[CI] add CI for Huawei Ascend NPU

.github/workflows/huawei-ascend-npu.yml

@@ -0,0 +1,124 @@
+name: huawei-ascend-npu
+
+defaults:
+  run:
+    shell: bash -ieo pipefail {0}
+on:
+  workflow_dispatch:
+  pull_request:
+    paths:
+      - '.github/workflows/huawei-ascend-npu.yml'
+      - 'requirements/**'
+  schedule:
+    - cron: "0 0 * * *"
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+permissions:
+  contents: read
+  issues: write
+
+jobs:
+  unit-tests:
+    runs-on: [self-hosted, ascend, npu]
+    container:
+      image: ascendai/cann
+      ports:
+        - 80
+      volumes:
+        - /usr/local/dcmi:/usr/local/dcmi
+        - /usr/local/bin/npu-smi:/usr/local/bin/npu-smi 
+        - /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/
+        - /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info
+        - /etc/ascend_install.info:/etc/ascend_install.info   
+      options: --network host
+               --name deepspeed_unit-tests
+               --device /dev/davinci4
+               --device /dev/davinci_manager
+               --device /dev/devmm_svm
+               --device /dev/hisi_hdc
+               --shm-size "20g"
+               --entrypoint /bin/bash
+
+    env:
+      PT_HPU_LAZY_MODE: 0
+      TORCHINDUCTOR_COMPILE_THREADS: 1
+      TEST_LIST: |
+        test_accelerator.py
+        test_autotuning.py
+        test_compression.py
+
+    steps:
+    - uses: actions/checkout@v4
+
+    - name: Install pytorch 
+      run: |
+        npu-smi info
+        apt-get update   
+        source /root/.bashrc
+
+        pip install torch==2.2.0 torchvision==0.17.0 torch_npu==2.2.0 numpy==1.26.4 cloudpickle tornado ml-dtypes -i https://pypi.tuna.tsinghua.edu.cn/simple
+
+        python << EOF 
+        if __name__ == '__main__':
+            import torch
+            import torch_npu
+            torch_npu.npu.set_device("npu:0") 
+            print(f"Device Name: {torch.npu.get_device_name(0)}")
+            print(f"Device Count: {torch.npu.device_count()}")
+            print(f"Device Available: {torch.npu.is_available()}")
+        EOF
+
+    - name: Install transformers
+      run: |
+        source /root/.bashrc
+        echo "y" | apt-get install git
+        git clone https://github.com/huggingface/transformers
+        cd transformers
+        git rev-parse --short HEAD
+        pip install . -i https://pypi.tuna.tsinghua.edu.cn/simple
+
+    - name: Install deepspeed
+      run: |
+        source /root/.bashrc
+        pip install .[1bit,npu,autotuning,inf] -i  https://pypi.tuna.tsinghua.edu.cn/simple
+        ds_report
+
+    - name: Python environment
+      run: |
+        source /root/.bashrc
+        pip list
+
+    - name: Unit tests
+      run: |
+        source /root/.bashrc
+        unset TORCH_CUDA_ARCH_LIST # only jit compile for current arch
+
+        cd tests/unit
+        pytest --verbose accelerator/*
+        pytest --verbose autotuning/*
+        pytest --verbose checkpoint/test_reshape_checkpoint.py
+        pytest --verbose checkpoint/test_moe_checkpoint.py
+        pytest --verbose checkpoint/test_shared_weights.py
+        pytest --verbose launcher/test_ds_arguments.py launcher/test_run.py
+        pytest --verbose model_parallelism/*
+        pytest --verbose moe/test_moe_tp.py
+        pytest --verbose monitor/*
+        pytest --verbose utils/*
+        pytest --verbose runtime/test_ds_config_model.py
+        pytest --verbose runtime/pipe/test_pipe_schedule.py
+        pytest --verbose runtime/zero/test_zero_config.py
+        pytest --verbose runtime/zero/test_zero_tiled.py
+        pytest --verbose runtime/zero/test_zeropp.py
+        pytest --verbose runtime/test_autocast.py
+        pytest --verbose runtime/test_data.py
+        pytest --verbose runtime/test_runtime_utils.py
+        pytest --verbose runtime/activation_checkpointing/*
+        pytest --verbose runtime/utils/*
+        pytest --verbose runtime/zero/test_zero_dynamic_class.py
+
+
+
+

README.md

@@ -141,6 +141,7 @@ DeepSpeed has been integrated with several different popular open-source DL fram
 | PyTorch Nightly | [![nv-torch-nightly-v100](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-torch-nightly-v100.yml/badge.svg?branch=master)](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-torch-nightly-v100.yml) |
 | Integrations | [![nv-transformers-v100](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-transformers-v100.yml/badge.svg?branch=master)](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-transformers-v100.yml) [![nv-lightning-v100](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-lightning-v100.yml/badge.svg?branch=master)](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-lightning-v100.yml) [![nv-accelerate-v100](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-accelerate-v100.yml/badge.svg?branch=master)](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-accelerate-v100.yml) [![nv-mii](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-mii.yml/badge.svg?branch=master)](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-mii.yml) [![nv-ds-chat](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-ds-chat.yml/badge.svg?branch=master)](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-ds-chat.yml) [![nv-sd](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-sd.yml/badge.svg)](https://github.com/microsoft/DeepSpeed/actions/workflows/nv-sd.yml) |
 | Misc | [![Formatting](https://github.com/microsoft/DeepSpeed/actions/workflows/formatting.yml/badge.svg?branch=master)](https://github.com/microsoft/DeepSpeed/actions/workflows/formatting.yml) [![pages-build-deployment](https://github.com/microsoft/DeepSpeed/actions/workflows/pages/pages-build-deployment/badge.svg)](https://github.com/microsoft/DeepSpeed/actions/workflows/pages/pages-build-deployment) [![Documentation Status](https://readthedocs.org/projects/deepspeed/badge/?version=latest)](https://deepspeed.readthedocs.io/en/latest/?badge=latest)[![python](https://github.com/microsoft/DeepSpeed/actions/workflows/python.yml/badge.svg?branch=master)](https://github.com/microsoft/DeepSpeed/actions/workflows/python.yml) |
+| Huawei Ascend NPU |  [![Huawei Ascend NPU](https://github.com/xuedinge233/DeepSpeed/actions/workflows/huawei-ascend-npu.yml/badge.svg?branch=master)](https://github.com/xuedinge233/DeepSpeed/actions/workflows/huawei-ascend-npu.yml) 
 
 # Installation
 

requirements/requirements-npu.txt

@@ -0,0 +1,21 @@
+accelerate
+clang-format==18.1.3
+comet_ml>=3.41.0
+# deepspeed-kernels ; sys_platform == 'linux'
+docutils<0.18
+future
+importlib-metadata>=4
+mup
+pre-commit>=2.20.0
+pytest>=7.2.0
+pytest-forked
+pytest-randomly
+pytest-xdist
+qtorch==0.3.0
+recommonmark
+sphinx>=5.3.0
+sphinx-rtd-theme
+tensorboard
+torchvision
+transformers>=4.39.0
+wandb

setup.py

@@ -91,6 +91,7 @@ def get_env_if_set(key, default: typing.Any = ""):
     'inf': fetch_requirements('requirements/requirements-inf.txt'),
     'sd': fetch_requirements('requirements/requirements-sd.txt'),
     'triton': fetch_requirements('requirements/requirements-triton.txt'),
+    'npu': fetch_requirements('requirements/requirements-npu.txt'),
 }
 
 # Only install pynvml on nvidia gpus.