+1. Fork the repo
+2. Create a feature branch
+3. Make changes with tests
+4. Submit a PR
+
+## License
+
+By contributing, you agree that your contributions will be licensed under Apache 2.0.

sie_server-0.1.10/Dockerfile.cpu

@@ -0,0 +1,187 @@
+# syntax=docker/dockerfile:1.6
+# SIE Server - CPU-only Image (amd64 + arm64)
+# Build from repo root:
+#   docker build -f packages/sie_server/Dockerfile.cpu -t sie-server:cpu-default .
+#   docker build -f packages/sie_server/Dockerfile.cpu --build-arg BUNDLE=sglang -t sie-server:cpu-sglang .
+#   docker buildx build --platform linux/amd64,linux/arm64 -f packages/sie_server/Dockerfile.cpu -t sie-server:cpu .
+#
+# For faster builds with cache mounts (requires BuildKit):
+#   DOCKER_BUILDKIT=1 docker build -f packages/sie_server/Dockerfile.cpu -t sie-server:cpu-default .
+#
+# Layer caching strategy:
+#   - Stage 1 (deps): Install all pip dependencies from pyproject.toml (cached unless deps change)
+#   - Stage 2 (builder): Copy source code and do editable install (quick, uses cached deps)
+#   This ensures code-only changes rebuild only Stage 2, not the slow pip install step.
+
+ARG BUNDLE=default
+
+# =============================================================================
+# Stage 1: Dependencies (changes rarely - cached for code-only changes)
+# =============================================================================
+FROM ubuntu:22.04 AS deps
+
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Install Python 3.12 and build tools
+RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
+    apt-get update && apt-get install -y --no-install-recommends \
+    software-properties-common \
+    gnupg \
+    ca-certificates \
+    && add-apt-repository -y ppa:deadsnakes/ppa \
+    && apt-get update && apt-get install -y --no-install-recommends \
+    python3.12 \
+    python3.12-venv \
+    python3.12-dev \
+    curl \
+    git \
+    build-essential \
+    zlib1g-dev \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install uv - pin to specific version for reproducibility
+COPY --from=ghcr.io/astral-sh/uv:0.5 /uv /uvx /bin/
+
+WORKDIR /app
+
+# UV environment variables for optimal builds
+ENV UV_NO_CACHE=1 \
+    UV_COMPILE_BYTECODE=1 \
+    UV_LINK_MODE=copy
+
+# Copy ONLY dependency specifications first (for layer caching)
+COPY packages/sie_sdk/pyproject.toml /tmp/sie_sdk/pyproject.toml
+COPY packages/sie_server/pyproject.toml ./pyproject.toml
+
+# Create minimal src directories so pip install -e works for dep resolution
+RUN mkdir -p /tmp/sie_sdk/src/sie_sdk src/sie_server \
+    && touch /tmp/sie_sdk/src/sie_sdk/__init__.py \
+    && touch src/sie_server/__init__.py
+
+# Create venv and install dependencies only (without actual source code)
+# This layer is cached as long as pyproject.toml files don't change
+RUN --mount=type=cache,target=/root/.cache/pip \
+    python3.12 -m venv .venv \
+    && .venv/bin/pip install --upgrade pip \
+    && .venv/bin/pip install \
+        --index-url https://download.pytorch.org/whl/cpu \
+        --extra-index-url https://pypi.org/simple \
+        -e "/tmp/sie_sdk[storage]" \
+        -e ".[gpu-metrics]"
+
+# =============================================================================
+# Stage 2: Source code (changes often - quick rebuild)
+# =============================================================================
+FROM deps AS base
+
+# Now copy the actual source code (this layer rebuilds on code changes)
+COPY packages/sie_sdk/src /tmp/sie_sdk/src
+COPY packages/sie_server/src src/
+
+# Re-install editable packages with actual source (quick, deps already installed)
+RUN --mount=type=cache,target=/root/.cache/pip \
+    .venv/bin/pip install --no-deps \
+        -e /tmp/sie_sdk \
+        -e .
