Skip to content

add reflex_episode_chache_bytes_total gauge + helper and EpisodeCache… #46

add reflex_episode_chache_bytes_total gauge + helper and EpisodeCache…

add reflex_episode_chache_bytes_total gauge + helper and EpisodeCache… #46

Workflow file for this run

name: Pytest CI
on:
push:
branches: [main]
pull_request:
branches: [main]
jobs:
test:
name: pytest (ubuntu py${{ matrix.python-version }})
runs-on: ubuntu-latest
strategy:
fail-fast: false
matrix:
# Linux-only: every reflex tester runs Linux. Macs use the [onnx] extra
# path locally but CI doesn't add coverage there. Saves ~50% runner time.
python-version: ["3.10", "3.12"]
steps:
- uses: actions/checkout@v4
- name: Set up Python ${{ matrix.python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ matrix.python-version }}
# Cache pip downloads keyed on pyproject.toml. Cuts ~3 min off
# subsequent runs since torch/onnxruntime/etc don't re-download.
cache: 'pip'
cache-dependency-path: 'pyproject.toml'
- name: Install dependencies
run: |
python -m pip install --upgrade pip
# [serve] pulls fastapi/uvicorn/Pillow which the import + CLI tests need.
# We deliberately skip [gpu] since GitHub-hosted runners have no GPU
# and tensorrt is a 2 GB install for tests that would all skip via
# @pytest.mark.skipif anyway.
pip install -e ".[dev,onnx,serve]"
- name: Run tests
run: |
# Hardware-specific tests skip automatically via @pytest.mark.skipif.
# CI coverage is the import + CLI + lazy-load + chat backend surface.
pytest tests/ -v