Merge branch 'master' into reduce_coalesced_fetch_bubble

.github/workflows/cpu-torch-latest.yml

@@ -37,6 +37,15 @@ jobs:
           python -c "import torch; print('torch:', torch.__version__, torch)"
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
 
+      - name: Install transformers
+        run: |
+          git clone https://github.com/huggingface/transformers
+          cd transformers
+          # if needed switch to the last known good SHA until transformers@master is fixed
+          # git checkout 6c3f168b3
+          git rev-parse --short HEAD
+          pip install .
+
       - name: Install deepspeed
         run: |
           pip install .[dev,autotuning]

.github/workflows/hpu-gaudi2.yml

@@ -111,7 +111,8 @@ jobs:
         run: |
           git clone https://github.com/huggingface/transformers
           cd transformers
-          git checkout 7df93d6ffb48946e532c7d766fe10372e98d78b6
+          # if needed switch to the last known good SHA until transformers@master is fixed
+          git checkout 6c3f168b3
           git rev-parse --short HEAD
           pip install .
 

.github/workflows/no-torch.yml

@@ -4,6 +4,7 @@ on:
   workflow_dispatch:
   pull_request:
     paths:
+      - 'accelerator/**'
       - '.github/workflows/no-torch.yml'
       - 'op_builder/**'
   schedule:

.github/workflows/nv-a6000.yml

@@ -40,9 +40,8 @@ jobs:
           python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
       - name: Install transformers
         run: |
-          git clone https://github.com/huggingface/transformers
+          git clone --depth=1 https://github.com/huggingface/transformers
           cd transformers
-          git checkout 7df93d6ffb48946e532c7d766fe10372e98d78b6
           git rev-parse --short HEAD
           python -m pip install .
       - name: Install deepspeed

.github/workflows/nv-accelerate-v100.yml

@@ -52,7 +52,5 @@ jobs:
           pip install .[testing]
           # force protobuf version due to issues
           pip install "protobuf<4.21.0"
-          # tmp fix: force newer datasets version
-          #pip install "datasets>=2.0.0"
           pip list
           pytest $PYTEST_OPTS --color=yes --durations=0 --verbose tests/deepspeed

.github/workflows/nv-flash-attn.yml

@@ -0,0 +1,64 @@
+name: nv-flash-attn
+
+on:
+  workflow_dispatch:
+  pull_request:
+    paths:
+      - 'deepspeed/sequence/**'
+      - 'tests/unit/sequence_parallelism/**'
+      - '.github/workflows/nv-flash-attn.yml'
+  schedule:
+    - cron: "0 0 * * *"
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  unit-tests:
+    runs-on: [self-hosted, nvidia, a6000]
+    container:
+      image: nvcr.io/nvidia/pytorch:24.03-py3
+      ports:
+        - 80
+      options: --gpus all --shm-size "8G"
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Check container state
+        run: |
+          ldd --version
+          nvcc --version
+          nvidia-smi
+          python -c "import torch; print('torch:', torch.__version__, torch)"
+          python -c "import torch; print('CUDA available:', torch.cuda.is_available())"
+      - name: Install transformers
+        run: |
+          git clone --depth=1 https://github.com/huggingface/transformers
+          cd transformers
+          git rev-parse --short HEAD
+          python -m pip install .
+      - name: Install deepspeed
+        run: |
+          python -m pip install .[dev]
+          ds_report
+      - name: Install FlashAttention
+        run: |
+          python -m pip install flash-attn
+      - name: Python environment
+        run: |
+          python -m pip list
+      - name: Unit tests
+        run: |
+          unset TORCH_CUDA_ARCH_LIST # only jit compile for current arch
+          cd tests
+          python -m pytest --color=yes --durations=0 --verbose -rF unit/sequence_parallelism/test_ulysses.py --torch_ver="2.3" --cuda_ver="12"
+      - name: Open GitHub issue if nightly CI fails
+        if: ${{ failure() && (github.event_name == 'schedule') }}
+        uses: JasonEtco/create-an-issue@v2
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          filename: .github/ISSUE_TEMPLATE/ci_failure_report.md
+          update_existing: true

.github/workflows/nv-torch-latest-v100.yml

@@ -38,7 +38,7 @@ jobs:
           git clone https://github.com/huggingface/transformers
           cd transformers
           # if needed switch to the last known good SHA until transformers@master is fixed
-          # git checkout 1cc453d33
+          # git checkout 6c3f168b3
           git rev-parse --short HEAD
           pip install .
 