+
+# =============================================================================
+# Stage 2: Builder - bundle-specific deps
+# =============================================================================
+FROM base AS builder
+
+ARG BUNDLE
+
+# Resolve bundle-specific dependencies
+COPY packages/sie_server/bundles bundles/
+COPY packages/sie_server/models models/
+
+# Use --cpu flag to exclude CUDA-only dependencies (flash-attn, xformers)
+RUN .venv/bin/python -m sie_server.cli resolve-deps \
+    --bundle "${BUNDLE}" \
+    --models-dir models \
+    --cpu \
+    > /tmp/bundle-requirements.txt \
+    && echo "Bundle ${BUNDLE} requirements:" \
+    && cat /tmp/bundle-requirements.txt
+
+# Install bundle-specific deps
+RUN --mount=type=cache,target=/root/.cache/pip \
+    if [ -s /tmp/bundle-requirements.txt ]; then \
+        .venv/bin/pip install \
+            --index-url https://download.pytorch.org/whl/cpu \
+            --extra-index-url https://pypi.org/simple \
+            -r /tmp/bundle-requirements.txt; \
+    fi
+
+# Clean up to reduce image size
+RUN find .venv -type d -name "__pycache__" -exec rm -rf {} + 2>/dev/null || true \
+    && find .venv -type f -name "*.pyc" -delete 2>/dev/null || true \
+    && find .venv -type d -name "tests" -exec rm -rf {} + 2>/dev/null || true \
+    && find .venv -type d -name "test" -exec rm -rf {} + 2>/dev/null || true \
+    && rm -rf .venv/share/jupyter 2>/dev/null || true
+
+# =============================================================================
+# Stage 2: Runtime
+# =============================================================================
+FROM ubuntu:22.04 AS runtime
+
+ARG BUNDLE
+
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Install Python 3.12 runtime
+RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
+    apt-get update && apt-get install -y --no-install-recommends \
+    software-properties-common \
+    gnupg \
+    ca-certificates \
+    && add-apt-repository -y ppa:deadsnakes/ppa \
+    && apt-get update && apt-get install -y --no-install-recommends \
+    python3.12 \
+    python3.12-venv \
+    libjpeg8 \
+    libpng16-16 \
+    && rm -rf /var/lib/apt/lists/*
+
+# Create non-root user
+RUN groupadd -g 1000 sie && useradd -u 1000 -g sie -m sie
+
+WORKDIR /app
+
+# Copy virtual environment from builder (with correct ownership to avoid chown layer)
+COPY --from=builder --chown=sie:sie /app/.venv /app/.venv
+
+# Copy source code (uv sync creates editable install with .pth pointing to /app/src)
+COPY --from=builder --chown=sie:sie /app/src /app/src
+
+# Copy sie-sdk source (installed as editable in /tmp)
+COPY --from=builder --chown=sie:sie /tmp/sie_sdk/src /tmp/sie_sdk/src
+
+# Copy model configs and bundles (needed for runtime model discovery)
+COPY --from=builder --chown=sie:sie /app/models /app/models
+COPY --from=builder --chown=sie:sie /app/bundles /app/bundles
+
+# Create directories for runtime
+RUN mkdir -p /app/evals /app/.cache/huggingface && chown -R sie:sie /app/evals /app/.cache
+
+# Labels
+LABEL org.opencontainers.image.title="SIE Server" \
+      org.opencontainers.image.description="Search Inference Engine - ${BUNDLE} bundle (CPU)" \
+      sie.bundle="${BUNDLE}"
+
+ENV PATH="/app/.venv/bin:$PATH" \
+    PYTHONUNBUFFERED=1 \
+    PYTHONDONTWRITEBYTECODE=1 \
+    HF_HOME=/app/.cache/huggingface \
+    OMP_NUM_THREADS=4 \
+    MKL_NUM_THREADS=4 \
+    SIE_BUNDLE="${BUNDLE}"
+
+USER sie
+EXPOSE 8080
+
+HEALTHCHECK --interval=30s --timeout=10s --start-period=60s --retries=3 \
+    CMD python -c "import urllib.request; urllib.request.urlopen('http://localhost:8080/healthz')" || exit 1
+
+ENTRYPOINT ["python", "-m", "sie_server.cli"]
+CMD ["serve", "--host", "0.0.0.0", "--port", "8080", "--models-dir", "/app/models", "--device", "cpu"]

sie_server-0.1.10/Dockerfile.cuda11

@@ -0,0 +1,165 @@
+# syntax=docker/dockerfile:1.6
+# SIE Server - CUDA 11.8 Image (for older drivers 470+)
+# Build from repo root:
+#   docker build -f packages/sie_server/Dockerfile.cuda11 -t sie-server:cuda11-default .
