GPU CI

.github/packaging/vllm_reqs.txt

@@ -0,0 +1,147 @@
+# These requirements were generated by running steps 1-3 of scripts/build_wheels.shell
+# then running pip freeze and manually removing the vllm dependency.
+# The intention of this file is to use these known requirements for a fixed
+# vLLM build to supplement a vLLM install from download.pytorch.org without
+# resorting to --extra-index-url https://download.pytorch.org/whl/nightly to find
+# vLLM dependencies (as this results in a ResolutionTooDeep error from pip).
+# See the file .github/workflows/gpu_test.yaml for an E2E forge installation using this approach.
+# TODO: this should be done way less hackily
+aiohappyeyeballs==2.6.1
+aiohttp==3.13.0
+aiosignal==1.4.0
+annotated-types==0.7.0
+anyio==4.11.0
+astor==0.8.1
+async-timeout==5.0.1
+attrs==25.4.0
+blake3==1.0.7
+cachetools==6.2.0
+cbor2==5.7.0
+certifi==2025.10.5
+cffi==2.0.0
+charset-normalizer==3.4.3
+click==8.3.0
+cloudpickle==3.1.1
+cmake==4.1.0
+compressed-tensors==0.10.2
+cupy-cuda12x==13.6.0
+depyf==0.19.0
+dill==0.4.0
+diskcache==5.6.3
+distro==1.9.0
+dnspython==2.8.0
+einops==0.8.1
+email-validator==2.3.0
+exceptiongroup==1.3.0
+fastapi==0.118.3
+fastapi-cli==0.0.13
+fastapi-cloud-cli==0.3.1
+fastrlock==0.8.3
+filelock==3.19.1
+frozenlist==1.8.0
+fsspec==2025.9.0
+gguf==0.17.1
+h11==0.16.0
+hf-xet==1.1.10
+httpcore==1.0.9
+httptools==0.7.1
+httpx==0.28.1
+huggingface-hub==0.35.3
+idna==3.10
+interegular==0.3.3
+Jinja2==3.1.6
+jiter==0.11.0
+jsonschema==4.25.1
+jsonschema-specifications==2025.9.1
+lark==1.2.2
+llguidance==0.7.30
+llvmlite==0.44.0
+lm-format-enforcer==0.10.12
+markdown-it-py==4.0.0
+MarkupSafe==3.0.2
+mdurl==0.1.2
+mistral_common==1.8.5
+mpmath==1.3.0
+msgpack==1.1.2
+msgspec==0.19.0
+multidict==6.7.0
+networkx==3.4.2
+ninja==1.13.0
+numba==0.61.2
+numpy==2.2.6
+nvidia-cublas-cu12==12.9.1.4
+nvidia-cuda-cupti-cu12==12.9.79
+nvidia-cuda-nvrtc-cu12==12.9.86
+nvidia-cuda-runtime-cu12==12.9.79
+nvidia-cudnn-cu12==9.10.2.21
+nvidia-cufft-cu12==11.4.1.4
+nvidia-cufile-cu12==1.14.1.1
+nvidia-curand-cu12==10.3.10.19
+nvidia-cusolver-cu12==11.7.5.82
+nvidia-cusparse-cu12==12.5.10.65
+nvidia-cusparselt-cu12==0.7.1
+nvidia-nccl-cu12==2.27.5
+nvidia-nvjitlink-cu12==12.9.86
+nvidia-nvshmem-cu12==3.3.20
+nvidia-nvtx-cu12==12.9.79
+openai==1.90.0
+opencv-python-headless==4.12.0.88
+outlines_core==0.2.10
+packaging==25.0
+partial-json-parser==0.2.1.1.post6
+pillow==11.3.0
+prometheus-fastapi-instrumentator==7.1.0
+prometheus_client==0.23.1
+propcache==0.4.1
+protobuf==6.32.1
+psutil==7.1.0
+py-cpuinfo==9.0.0
+pybase64==1.4.2
+pycountry==24.6.1
+pycparser==2.23
+pydantic==2.12.0
+pydantic-extra-types==2.10.6
+pydantic_core==2.41.1
+Pygments==2.19.2
+python-dotenv==1.1.1
+python-json-logger==4.0.0
+python-multipart==0.0.20
+pytorch-triton==3.4.0+gitf7888497
+PyYAML==6.0.3
+pyzmq==27.1.0
+ray==2.49.2
+referencing==0.36.2
+regex==2025.9.18
+requests==2.32.5
+rich==14.2.0
+rich-toolkit==0.15.1
+rignore==0.7.0
+rpds-py==0.27.1
+safetensors==0.6.2
+scipy==1.15.3
+sentencepiece==0.2.1
+sentry-sdk==2.41.0
+setuptools-scm==9.2.0
+shellingham==1.5.4
+sniffio==1.3.1
+soundfile==0.13.1
+soxr==1.0.0
+starlette==0.48.0
+sympy==1.14.0
+tiktoken==0.12.0
+tokenizers==0.22.1
+tomli==2.3.0
+torch==2.9.0.dev20250905+cu129
+tqdm==4.67.1
+transformers==4.57.0
+triton==3.4.0
+typer==0.19.2
+typing-inspection==0.4.2
+typing_extensions==4.15.0
+urllib3==2.5.0
+uvicorn==0.37.0
+uvloop==0.21.0
+watchfiles==1.1.0
+websockets==15.0.1
+xgrammar==0.1.21
+yarl==1.22.0

