# # Tests layout

# Each folder under tests/ corresponds to a test category for a sub-namespace in verl. For instance:
# - `tests/trainer` for testing functionality related to `verl/trainer`
# - `tests/models` for testing functionality related to `verl/models`
# - ...

# There are a few folders with `special_` prefix, created for special purposes:
# - `special_distributed`: unit tests that must run with multiple GPUs
# - `special_e2e`: end-to-end tests with training/generation scripts
# - `special_npu`: tests for NPUs
# - `special_sanity`: a suite of quick sanity tests
# - `special_standalone`: a set of test that are designed to run in dedicated environments

# Accelerators for tests 
# - By default tests are run with GPU available, except for the ones under `special_npu`, and any test script whose name ends with `on_cpu.py`.
# - For test scripts with `on_cpu.py` name suffix would be tested on CPU resources in linux environment.

# # Workflow layout

# All CI tests are configured by yaml files in `.github/workflows/`. Here's an overview of all test configs:
# 1. A list of always triggered CPU sanity tests: `check-pr-title.yml`, `secrets_scan.yml`, `check-pr-title,yml`, `pre-commit.yml`, `doc.yml`
# 2. Some heavy multi-GPU unit tests, such as `model.yml`, `vllm.yml`, `sgl.yml`
# 3. End-to-end tests: `e2e_*.yml`
# 4. Unit tests
#   - `cpu_unit_tests.yml`, run pytest on all scripts with file name pattern `tests/**/test_*_on_cpu.py`
#   - `gpu_unit_tests.yml`, run pytest on all scripts with file without the `on_cpu.py` suffix.
#   - Since cpu/gpu unit tests by default runs all tests under `tests`, please make sure tests are manually excluded in them when
#     - new workflow yaml is added to `.github/workflows`
#     - new tests are added to workflow mentioned in 2.
# name: Check PR Title

name: model_ascend

on:
  # Trigger the workflow on push or pull request,
  # but only for the main branch
  push:
    branches:
      - main
      - v0.*
  pull_request:
    branches:
      - main
      - v0.*
    paths:
      - "verl/**/*.py"
      # Entrypoints
      - ".github/workflows/model_ascend.yml"
      - "tests/special_distributed/test_fsdp_ckpt.py"
      - "tests/special_distributed/test_tensor_dict.py"
      - "tests/models/**"
      - "tests/special_distributed/run_all.sh"

# Cancel jobs on the same ref if a new one is triggered
concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: ${{ github.ref != 'refs/heads/main' }}

permissions:
  contents: read

jobs:
  model_rmpad_ascend:
    if: github.repository_owner == 'verl-project'
    runs-on: linux-aarch64-a2b3-8
    timeout-minutes: 60 # Increase this timeout value as needed
    container:
      image: swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/ascend-ci/verl/verl:verl-8.5.0-910b-ubuntu22.04-py3.11-latest
      options: >-
        --shm-size 16g
    env:
      HF_ENDPOINT: "https://hf-mirror.com"
      HF_HUB_ENABLE_HF_TRANSFER: "0" # This is more stable
    steps:
      - name: Check npu and CANN info
        run: |
          cat /usr/local/Ascend/ascend-toolkit/latest/"$(uname -i)"-linux/ascend_toolkit_install.info
          npu-smi info
      - name: Check initial pip list from image
        run: |
          pip list
      - name: Checkout verl-project/verl repo
        uses: actions/checkout@v4
        with:
          fetch-depth: 0
          clean: true
      - name: Install the current repository
        run: |
          pip install -r requirements-npu.txt
          pip install --no-deps -e .[test]
      - name: Check final pip list
        run: |
          pip list
      - name: Prepare weights
        run: |
          ln -s /root/.cache/models ~/models
      - name: Running rmpad model tests on 8 NPUs
        run: |
          pytest -s tests/models/test_transformer.py
      - name: Running FSDP rmpad model tests on 8 NPUs
        run: |
          STRATEGY=fsdp torchrun --nproc_per_node=8 tests/special_distributed/test_fsdp_ckpt.py
      - name: Running transformers ulysses tests on 8 NPUs
        run: |
          torchrun --nproc_per_node=8 -m pytest tests/models/test_transformers_ulysses.py
      - name: Run distributed test
        run: |
          bash tests/special_distributed/run_all.sh

  # TODO: Move this back to model_rmpad once FSDP2 is stable.
  # NOTE: List as an independent job to make rerun easier.
  model_rmpad_fsdp2_unstable_ascend:
    if: github.repository_owner == 'verl-project'
    runs-on: linux-aarch64-a2b3-8
    timeout-minutes: 60
    container:
      image: swr.cn-southwest-2.myhuaweicloud.com/modelfoundry/ascend-ci/verl/verl:verl-8.5.0-910b-ubuntu22.04-py3.11-latest
      options: >-
        --shm-size 16g
    env:
      HF_ENDPOINT: "https://hf-mirror.com"
      HF_HUB_ENABLE_HF_TRANSFER: "0" # This is more stable
    steps:
      - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
        with:
          fetch-depth: 0
      - name: Install the current repository
        run: |
          pip install -r requirements-npu.txt
          pip install --no-deps -e .[test]
      - name: Prepare weights
        run: |
          ln -s /root/.cache/models ~/models
      - name: Running FSDP2 rmpad model tests on 8 NPUs
        run: |
          STRATEGY=fsdp2 torchrun --nproc_per_node=8 tests/special_distributed/test_fsdp_ckpt.py