+#   docker build -f packages/sie_server/Dockerfile.cuda11 --build-arg BUNDLE=sglang -t sie-server:cuda11-sglang .
+#
+# For faster builds with cache mounts (requires BuildKit):
+#   DOCKER_BUILDKIT=1 docker build -f packages/sie_server/Dockerfile.cuda11 -t sie-server:cuda11-default .
+
+ARG BUNDLE=default
+
+# =============================================================================
+# Stage 1: Base deps (shared across bundles)
+# =============================================================================
+FROM nvidia/cuda:11.8.0-cudnn8-devel-ubuntu22.04 AS base
+
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Install Python 3.12 and build tools
+RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
+    apt-get update && apt-get install -y --no-install-recommends \
+    software-properties-common \
+    gnupg \
+    ca-certificates \
+    && add-apt-repository -y ppa:deadsnakes/ppa \
+    && apt-get update && apt-get install -y --no-install-recommends \
+    python3.12 \
+    python3.12-venv \
+    python3.12-dev \
+    curl \
+    git \
+    zlib1g-dev \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install uv - pin to specific version for reproducibility
+COPY --from=ghcr.io/astral-sh/uv:0.5 /uv /uvx /bin/
+
+WORKDIR /app
+
+# UV environment variables for optimal builds
+ENV UV_NO_CACHE=1 \
+    UV_COMPILE_BYTECODE=1 \
+    UV_LINK_MODE=copy
+
+# Copy sie-sdk (dependency of sie-server)
+COPY packages/sie_sdk /tmp/sie_sdk
+
+# Copy sie-server files (delay bundles/models to keep deps cached)
+COPY packages/sie_server/pyproject.toml ./
+COPY packages/sie_server/src src/
+
+# Create venv and install everything with pip (simpler, no workspace issues)
+# Note: flash-attn prebuilt wheels not available for cu118, uses SDPA fallback
+RUN --mount=type=cache,target=/root/.cache/pip \
+    python3.12 -m venv .venv \
+    && .venv/bin/pip install --upgrade pip \
+    && .venv/bin/pip install \
+        --index-url https://download.pytorch.org/whl/cu118 \
+        --extra-index-url https://pypi.org/simple \
+        -e "/tmp/sie_sdk[storage]" \
+        -e ".[gpu-metrics]"
+
+# =============================================================================
+# Stage 2: Builder - bundle-specific deps
+# =============================================================================
+FROM base AS builder
+
+ARG BUNDLE
+
+# Resolve bundle-specific dependencies
+COPY packages/sie_server/bundles bundles/
+COPY packages/sie_server/models models/
+
+RUN .venv/bin/python -m sie_server.cli resolve-deps \
+    --bundle "${BUNDLE}" \
+    --models-dir models \
+    > /tmp/bundle-requirements.txt \
+    && echo "Bundle ${BUNDLE} requirements:" \
+    && cat /tmp/bundle-requirements.txt
+
+# Install bundle-specific deps
+RUN --mount=type=cache,target=/root/.cache/pip \
+    if [ -s /tmp/bundle-requirements.txt ]; then \
+        .venv/bin/pip install \
+            --index-url https://download.pytorch.org/whl/cu118 \
+            --extra-index-url https://pypi.org/simple \
+            -r /tmp/bundle-requirements.txt; \