.github/workflows/gpu_test.yaml

@@ -0,0 +1,67 @@
+name: GPU tests
+
+on:
+  schedule:
+    # Runs at midnight every day
+    - cron:  '0 0 * * *'
+  push:
+    branches: [ main ]
+  pull_request:
+  workflow_dispatch:
+
+concurrency:
+  group: gpu-test-${{ github.workflow }}-${{ github.ref == 'refs/heads/main' && github.run_number || github.ref }}
+  cancel-in-progress: true
+
+permissions:
+  id-token: write
+  contents: read
+
+defaults:
+  run:
+    shell: bash -l -eo pipefail {0}
+
+jobs:
+  gpu_test:
+    if: github.repository_owner == 'meta-pytorch'
+    runs-on: linux.g5.12xlarge.nvidia.gpu
+    strategy:
+      matrix:
+        python-version: ['3.10']
+    steps:
+      - name: Check out repo
+        uses: actions/checkout@v4
+      - name: Setup conda env
+        uses: conda-incubator/setup-miniconda@v2
+        with:
+          auto-update-conda: true
+          miniconda-version: "latest"
+          activate-environment: test
+          python-version: ${{ matrix.python-version }}
+      - name: Update pip
+        run: python -m pip install --upgrade pip
+      - name: Install pinned torch nightly
+        run: python -m pip install --pre torch==2.9.0.dev20250905 --no-cache-dir --index-url https://download.pytorch.org/whl/nightly/cu129
+      - name: Download and install vLLM and its dependencies
+        # TODO: this honestly could not be hackier if I tried
+        run: |
+          python -m pip install -r .github/packaging/vllm_reqs.txt
+          python -m pip install vllm==0.10.1.dev0+g6d8d0a24c.d20251009.cu129 --no-cache-dir --index-url https://download.pytorch.org/whl/preview/forge
+      - name: Download and install monarch and its dependencies
+        run: |
+          python -m pip install -r https://raw.githubusercontent.com/meta-pytorch/monarch/main/requirements.txt
+          python -m pip install torchmonarch --extra-index-url https://download.pytorch.org/whl/preview/forge
+      - name: Install torchtitan and torchstore
+        run: |
+          python -m pip install git+https://github.com/pytorch/torchtitan.git
+          python -m pip install git+https://github.com/meta-pytorch/torchstore.git
+      - name: Install dependencies
+        run: python -m pip install --no-build-isolation -e ".[dev]"
+      - name: Run unit tests with coverage
+        # TODO add all tests
+        run: |
+          export LD_PRELOAD=$CONDA/envs/test/lib/libpython3.10.so.1.0
+          export LD_LIBRARY_PATH=$CONDA/envs/test/lib/libpython3.10.so.1.0
+          pytest tests/unit_tests --cov=. --cov-report=xml --durations=20 -vv
+      - name: Upload Coverage to Codecov
+        uses: codecov/codecov-action@v3