Skip to content

GitLab

Commit 7d1a83a9 authored by aiss's avatar aiss
Browse files

push Deepspeed 0.6.3 rocm version

parent ab5534fc
Hide whitespace changes
Inline Side-by-side
Showing with 184 additions and 107 deletions
CODE_OF_CONDUCT.md
View file @ 7d1a83a9
# Microsoft Open Source Code of Conduct # Microsoft Open Source Code of Conduct
This project has adopted the [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/). This project has adopted the [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/).
Resources: Resources:
- [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/) - [Microsoft Open Source Code of Conduct](https://opensource.microsoft.com/codeofconduct/)
- [Microsoft Code of Conduct FAQ](https://opensource.microsoft.com/codeofconduct/faq/) - [Microsoft Code of Conduct FAQ](https://opensource.microsoft.com/codeofconduct/faq/)
- Contact [opencode@microsoft.com](mailto:opencode@microsoft.com) with questions or concerns - Contact [opencode@microsoft.com](mailto:opencode@microsoft.com) with questions or concerns
CONTRIBUTING.md
View file @ 7d1a83a9
...@@ -65,3 +65,31 @@ Conduct](https://opensource.microsoft.com/codeofconduct/). For more information ...@@ -65,3 +65,31 @@ Conduct](https://opensource.microsoft.com/codeofconduct/). For more information
[Code of Conduct FAQ](https://opensource.microsoft.com/codeofconduct/faq/) or contact [Code of Conduct FAQ](https://opensource.microsoft.com/codeofconduct/faq/) or contact
[opencode@microsoft.com](mailto:opencode@microsoft.com) with any additional questions or [opencode@microsoft.com](mailto:opencode@microsoft.com) with any additional questions or
comments. comments.
## New Feature Contribution Guidelines
Unlike bug fix or improving existing feature (where users usually directly submit a PR and we review it), adding a new feature to DeepSpeed requires several steps: (1) proposal and discussion, (2) implementation and verification, (3) release and maintenance. This general guideline applies to all new feature contributions. Core DeepSpeed team member contributions may complete step 1 internally.
### Step 1: proposal and discussion
We ask users to first post your intended feature in an issue. This issue needs to include:
* A description of the proposed feature.
* A motivation of why it will be useful to DeepSpeed users.
* A rough design of how you implement the feature inside DeepSpeed.
* (Important) Results or planned experiments to demonstrate the effectiveness and correctness of the feature.
* If this is a general feature applicable to different tasks, we require testing it on at least one CV task (e.g., [CIFAR](https://www.deepspeed.ai/tutorials/cifar-10/)) and one NLP task (e.g., [SQuAD](https://www.deepspeed.ai/tutorials/bert-finetuning/)). If this is a feature for one kind of task only, it is fine to just test on the specific task.
* If the feature only affects performance and does not affect training convergence, we require testing on a fraction of training to demonstrate that the training/validation loss are consistent with baseline, and that the performance is better than baseline.
* If the feature does affect training convergence, we require testing the whole training to demonstrate that the feature achieves better/on-par final model quality and training performance compared to baseline.
Based on the issue we shall discuss the merit of the new feature and decide whether accept or decline the proposal. Once accepted and after we confirm the design and implementation plan, we are ready for step 2.
### Step 2: implementation and verification
Contributor will go ahead and implement the feature, and the DeepSpeed team will provide guidance/helps as needed. The required deliverables include:
* A PR to [microsoft/DeepSpeed](https://github.com/microsoft/DeepSpeed) including (1) the feature implementation (2) unit tests (3) documentation (4) tutorial
* A PR to [microsoft/DeepSpeedExamples](https://github.com/microsoft/DeepSpeedExamples) or [microsoft/Megatron-DeepSpeed](https://github.com/microsoft/Megatron-DeepSpeed) including the examples of how to use the feature (this is related to the planned testing experiments in proposal)
* In the implementation (code, documentation, tutorial), we require the feature author to record their GitHub username as a contact method for future questions/maintenance.
After receiving the PRs, we will review them and merge them after necessary tests/fixes.
### Step 3: release and maintenance
After the PRs are merged, we will announce the feature on our website (with credit to the feature author). We ask the feature author to commit to the maintenance of the feature.
LICENSE
View file @ 7d1a83a9
MIT License MIT License
Copyright (c) Microsoft Corporation. Copyright (c) Microsoft Corporation.