+    fi
+
+# Clean up to reduce image size
+RUN find .venv -type d -name "__pycache__" -exec rm -rf {} + 2>/dev/null || true \
+    && find .venv -type f -name "*.pyc" -delete 2>/dev/null || true \
+    && find .venv -type d -name "tests" -exec rm -rf {} + 2>/dev/null || true \
+    && find .venv -type d -name "test" -exec rm -rf {} + 2>/dev/null || true \
+    && rm -rf .venv/share/jupyter 2>/dev/null || true
+
+# =============================================================================
+# Stage 2: Runtime
+# =============================================================================
+# Use base image - PyTorch bundles CUDA libs
+FROM nvidia/cuda:11.8.0-base-ubuntu22.04 AS runtime
+
+ARG BUNDLE
+
+ENV DEBIAN_FRONTEND=noninteractive
+
+# Install Python 3.12 runtime + gcc (needed for triton kernel compilation)
+# libnuma1 is required by sgl_kernel (SGLang bundle) - without it, imports fail with misleading SM arch errors
+RUN --mount=type=cache,target=/var/cache/apt,sharing=locked \
+    apt-get update && apt-get install -y --no-install-recommends \
+    software-properties-common \
+    gnupg \
+    ca-certificates \
+    && add-apt-repository -y ppa:deadsnakes/ppa \
+    && apt-get update && apt-get install -y --no-install-recommends \
+    python3.12 \
+    python3.12-venv \
+    gcc \
+    libc6-dev \
+    libnuma1 \
+    && rm -rf /var/lib/apt/lists/*
+
+# Create non-root user
+RUN groupadd -g 1000 sie && useradd -u 1000 -g sie -m sie
+
+WORKDIR /app
+
+# Copy virtual environment from builder (with correct ownership to avoid chown layer)
+COPY --from=builder --chown=sie:sie /app/.venv /app/.venv
+
+# Copy source code (uv sync creates editable install with .pth pointing to /app/src)
+COPY --from=builder --chown=sie:sie /app/src /app/src
+
+# Copy sie-sdk source (installed as editable in /tmp)
+COPY --from=builder --chown=sie:sie /tmp/sie_sdk/src /tmp/sie_sdk/src
+
+# Copy model configs and bundles (needed for runtime model discovery)
+COPY --from=builder --chown=sie:sie /app/models /app/models
+COPY --from=builder --chown=sie:sie /app/bundles /app/bundles
+
+# Create directories for runtime
+RUN mkdir -p /app/evals /app/.cache/huggingface && chown -R sie:sie /app/evals /app/.cache
+
+# Labels
+LABEL org.opencontainers.image.title="SIE Server" \
+      org.opencontainers.image.description="Search Inference Engine - ${BUNDLE} bundle (CUDA 11.8)" \
+      sie.bundle="${BUNDLE}"
+
+ENV PATH="/app/.venv/bin:$PATH" \
+    PYTHONUNBUFFERED=1 \
+    PYTHONDONTWRITEBYTECODE=1 \
+    HF_HOME=/app/.cache/huggingface \
+    NVIDIA_VISIBLE_DEVICES=all \
+    NVIDIA_DRIVER_CAPABILITIES=compute,utility \
+    SIE_BUNDLE="${BUNDLE}" \
+    SIE_ATTENTION_BACKEND=sdpa
+
+USER sie
+EXPOSE 8080
+
+HEALTHCHECK --interval=30s --timeout=10s --start-period=60s --retries=3 \
+    CMD python -c "import urllib.request; urllib.request.urlopen('http://localhost:8080/healthz')" || exit 1
+
+ENTRYPOINT ["python", "-m", "sie_server.cli"]
+CMD ["serve", "--host", "0.0.0.0", "--port", "8080", "--models-dir", "/app/models"]