.github/workflows/nv-torch-nightly-v100.yml

@@ -2,6 +2,9 @@ name: nv-torch-nightly-v100
 
 on:
   workflow_dispatch:
+    pull_request:
+    paths:
+      - '.github/workflows/nv-torch-nightly-v100.yml'
   schedule:
     - cron: "0 0 * * *"
 
@@ -34,7 +37,7 @@ jobs:
           git clone https://github.com/huggingface/transformers
           cd transformers
           # if needed switch to the last known good SHA until transformers@master is fixed
-          # git checkout 1cc453d33
+          # git checkout 6c3f168b3
           git rev-parse --short HEAD
           pip install .
 

.github/workflows/release.yml

@@ -25,6 +25,7 @@ jobs:
         python release/check_release_version.py --release_version ${{ env.RELEASE_VERSION }}
     - name: Build DeepSpeed
       run: |
+        pip install setuptools
         DS_BUILD_STRING=" " python setup.py sdist
     - name: Publish to PyPI
       uses: pypa/gh-action-pypi-publish@release/v1

.github/workflows/xpu-compile.yml

@@ -31,10 +31,10 @@ jobs:
       run: |
         apt-get update
         apt-get install clinfo libaio-dev python3-pip -y
-        pip install torch==2.3.1 -f https://pytorch-extension.intel.com/release-whl/stable/xpu/us/torch/
-        pip install intel-extension-for-pytorch==2.3.110+xpu -f https://pytorch-extension.intel.com/release-whl/stable/xpu/us/intel-extension-for-pytorch/
-        pip install oneccl_bind_pt==2.3.100+xpu -f https://pytorch-extension.intel.com/release-whl/stable/xpu/us/oneccl-bind-pt/
-        pip install torchvision==0.18.1 -f https://pytorch-extension.intel.com/release-whl/stable/xpu/us/torchvision/
+        pip install torch==2.3.1 -f https://pytorch-extension.intel.com/release-whl/stable/xpu/cn/torch/
+        pip install intel-extension-for-pytorch==2.3.110+xpu -f https://pytorch-extension.intel.com/release-whl/stable/xpu/cn/intel-extension-for-pytorch/
+        pip install oneccl_bind_pt==2.3.100+xpu -f https://pytorch-extension.intel.com/release-whl/stable/xpu/cn/oneccl-bind-pt/
+        pip install torchvision==0.18.1 -f https://pytorch-extension.intel.com/release-whl/stable/xpu/cn/torchvision/
         pip install https://github.com/intel/intel-xpu-backend-for-triton/releases/download/v3.0.0b2/triton_xpu-3.0.0b2-cp310-cp310-linux_x86_64.whl
         pip install py-cpuinfo numpy
         pip install .[dev,autotuning]

.github/workflows/xpu-max1100.yml

@@ -47,10 +47,10 @@ jobs:
       run: |
         apt-get update
         apt-get install clinfo libaio-dev python3-pip -y
-        pip install torch==2.3.1 -f https://pytorch-extension.intel.com/release-whl/stable/xpu/us/torch/
-        pip install intel-extension-for-pytorch==2.3.110+xpu -f https://pytorch-extension.intel.com/release-whl/stable/xpu/us/intel-extension-for-pytorch/
-        pip install oneccl_bind_pt==2.3.100+xpu -f https://pytorch-extension.intel.com/release-whl/stable/xpu/us/oneccl-bind-pt/
-        pip install torchvision==0.18.1 -f https://pytorch-extension.intel.com/release-whl/stable/xpu/us/torchvision/
+        pip install torch==2.3.1 -f https://pytorch-extension.intel.com/release-whl/stable/xpu/cn/torch/
+        pip install intel-extension-for-pytorch==2.3.110+xpu -f https://pytorch-extension.intel.com/release-whl/stable/xpu/cn/intel-extension-for-pytorch/
+        pip install oneccl_bind_pt==2.3.100+xpu -f https://pytorch-extension.intel.com/release-whl/stable/xpu/cn/oneccl-bind-pt/
+        pip install torchvision==0.18.1 -f https://pytorch-extension.intel.com/release-whl/stable/xpu/cn/torchvision/
         pip install py-cpuinfo numpy
         pip install .[dev,autotuning]
 