Permission is hereby granted, free of charge, to any person obtaining a copy Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions: furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software. copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE SOFTWARE
MANIFEST.in
View file @ 7d1a83a9
include *.txt README.md include *.txt README.md
recursive-include requirements *.txt recursive-include requirements *.txt
recursive-include deepspeed *.cpp *.h *.cu *.tr *.cuh *.cc recursive-include deepspeed *.cpp *.h *.cu *.hip *.tr *.cuh *.cc *.json
recursive-include csrc *.cpp *.h *.cu *.tr *.cuh *.cc recursive-include csrc *.cpp *.h *.cu *.tr *.cuh *.cc
MANIFEST_win.in 0 → 100644
View file @ 7d1a83a9
include *.txt README.md
recursive-include requirements *.txt
# this is for Windows only
recursive-include deepspeed *.tr
recursive-exclude deepspeed/ops/csrc *.cpp *.h *.cu *.cuh *.cc
prune csrc
prune op_builder
README.md 100755 → 100644
View file @ 7d1a83a9
...@@ -2,9 +2,28 @@ ...@@ -2,9 +2,28 @@
[![PyPI version](https://badge.fury.io/py/deepspeed.svg)](https://pypi.org/project/deepspeed/) [![PyPI version](https://badge.fury.io/py/deepspeed.svg)](https://pypi.org/project/deepspeed/)
[![Documentation Status](https://readthedocs.org/projects/deepspeed/badge/?version=latest)](https://deepspeed.readthedocs.io/en/latest/?badge=latest) [![Documentation Status](https://readthedocs.org/projects/deepspeed/badge/?version=latest)](https://deepspeed.readthedocs.io/en/latest/?badge=latest)
[![License MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://github.com/Microsoft/DeepSpeed/blob/master/LICENSE) [![License MIT](https://img.shields.io/badge/License-MIT-blue.svg)](https://github.com/Microsoft/DeepSpeed/blob/master/LICENSE)
[![Docker Pulls](https://img.shields.io/docker/pulls/deepspeed/deepspeed)](https://hub.docker.com/r/deepspeed/deepspeed)
### 03/2021: DeepSpeed is hiring! Come join us: [SDE 2](https://careers.microsoft.com/us/en/job/1013160/Software-Engineer-2), [Sr. SDE](https://careers.microsoft.com/us/en/job/1017151/Senior-Software-Engineer), [Sr. Researcher](https://careers.microsoft.com/us/en/job/1016440/Senior-Researcher) <div align="center">
<img src="docs/assets/images/DeepSpeed_light.svg#gh-light-mode-only" width="400px">
<img src="docs/assets/images/DeepSpeed_dark_transparent.svg#gh-dark-mode-only" width="400px">
</div>
<!--
Remove until pypi issue is resolved: https://status.python.org/incidents/2jj696st6yn5
[![Downloads](https://pepy.tech/badge/deepspeed/month)](https://pepy.tech/project/deepspeed)
-->
## Latest News
* [2022/03/21] [Supporting efficient large model training on AMD Instinct GPUs with DeepSpeed](https://cloudblogs.microsoft.com/opensource/2022/03/21/supporting-efficient-large-model-training-on-amd-instinct-gpus-with-deepspeed/)
* [2022/03/07] [Maximizing Communication Efficiency for Large-scale Training via 0/1 Adam](https://www.deepspeed.ai/tutorials/zero-one-adam/)
* [2022/01/19] [DeepSpeed: Advancing MoE inference and training to power next-generation AI scale](https://www.microsoft.com/en-us/research/blog/deepspeed-advancing-moe-inference-and-training-to-power-next-generation-ai-scale/)
* [Mixture of Experts (MoE) for NLG tutorial](https://www.deepspeed.ai/tutorials/mixture-of-experts-nlg/).
* [Mixture of Experts (MoE) Inference tutorial](https://www.deepspeed.ai/tutorials/moe-inference-tutorial).
