Add vLLM to dependency list since it's OSS-ed. #37
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| # Copyright 2025 Google LLC | ||
| # | ||
| # Licensed under the Apache License, Version 2.0 (the "License"); | ||
| # you may not use this file except in compliance with the License. | ||
| # You may obtain a copy of the License at | ||
| # | ||
| # https://www.apache.org/licenses/LICENSE-2.0 | ||
| # | ||
| # Unless required by applicable law or agreed to in writing, software | ||
| # distributed under the License is distributed on an "AS IS" BASIS, | ||
| # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
| # See the License for the specific language governing permissions and | ||
| # limitations under the License. | ||
| # This workflow will install Python dependencies, run tests and lint with a variety of Python versions | ||
| # For more information see: https://docs.github.com/en/actions/automating-builds-and-tests/building-and-testing-python | ||
| name: TPU Tests | ||
| on: | ||
| workflow_call: | ||
| secrets: | ||
| HF_TOKEN: | ||
| required: true | ||
| description: 'HuggingFace token for model downloads' | ||
| concurrency: | ||
| # Dedup pull requests (canceling previous runs of the same workflow for same PR), and scheduled runs but nothing else | ||
| group: ${{ github.event_name == 'pull_request' && format('{0}-pr-{1}', github.workflow, github.event.pull_request.number) || github.event_name == 'schedule' && format('{0}-schedule', github.workflow) || github.run_id }} | ||
| cancel-in-progress: true | ||
| env: | ||
| HF_HOME: ~/.cache/huggingface | ||
| HF_HUB_ENABLE_HF_TRANSFER: "1" | ||
| jobs: | ||
| run-prod: | ||
| runs-on: [linux-x86-ct5lp-224-8tpu] | ||
| environment: testing | ||
| container: | ||
| image: us-docker.pkg.dev/tpu-prod-env-multipod/jax-stable-stack/candidate/tpu:jax0.7.1_rev1 | ||
| options: --privileged | ||
| env: | ||
| CLOUD_TPU_ACCELERATOR: v5e-8 | ||
| JAX_PLATFORMS: tpu | ||
| steps: | ||
| # Cache Hugging Face hub | ||
| - name: Cache HF hub | ||
| uses: actions/cache@v4 | ||
| with: | ||
| path: ~/.cache/huggingface | ||
| key: hf-${{ runner.os }}-${{ hashFiles('pyproject.toml', 'requirements*.txt', 'constraints*.txt') }} | ||
| restore-keys: | | ||
| hf-${{ runner.os }}- | ||
| - name: Checkout code | ||
| uses: actions/checkout@v4 | ||
| with: | ||
| fetch-depth: 0 | ||
| - name: Install tunix dependencies | ||
| run: | | ||
| pip install -e .[prod] | ||
| pip install pytest pytest-xdist | ||
| - name: Verify TPU availability | ||
| run: | | ||
| python -c " | ||
| import jax | ||
| print(f'JAX version: {jax.__version__}') | ||
| print(f'JAX devices: {jax.devices()}') | ||
| # Check if we have TPU devices specifically | ||
| devices = jax.devices() | ||
| has_tpu = len(devices) > 0 and all(device.platform == 'tpu' for device in devices) | ||
| print(f'TPU available: {has_tpu}') | ||
| if not has_tpu: | ||
| print('ERROR: No TPU devices found! Expected TPU devices but got:', [device.platform for device in devices]) | ||
| exit(1) | ||
| else: | ||
| print(f'SUCCESS: Found {len(devices)} TPU device(s)') | ||
| " | ||
| - name: Run tunix model tests | ||
| run: | | ||
| python -m pytest tests/models/ -v --tb=short -m "not cpu_only and not gpu_only" | ||
| - name: Run tunix cli tests | ||
| env: | ||
| HF_TOKEN: ${{ secrets.HF_TOKEN }} | ||
| KAGGLE_USERNAME: ${{ secrets.KAGGLE_USERNAME }} | ||
| KAGGLE_KEY: ${{ secrets.KAGGLE_KEY }} | ||
| run: | | ||
| # Config tests that passed | ||
| python -m pytest tests/cli/ -v --tb=short \ | ||
| --ignore=tests/cli/utils/model_test.py | ||
| - name: Run tunix generation tests (PASSED only) | ||
| run: | | ||
| # tokenizer_adapter_test requires access to gated repo | ||
| python -m pytest tests/generate/ -v --tb=short \ | ||
| --ignore=tests/generate/vllm_sampler_test.py \ | ||
| --ignore=tests/generate/tokenizer_adapter_test.py | ||
| - name: Run tunix SFT tests | ||
| run: | | ||
| python -m pytest tests/sft/ -v --tb=short | ||
| - name: Run tunix SFT integration tests | ||
| env: | ||
| HF_TOKEN: ${{ secrets.HF_TOKEN }} | ||
| run: | | ||
| ./tests/sft/sft_tpu_smoke_test.sh | ||
| - name: Run tunix distillation tests | ||
| run: | | ||
| python -m pytest tests/distillation/ -v --tb=short | ||
| - name: Run tunix RL tests | ||
| run: | | ||
| # RL common tests that passed | ||
| # b/448133814: test_grpo_with_lora_model fails | ||
| python -m pytest tests/rl/ -v --tb=short -k "not test_grpo_with_lora_model" --ignore=tests/rl/experimental/agentic | ||
| - name: GRPO Integration Test | ||
| env: | ||
| HF_TOKEN: ${{ secrets.HF_TOKEN }} | ||
| run: | | ||
| # Download GSM8K dataset | ||