.pre-commit-config.yaml

@@ -5,7 +5,7 @@ repos:
     -   id: check-useless-excludes
 
 -   repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.0.1
+    rev: v5.0.0
     hooks:
     -   id: check-case-conflict
     -   id: check-json

CODEOWNERS

@@ -8,49 +8,52 @@
 
 # top-level repo folders
 /.github/ @loadams
-/azure/ @awan-10
-/benchmarks/ @awan-10 @tjruwase
+/azure/ @loadams
+/benchmarks/ @guanhuawang @tjruwase
 /bin/ @loadams
-/csrc/ @awan-10
+/csrc/ @tjruwase
 /deepspeed/ @loadams @tjruwase
-/docker/ @awan-10
+/docker/ @loadams @guanhuawang
 /docs/ @loadams @tjruwase
-/examples/ @awan-10 @tohtana
+/examples/ @jomayeri @tohtana
 /op_builder/ @loadams @tjruwase @jomayeri
-/release/ @loadams
+/release/ @loadams @jomayeri
 /requirements/ @loadams
-/scripts/ @awan-10
+/scripts/ @loadams @tjruwase
 /tests/ @tjruwase @loadams @tohtana
 
 # deepspeed
 /deepspeed/autotuning/ @loadams
 /deepspeed/checkpoint/ @tjruwase
-/deepspeed/comm/ @awan-10
+/deepspeed/comm/ @guanhuawang
 /deepspeed/compression/ @tjruwase
-/deepspeed/elasticity/ @awan-10
+/deepspeed/elasticity/ @tjruwase
 /deepspeed/launcher/ @loadams
-/deepspeed/module_inject/ @awan-10
+/deepspeed/module_inject/ @hwchen2017 @loadams
 /deepspeed/moe/ @tohtana
-/deepspeed/monitor/ @awan-10
+/deepspeed/monitor/ @tjruwase
 /deepspeed/nebula/ @tjruwase
+/deepspeed/nvme/ @tjruwase @jomayeri
 /deepspeed/ops/ @tohtana
 /deepspeed/pipe/ @tohtana @loadams
 /deepspeed/profiling/ @loadams
-/deepspeed/utils/ @tjruwase @awan-10
+/deepspeed/sequence/ @tohtana
+/deepspeed/utils/ @tjruwase @tohtana
 
 # inference
-/deepspeed/inference/ @awan-10
-/deepspeed/model_implementations/ @awan-10
+/deepspeed/inference/ @hwchen2017 @tohtana
+/deepspeed/model_implementations/@tohtana @loadams
 
 # training
 /deepspeed/runtime/ @tjruwase @tohtana
 /deepspeed/runtime/activation_checkpointing/ @tjruwase
 /deepspeed/runtime/checkpoint_engine/ @tjruwase
-/deepspeed/runtime/comm/ @awan-10
-/deepspeed/runtime/compression/ @awan-10
+/deepspeed/runtime/comm/ @guanhuawang
+/deepspeed/runtime/compression/ @tjruwase
 /deepspeed/runtime/data_pipeline/ @tjruwase
-/deepspeed/runtime/fp16/ @tjruwase
-/deepspeed/runtime/fp16/onebit/ @awan-10
-/deepspeed/runtime/pipe/ @loadams
-/deepspeed/runtime/swap_tensor/ @tjruwase
-/deepspeed/runtime/zero/ @tjruwase
+/deepspeed/runtime/domino/ @guanhuawang @hwchen2017
+/deepspeed/runtime/fp16/ @tjruwase @tohtana
+/deepspeed/runtime/fp16/onebit/ @tjruwase
+/deepspeed/runtime/pipe/ @loadams @tohtana
+/deepspeed/runtime/swap_tensor/ @tjruwase @jomayeri
+/deepspeed/runtime/zero/ @tjruwase @tohtana