* [2021/11/15] [Autotuning: Automatically discover the optimal DeepSpeed configuration that delivers good training speed](https://www.deepspeed.ai/news/2021/11/15/autotuning.html)
* [2021/10/11] [Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, the World’s Largest and Most Powerful Generative Language Model](https://www.microsoft.com/en-us/research/blog/using-deepspeed-and-megatron-to-train-megatron-turing-nlg-530b-the-worlds-largest-and-most-powerful-generative-language-model/)
* Read more on how to [train large models with DeepSpeed](https://www.deepspeed.ai/tutorials/large-models-w-deepspeed/)
### DeepSpeed is hiring, [come join us!](https://careers.microsoft.com/us/en/search-results?keywords=http:%2F%2Fdeepspeed.ai)
---
[DeepSpeed](https://www.deepspeed.ai/) is a deep learning optimization [DeepSpeed](https://www.deepspeed.ai/) is a deep learning optimization
library that makes distributed training easy, efficient, and effective. library that makes distributed training easy, efficient, and effective.
...@@ -14,10 +33,10 @@ library that makes distributed training easy, efficient, and effective. ...@@ -14,10 +33,10 @@ library that makes distributed training easy, efficient, and effective.
<p align="center"><i><b>Minimal Code Change</b></i></p> <p align="center"><i><b>Minimal Code Change</b></i></p>
DeepSpeed delivers extreme-scale model training for everyone, from data scientists training on massive supercomputers to those training on low-end clusters or even on a single GPU: DeepSpeed delivers extreme-scale model training for everyone, from data scientists training on massive supercomputers to those training on low-end clusters or even on a single GPU:
* Extreme scale: Using current generation of GPU clusters with hundreds of devices, 3D parallelism of DeepSpeed can efficiently train deep learning models with trillions of parameters. * Extreme scale: Using current generation of GPU clusters with hundreds of devices, 3D parallelism of DeepSpeed can efficiently train deep learning models with trillions of parameters.
* Extremely memory efficient: With just a single GPU, ZeRO-Offload of DeepSpeed can train models with over 10B parameters, 10x bigger than the state of arts, democratizing multi-billion-parameter model training such that many deep learning scientists can explore bigger and better models. * Extremely memory efficient: With just a single GPU, ZeRO-Offload of DeepSpeed can train models with over 10B parameters, 10x bigger than the state of arts, democratizing multi-billion-parameter model training such that many deep learning scientists can explore bigger and better models.
* Extremely long sequence length: Sparse attention of DeepSpeed powers an order-of-magnitude longer input sequence and obtains up to 6x faster execution comparing with dense transformers. * Extremely long sequence length: Sparse attention of DeepSpeed powers an order-of-magnitude longer input sequence and obtains up to 6x faster execution comparing with dense transformers.
* Extremely communication efficient: 3D parallelism improves communication efficiency allows users to train multi-billion-parameter models 2–7x faster on clusters with limited network bandwidth. 1-bit Adam reduces communication volume by up to 5x while achieving similar convergence efficiency to Adam, allowing for scaling to different types of GPU clusters and networks. * Extremely communication efficient: 3D parallelism improves communication efficiency allows users to train multi-billion-parameter models 2–7x faster on clusters with limited network bandwidth. 1-bit Adam, 0/1 Adam and 1-bit LAMB reduce communication volume by up to 26x while achieving similar convergence efficiency to Adam/LAMB, allowing for scaling to different types of GPU clusters and networks.
Early adopters of DeepSpeed have already produced Early adopters of DeepSpeed have already produced
a language model (LM) with over 17B parameters called a language model (LM) with over 17B parameters called
...@@ -31,22 +50,6 @@ information [here](https://innovation.microsoft.com/en-us/exploring-ai-at-scale) ...@@ -31,22 +50,6 @@ information [here](https://innovation.microsoft.com/en-us/exploring-ai-at-scale)
**_For further documentation, tutorials, and technical deep-dives please see [deepspeed.ai](https://www.deepspeed.ai/)!_** **_For further documentation, tutorials, and technical deep-dives please see [deepspeed.ai](https://www.deepspeed.ai/)!_**
# News
* [2021/04/01] [[DeepSpeed on AzureML] Transformers and CIFAR examples are now available on AzureML GitHub](https://github.com/Azure/azureml-examples/tree/main/workflows/train/deepspeed)
* [2021/03/30] [[PyTorch Lightning Blog] Accessible Multi-Billion Parameter Model Training with PyTorch Lightning + DeepSpeed](https://medium.com/pytorch-lightning/accessible-multi-billion-parameter-model-training-with-pytorch-lightning-deepspeed-c9333ac3bb59)
* [2021/03/16] [1-bit Adam v2: NCCL-based implementation and more](https://www.deepspeed.ai/tutorials/onebit-adam/)