| mkdir -p /tmp/grpo_test/rl/grpo/data | ||
| python3 -c " | ||
| from datasets import load_dataset | ||
| import json | ||
| # Download and save GSM8K train split | ||
| dataset = load_dataset('openai/gsm8k', 'main', split='train') | ||
| train_data = [{'question': item['question'], 'answer': item['answer']} for item in dataset] | ||
| with open('/tmp/grpo_test/rl/grpo/data/gsm8k_train.json', 'w') as f: | ||
| json.dump(train_data, f) | ||
| # Download and save GSM8K test split | ||
| dataset = load_dataset('openai/gsm8k', 'main', split='test') | ||
| test_data = [{'question': item['question'], 'answer': item['answer']} for item in dataset] | ||
| with open('/tmp/grpo_test/rl/grpo/data/gsm8k_test.json', 'w') as f: | ||
| json.dump(test_data, f) | ||
| print('GSM8K dataset downloaded successfully') | ||
| " | ||
| # Run GRPO demo script with minimal configuration | ||
| python3 scripts/grpo_demo_llama3_qwen2.py \ | ||
| --root-dir=/tmp/grpo_test \ | ||
| --model-version=Qwen/Qwen2.5-0.5B-Instruct \ | ||
| --num-batches=8 \ | ||
| --num-test-batches=4 \ | ||
| --rollout-engine=vanilla | ||
| - name: Run tunix tests not covered by the above categories | ||
| run: | | ||
| # This category is to catch tests added but not covered by CI yet. Whenever you add new folders under tests/, please add a new category above and skip those tests here. | ||
| python -m pytest tests/ -v --tb=short --ignore=tests/models/ --ignore=tests/cli/ --ignore=tests/generate/ --ignore=tests/sft/ --ignore=tests/distillation/ --ignore=tests/rl/ || code=$? | ||
| if [ "${code:-0}" = "5" ]; then | ||
| echo "No tests collected (expected)." | ||
| exit 0 | ||
| else | ||
| exit "${code:-0}" | ||
| fi | ||
| run-dev: | ||
| runs-on: [linux-x86-ct5lp-224-8tpu] # your existing k8s-based runner labels | ||
| environment: testing | ||
| # IMPORTANT: remove the 'container:' block here; we want the job to run | ||
| # directly on the runner pod so it can call kubectl. | ||
| steps: | ||
| - name: Cache HF hub (runner-side only; pod won’t use this cache) | ||
| uses: actions/cache@v4 | ||
| with: | ||
| path: ~/.cache/huggingface | ||
| key: hf-${{ runner.os }}-${{ hashFiles('pyproject.toml', 'requirements*.txt', 'constraints*.txt') }} | ||
| restore-keys: | | ||
| hf-${{ runner.os }}- | ||
| - name: Checkout code | ||
| uses: actions/checkout@v4 | ||
| with: | ||
| fetch-depth: 0 | ||
| - name: Create test pod | ||
| run: | | ||
| set -eux | ||
| cat <<'YAML' | kubectl apply -f - | ||
| apiVersion: v1 | ||
| kind: Pod | ||
| metadata: | ||
| name: vllm-tests | ||
| labels: | ||
| app: vllm-tests | ||
| spec: | ||
| restartPolicy: Never | ||
| containers: | ||
| - name: test | ||
| # Use your exact image tag/digest here: | ||
| image: vllm/vllm-tpu:785d8b6410c3f1dc138947ea861a194b061f0293 | ||
| command: ["bash","-lc","sleep infinity"] | ||
| # Add privileged if your cluster policy allows and you truly need it. | ||
| securityContext: | ||
| privileged: true | ||
| YAML | ||
| - name: Wait for pod Ready | ||
| run: kubectl wait --for=condition=Ready pod/vllm-tests --timeout=300s | ||
| - name: Copy workspace into pod | ||
| run: | | ||
| set -eux | ||
| kubectl exec vllm-tests -- mkdir -p /workspace | ||
| # Copy your repo into the container | ||
| kubectl cp "${GITHUB_WORKSPACE}/." vllm-tests:/workspace | ||
| - name: Run tests inside the pod | ||
| env: | ||
| HF_TOKEN: ${{ secrets.HF_TOKEN }} | ||
| run: | | ||
| set -eux | ||
| # Everything below runs inside the container | ||
| kubectl exec -i vllm-tests -- bash -lc ' | ||
| set -eux | ||
| export HF_TOKEN="${HF_TOKEN:-}" | ||
| cd /workspace | ||
| python3 -V || true | ||
| # Base tooling | ||
| pip install --upgrade pip setuptools wheel | ||
| # Install Tunix (from PR checkout) | ||
| # Your original steps included removing TPU/torch stacks before installing. | ||
| pip uninstall -y torch torch-xla libtpu jax jaxlib || true | ||
| if [ -f constraints_dev.txt ]; then | ||
| pip install -c constraints_dev.txt -e .[dev] | ||
| else | ||
| pip install -e .[dev] | ||
| fi | ||
| # Install tpu-inference and test deps | ||
| pip uninstall -y torch torch-xla libtpu jax jaxlib || true | ||
| pip install tpu-inference==v0.11.1 | ||
| pip install pytest pytest-xdist | ||
| # Run the specific vLLM tests | ||
| pytest tests/generate/vllm_sampler_test.py -v --tb=short | ||
| pytest tests/generate/vllm_driver_test.py -v --tb=short | ||
| ' | ||
| - name: Pod logs (on failure) | ||
| if: failure() | ||
| run: kubectl logs vllm-tests --all-containers=true || true | ||
| - name: Cleanup pod | ||
| if: always() | ||
| run: kubectl delete pod vllm-tests --ignore-not-found | ||