COMMITTERS.md

@@ -5,5 +5,7 @@
 | Olatunji Ruwase | [tjruwase](https://github.com/tjruwase)     | Microsoft |
 | Logan Adams     | [loadams](https://github.com/loadams)      | Microsoft |
 | Masahiro Tanaka | [tohtana](https://github.com/tohtana)      | Microsoft |
-| Jeff Rasley     | [jeffra](https://github.com/jeffra)       | SnowFlake |
-| Minjia Zhang    | [minjiazhang](https://github.com/minjiazhang)  | UIUC      |
+| Jeff Rasley     | [jeffra](https://github.com/jeffra)       | SnowFlake  |
+| Minjia Zhang    | [minjiazhang](https://github.com/minjiazhang)  | UIUC  |
+| Ashwin Aji      | [ashwinma](https://github.com/ashwinma)        | AMD   |
+| Sam Foreman     | [saforem2](https://github.com/saforem2)        | Argonne National Laboratory |

README.md

@@ -16,7 +16,8 @@
 ## Latest News
 <b> <span style="color:orange" > DeepSpeed empowers ChatGPT-like model training with a single click, offering 15x speedup over SOTA RLHF systems with unprecedented cost reduction at all scales; [learn how](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-chat)</span>.</b>
 
-
+* [2024/12] [Ulysses-Offload: Democratizing Long Context LLM Training ](https://github.com/microsoft/DeepSpeed/blob/master/blogs/ulysses-offload/README.md)
+* [2024/12] [DeepSpeed-Domino: Communication-Free LLM Training Engine](https://github.com/microsoft/DeepSpeed/blob/master/blogs/deepspeed-domino/README.md)
 * [2024/08] [DeepSpeed on Windows](https://github.com/microsoft/DeepSpeed/tree/master/blogs/windows/08-2024/README.md) [[日本語](https://github.com/microsoft/DeepSpeed/tree/master/blogs/windows/08-2024/japanese/README.md)]
 * [2024/08] [DeepNVMe: Improving DL Applications through I/O Optimizations](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-gds/README.md) [[日本語](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-gds/japanese/README.md)]
 * [2024/07] [DeepSpeed Universal Checkpointing: Efficient and Flexible Checkpointing for Large Scale Distributed Training](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-ucp/README.md) [[中文](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-ucp/chinese/README.md)] [[日本語](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-ucp/japanese/README.md)]
@@ -26,7 +27,7 @@
 * [2023/11] [DeepSpeed ZeRO-Offload++: 6x Higher Training Throughput via Collaborative CPU/GPU Twin-Flow](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-offloadpp)
 * [2023/11] [DeepSpeed-FastGen: High-throughput Text Generation for LLMs via MII and DeepSpeed-Inference](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-fastgen) [[English](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-fastgen)] [[中文](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-fastgen/chinese/README.md)] [[日本語](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-fastgen/japanese/README.md)]
 * [2023/10] [DeepSpeed-VisualChat: Improve Your Chat Experience with Multi-Round Multi-Image Inputs](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-visualchat/10-03-2023/README.md) [[English](https://github.com/microsoft/DeepSpeed/tree/master/blogs/deepspeed-visualchat/10-03-2023/README.md)] [[中文](https://github.com/microsoft/DeepSpeed/blob/master/blogs/deepspeed-visualchat/10-03-2023/README-Chinese.md)] [[日本語](https://github.com/microsoft/DeepSpeed/blob/master/blogs/deepspeed-visualchat/10-03-2023/README-Japanese.md)]
-* [2023/09] Announcing the DeepSpeed4Science Initiative: Enabling large-scale scientific discovery through sophisticated AI system technologies [[DeepSpeed4Science website](https://deepspeed4science.ai/)] [[Tutorials](https://www.deepspeed.ai/deepspeed4science/)] [[White paper](https://arxiv.org/abs/2310.04610)] [[Blog](https://www.microsoft.com/en-us/research/blog/announcing-the-deepspeed4science-initiative-enabling-large-scale-scientific-discovery-through-sophisticated-ai-system-technologies/)] [[中文](https://github.com/microsoft/DeepSpeed/blob/master/blogs/deepspeed4science/chinese/README.md)] [[日本語](https://github.com/microsoft/DeepSpeed/blob/master/blogs/deepspeed4science/japanese/README.md)]
+* [2023/09] Announcing the DeepSpeed4Science Initiative: Enabling large-scale scientific discovery through sophisticated AI system technologies [[Tutorials](https://www.deepspeed.ai/deepspeed4science/)] [[White paper](https://arxiv.org/abs/2310.04610)] [[Blog](https://www.microsoft.com/en-us/research/blog/announcing-the-deepspeed4science-initiative-enabling-large-scale-scientific-discovery-through-sophisticated-ai-system-technologies/)] [[中文](https://github.com/microsoft/DeepSpeed/blob/master/blogs/deepspeed4science/chinese/README.md)] [[日本語](https://github.com/microsoft/DeepSpeed/blob/master/blogs/deepspeed4science/japanese/README.md)]
 