* [2021/03/08] [ZeRO-3 Offload: Scale your models to trillion parameters without code changes while leveraging both CPUs & GPUs](https://www.deepspeed.ai/news/2021/03/07/zero3-offload.html)
* [2021/01/19] [[🤗Hugging Face Blog] Fit More and Train Faster With ZeRO via DeepSpeed and FairScale](https://huggingface.co/blog/zero-deepspeed-fairscale)
* [2020/11/12] [Simplified install, JIT compiled ops, PyPI releases, and reduced dependencies](#installation)
* [2020/11/10] [Efficient and robust compressed training through progressive layer dropping](https://www.deepspeed.ai/news/2020/10/28/progressive-layer-dropping-news.html)
* [2020/09/10] [DeepSpeed v0.3: Extreme-scale model training for everyone](https://www.microsoft.com/en-us/research/blog/deepspeed-extreme-scale-model-training-for-everyone/)
* [Powering 10x longer sequences and 6x faster execution through DeepSpeed Sparse Attention](https://www.deepspeed.ai/news/2020/09/08/sparse-attention-news.html)
* [Training a trillion parameters with pipeline parallelism](https://www.deepspeed.ai/news/2020/09/08/pipeline-parallelism.html)
* [Up to 5x less communication and 3.4x faster training through 1-bit Adam](https://www.deepspeed.ai/news/2020/09/08/onebit-adam-news.html)
* [10x bigger model training on a single GPU with ZeRO-Offload](https://www.deepspeed.ai/news/2020/09/08/ZeRO-Offload.html)
# Table of Contents # Table of Contents
| Section | Description | | Section | Description |
| --------------------------------------- | ------------------------------------------- | | --------------------------------------- | ------------------------------------------- |
...@@ -96,6 +99,12 @@ If you would like to pre-install any of the DeepSpeed extensions/ops (instead ...@@ -96,6 +99,12 @@ If you would like to pre-install any of the DeepSpeed extensions/ops (instead
of JIT compiling) or install pre-compiled ops via PyPI please see our [advanced of JIT compiling) or install pre-compiled ops via PyPI please see our [advanced
installation instructions](https://www.deepspeed.ai/tutorials/advanced-install/). installation instructions](https://www.deepspeed.ai/tutorials/advanced-install/).
On Windows you can build wheel with following steps, currently only inference mode is supported.
1. Install pytorch, such as pytorch 1.8 + cuda 11.1
2. Install visual cpp build tools, such as VS2019 C++ x64/x86 build tools
3. Launch cmd console with Administrator privilege for creating required symlink folders
4. Run `python setup.py bdist_wheel` to build wheel in `dist` folder
# Features # Features
Below we provide a brief feature list, see our detailed [feature Below we provide a brief feature list, see our detailed [feature
overview](https://www.deepspeed.ai/features/) for descriptions and usage. overview](https://www.deepspeed.ai/features/) for descriptions and usage.
...@@ -116,14 +125,14 @@ overview](https://www.deepspeed.ai/features/) for descriptions and usage. ...@@ -116,14 +125,14 @@ overview](https://www.deepspeed.ai/features/) for descriptions and usage.
* [ZeRO-Offload](https://www.deepspeed.ai/tutorials/zero-offload/) * [ZeRO-Offload](https://www.deepspeed.ai/tutorials/zero-offload/)
* Leverage both CPU/GPU memory for model training * Leverage both CPU/GPU memory for model training
* Support 10B model training on a single GPU * Support 10B model training on a single GPU
* [Ultra-fast dense transformer kernels](https://www.deepspeed.ai/news/2020/05/18/bert-record.html) * [Ultra-fast dense transformer kernels](https://www.deepspeed.ai/2020/05/18/bert-record.html)
* [Sparse attention](https://www.deepspeed.ai/news/2020/09/08/sparse-attention.html) * [Sparse attention](https://www.deepspeed.ai/2020/09/08/sparse-attention-news.html)
* Memory- and compute-efficient sparse kernels * Memory- and compute-efficient sparse kernels
* Support 10x longer sequences than dense * Support 10x longer sequences than dense
* Flexible support to different sparse structures * Flexible support to different sparse structures
* [1-bit Adam](https://www.deepspeed.ai/news/2020/09/08/onebit-adam-blog-post.html) * [1-bit Adam](https://www.deepspeed.ai/2020/09/08/onebit-adam-blog-post.html), [0/1 Adam](https://www.deepspeed.ai/tutorials/zero-one-adam/) and [1-bit LAMB](https://www.deepspeed.ai/tutorials/onebit-lamb/)
* Custom communication collective * Custom communication collective
* Up to 5x communication volume saving * Up to 26x communication volume saving
* [Additional Memory and Bandwidth Optimizations](https://www.deepspeed.ai/features/#additional-memory-and-bandwidth-optimizations) * [Additional Memory and Bandwidth Optimizations](https://www.deepspeed.ai/features/#additional-memory-and-bandwidth-optimizations)
* Smart Gradient Accumulation * Smart Gradient Accumulation
* Communication/Computation Overlap * Communication/Computation Overlap
...@@ -142,8 +151,12 @@ overview](https://www.deepspeed.ai/features/) for descriptions and usage. ...@@ -142,8 +151,12 @@ overview](https://www.deepspeed.ai/features/) for descriptions and usage.
* Learning Rate Range Test * Learning Rate Range Test
* 1Cycle Learning Rate Schedule * 1Cycle Learning Rate Schedule
* [Simplified Data Loader](https://www.deepspeed.ai/features/#simplified-data-loader) * [Simplified Data Loader](https://www.deepspeed.ai/features/#simplified-data-loader)
* [Curriculum Learning](https://www.deepspeed.ai/tutorials/curriculum-learning/)
* A curriculum learning-based data pipeline that presents easier or simpler examples earlier during training
* Stable and 3.3x faster GPT-2 pre-training with 8x/4x larger batch size/learning rate while maintaining token-wise convergence speed
* Complementary to many other DeepSpeed features
* [Performance Analysis and Debugging](https://www.deepspeed.ai/features/#performance-analysis-and-debugging) * [Performance Analysis and Debugging](https://www.deepspeed.ai/features/#performance-analysis-and-debugging)
* [Mixture of Experts (MoE)](https://www.deepspeed.ai/tutorials/mixture-of-experts/)
# Further Reading # Further Reading
...@@ -154,14 +167,14 @@ All DeepSpeed documentation can be found on our website: [deepspeed.ai](https:// ...@@ -154,14 +167,14 @@ All DeepSpeed documentation can be found on our website: [deepspeed.ai](https://
| Article | Description | | Article | Description |
| ---------------------------------------------------------------------------------------------- | -------------------------------------------- | | ---------------------------------------------------------------------------------------------- | -------------------------------------------- |
| [DeepSpeed Features](https://www.deepspeed.ai/features/) | DeepSpeed features | | [DeepSpeed Features](https://www.deepspeed.ai/features/) | DeepSpeed features |
| [Getting Started](https://www.deepspeed.ai/getting-started/) | First steps with DeepSpeed | | [Getting Started](https://www.deepspeed.ai/getting-started/) | First steps with DeepSpeed |
| [DeepSpeed JSON Configuration](https://www.deepspeed.ai/docs/config-json/) | Configuring DeepSpeed | | [DeepSpeed JSON Configuration](https://www.deepspeed.ai/docs/config-json/) | Configuring DeepSpeed |
| [API Documentation](https://deepspeed.readthedocs.io/en/latest/) | Generated DeepSpeed API documentation | | [API Documentation](https://deepspeed.readthedocs.io/en/latest/) | Generated DeepSpeed API documentation |
| [CIFAR-10 Tutorial](https://www.deepspeed.ai/tutorials/cifar-10) | Getting started with CIFAR-10 and DeepSpeed | | [CIFAR-10 Tutorial](https://www.deepspeed.ai/tutorials/cifar-10) | Getting started with CIFAR-10 and DeepSpeed |
| [Megatron-LM Tutorial](https://www.deepspeed.ai/tutorials/megatron/) | Train GPT2 with DeepSpeed and Megatron-LM | | [Megatron-LM Tutorial](https://www.deepspeed.ai/tutorials/megatron/) | Train GPT2 with DeepSpeed and Megatron-LM |
| [BERT Pre-training Tutorial](https://www.deepspeed.ai/tutorials/bert-pretraining/) | Pre-train BERT with DeepSpeed | | [BERT Pre-training Tutorial](https://www.deepspeed.ai/tutorials/bert-pretraining/) | Pre-train BERT with DeepSpeed |
| [Learning Rate Range Test Tutorial](https://www.deepspeed.ai/tutorials/lrrt/) | Faster training with large learning rates | | [Learning Rate Range Test Tutorial](https://www.deepspeed.ai/tutorials/lrrt/) | Faster training with large learning rates |