 
 <!-- NOTE: we must use html for news items otherwise links will be broken in the 'more news' section -->
@@ -77,7 +78,7 @@ To further increase the inference efficiency, DeepSpeed offers easy-to-use and f
 
 ## DeepSpeed4Science
 
-In line with Microsoft's mission to solve humanity's most pressing challenges, the DeepSpeed team at Microsoft is responding to this opportunity by launching a new initiative called *DeepSpeed4Science*, aiming to build unique capabilities through AI system technology innovations to help domain experts to unlock today's biggest science mysteries. Learn more: [DeepSpeed4Science website](https://deepspeed4science.ai/) and [tutorials](https://www.deepspeed.ai/deepspeed4science/)
+In line with Microsoft's mission to solve humanity's most pressing challenges, the DeepSpeed team at Microsoft is responding to this opportunity by launching a new initiative called *DeepSpeed4Science*, aiming to build unique capabilities through AI system technology innovations to help domain experts to unlock today's biggest science mysteries. Learn more: [tutorials](https://www.deepspeed.ai/deepspeed4science/)
 
 ---
 
@@ -121,7 +122,7 @@ DeepSpeed has been integrated with several different popular open-source DL fram
 
 |                                                                                                | Documentation                                |
 | ---------------------------------------------------------------------------------------------- | -------------------------------------------- |
-<img src="docs/assets/images/transformers-light.png#gh-light-mode-only" width="250px"><img src="docs/assets/images/transformers-dark.png#gh-dark-mode-only" width="250px"> | [Transformers with DeepSpeed](https://huggingface.co/docs/transformers/main/main_classes/deepspeed) |
+<img src="docs/assets/images/transformers-light.png#gh-light-mode-only" width="250px"><img src="docs/assets/images/transformers-dark.png#gh-dark-mode-only" width="250px"> | [Transformers with DeepSpeed](https://huggingface.co/docs/transformers/deepspeed) |
 | <img src="docs/assets/images/accelerate-light.png#gh-light-mode-only" width="250px"><img src="docs/assets/images/accelerate-dark.png#gh-dark-mode-only" width="250px"> | [Accelerate with DeepSpeed](https://huggingface.co/docs/accelerate/usage_guides/deepspeed) |
 | <img src="docs/assets/images/lightning-light.svg#gh-light-mode-only" width="200px"><img src="docs/assets/images/lightning-dark.svg#gh-dark-mode-only" width="200px"> | [Lightning with DeepSpeed](https://lightning.ai/docs/pytorch/stable/advanced/model_parallel.html#deepspeed) |
 | <img src="docs/assets/images/mosaicml.svg" width="200px"> | [MosaicML with DeepSpeed](https://docs.mosaicml.com/projects/composer/en/latest/trainer/using_the_trainer.html?highlight=deepspeed#deepspeed-integration) |

accelerator/cpu_accelerator.py

@@ -3,9 +3,15 @@
 
 # DeepSpeed Team
 
-import torch
 from .abstract_accelerator import DeepSpeedAccelerator
 
+# During setup stage torch may not be installed, pass on no torch will
+# allow op builder related API to be executed.
+try:
+    import torch
+except ImportError as e:
+    pass
+
 try:
     import oneccl_bindings_for_pytorch  # noqa: F401 # type: ignore
     oneccl_imported_p = True

accelerator/hpu_accelerator.py

@@ -21,8 +21,8 @@ def __init__(self):
         self.apply_hpu_workarounds()
         try:
             import habana_frameworks.torch.hpu as hpu
-            hpu.setDeterministic(True)
             self.hpu = hpu
+            torch.use_deterministic_algorithms(True)
         except ImportError as e:
             raise ValueError(
                 f"HPU_Accelerator requires habana_frameworks.torch.hpu, which is not installed on this system.")