| [1Cycle Tutorial](https://www.deepspeed.ai/tutorials/1Cycle/) | SOTA learning schedule in DeepSpeed | | [1Cycle Tutorial](https://www.deepspeed.ai/tutorials/one-cycle/) | SOTA learning schedule in DeepSpeed |
...@@ -192,7 +205,12 @@ Conduct](https://opensource.microsoft.com/codeofconduct/). For more information ...@@ -192,7 +205,12 @@ Conduct](https://opensource.microsoft.com/codeofconduct/). For more information
2. Jeff Rasley, Samyam Rajbhandari, Olatunji Ruwase, and Yuxiong He. (2020) DeepSpeed: System Optimizations Enable Training Deep Learning Models with Over 100 Billion Parameters. [In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (KDD '20, Tutorial)](https://dl.acm.org/doi/10.1145/3394486.3406703). 2. Jeff Rasley, Samyam Rajbhandari, Olatunji Ruwase, and Yuxiong He. (2020) DeepSpeed: System Optimizations Enable Training Deep Learning Models with Over 100 Billion Parameters. [In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (KDD '20, Tutorial)](https://dl.acm.org/doi/10.1145/3394486.3406703).
3. Minjia Zhang, Yuxiong He. (2020) Accelerating Training of Transformer-Based Language Models with Progressive Layer Dropping. [arXiv:2010.13369](https://arxiv.org/abs/2010.13369) and [NeurIPS 2020](https://proceedings.neurips.cc/paper/2020/hash/a1140a3d0df1c81e24ae954d935e8926-Abstract.html). 3. Minjia Zhang, Yuxiong He. (2020) Accelerating Training of Transformer-Based Language Models with Progressive Layer Dropping. [arXiv:2010.13369](https://arxiv.org/abs/2010.13369) and [NeurIPS 2020](https://proceedings.neurips.cc/paper/2020/hash/a1140a3d0df1c81e24ae954d935e8926-Abstract.html).
4. Jie Ren, Samyam Rajbhandari, Reza Yazdani Aminabadi, Olatunji Ruwase, Shuangyan Yang, Minjia Zhang, Dong Li, Yuxiong He. (2021) ZeRO-Offload: Democratizing Billion-Scale Model Training. [arXiv:2101.06840](https://arxiv.org/abs/2101.06840). 4. Jie Ren, Samyam Rajbhandari, Reza Yazdani Aminabadi, Olatunji Ruwase, Shuangyan Yang, Minjia Zhang, Dong Li, Yuxiong He. (2021) ZeRO-Offload: Democratizing Billion-Scale Model Training. [arXiv:2101.06840](https://arxiv.org/abs/2101.06840).
5. Hanlin Tang, Shaoduo Gan, Ammar Ahmad Awan, Samyam Rajbhandari, Conglong Li, Xiangru Lian, Ji Liu, Ce Zhang, Yuxiong He. (2021) 1-bit Adam: Communication Efficient Large-Scale Training with Adam's Convergence Speed. [arXiv:2102.02888](https://arxiv.org/abs/2102.02888). 5. Hanlin Tang, Shaoduo Gan, Ammar Ahmad Awan, Samyam Rajbhandari, Conglong Li, Xiangru Lian, Ji Liu, Ce Zhang, Yuxiong He. (2021) 1-bit Adam: Communication Efficient Large-Scale Training with Adam's Convergence Speed. [arXiv:2102.02888](https://arxiv.org/abs/2102.02888) and [ICML 2021](http://proceedings.mlr.press/v139/tang21a.html).
6. Samyam Rajbhandari, Olatunji Ruwase, Jeff Rasley, Shaden Smith, Yuxiong He. (2021) ZeRO-Infinity: Breaking the GPU Memory Wall for Extreme Scale Deep Learning. [arXiv:2104.07857](https://arxiv.org/abs/2104.07857).
7. Conglong Li, Ammar Ahmad Awan, Hanlin Tang, Samyam Rajbhandari, Yuxiong He. (2021) 1-bit LAMB: Communication Efficient Large-Scale Large-Batch Training with LAMB's Convergence Speed. [arXiv:2104.06069](https://arxiv.org/abs/2104.06069).
8. Conglong Li, Minjia Zhang, Yuxiong He. (2021) Curriculum Learning: A Regularization Method for Efficient and Stable Billion-Scale GPT Model Pre-Training. [arXiv:2108.06084](https://arxiv.org/abs/2108.06084).
9. Yucheng Lu, Conglong Li, Minjia Zhang, Christopher De Sa, Yuxiong He. (2022) Maximizing Communication Efficiency for Large-scale Training via 0/1 Adam. [arXiv:2202.06009](https://arxiv.org/abs/2202.06009).
10. Samyam Rajbhandari, Conglong Li, Zhewei Yao, Minjia Zhang, Reza Yazdani Aminabadi, Ammar Ahmad Awan, Jeff Rasley, Yuxiong He. (2022) DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale [arXiv:2201.05596](https://arxiv.org/abs/2201.05596).
# Videos # Videos
1. DeepSpeed KDD 2020 Tutorial 1. DeepSpeed KDD 2020 Tutorial
...@@ -206,3 +224,6 @@ Conduct](https://opensource.microsoft.com/codeofconduct/). For more information ...@@ -206,3 +224,6 @@ Conduct](https://opensource.microsoft.com/codeofconduct/). For more information
* Registration is free and all videos are available on-demand. * Registration is free and all videos are available on-demand.
* [ZeRO & Fastest BERT: Increasing the scale and speed of deep learning training in DeepSpeed](https://note.microsoft.com/MSR-Webinar-DeepSpeed-Registration-On-Demand.html). * [ZeRO & Fastest BERT: Increasing the scale and speed of deep learning training in DeepSpeed](https://note.microsoft.com/MSR-Webinar-DeepSpeed-Registration-On-Demand.html).
3. [DeepSpeed on AzureML](https://youtu.be/yBVXR8G8Bg8) 3. [DeepSpeed on AzureML](https://youtu.be/yBVXR8G8Bg8)
4. Community Tutorials
* [DeepSpeed: All the tricks to scale to gigantic models](https://www.youtube.com/watch?v=pDGI668pNg0)
* [Turing-NLG, DeepSpeed and the ZeRO optimizer](https://www.youtube.com/watch?v=tC01FRB0M7w)
SECURITY.md
View file @ 7d1a83a9
<!-- BEGIN MICROSOFT SECURITY.MD V0.0.3 BLOCK --> <!-- BEGIN MICROSOFT SECURITY.MD V0.0.3 BLOCK -->
## Security ## Security
Microsoft takes the security of our software products and services seriously, which includes all source code repositories managed through our GitHub organizations, which include [Microsoft](https://github.com/Microsoft), [Azure](https://github.com/Azure), [DotNet](https://github.com/dotnet), [AspNet](https://github.com/aspnet), [Xamarin](https://github.com/xamarin), and [our GitHub organizations](https://opensource.microsoft.com/). Microsoft takes the security of our software products and services seriously, which includes all source code repositories managed through our GitHub organizations, which include [Microsoft](https://github.com/Microsoft), [Azure](https://github.com/Azure), [DotNet](https://github.com/dotnet), [AspNet](https://github.com/aspnet), [Xamarin](https://github.com/xamarin), and [our GitHub organizations](https://opensource.microsoft.com/).
If you believe you have found a security vulnerability in any Microsoft-owned repository that meets Microsoft's [Microsoft's definition of a security vulnerability](https://docs.microsoft.com/en-us/previous-versions/tn-archive/cc751383(v=technet.10)) of a security vulnerability, please report it to us as described below. If you believe you have found a security vulnerability in any Microsoft-owned repository that meets Microsoft's [Microsoft's definition of a security vulnerability](https://docs.microsoft.com/en-us/previous-versions/tn-archive/cc751383(v=technet.10)) of a security vulnerability, please report it to us as described below.
## Reporting Security Issues ## Reporting Security Issues
**Please do not report security vulnerabilities through public GitHub issues.** **Please do not report security vulnerabilities through public GitHub issues.**
Instead, please report them to the Microsoft Security Response Center (MSRC) at [https://msrc.microsoft.com/create-report](https://msrc.microsoft.com/create-report). Instead, please report them to the Microsoft Security Response Center (MSRC) at [https://msrc.microsoft.com/create-report](https://msrc.microsoft.com/create-report).
If you prefer to submit without logging in, send email to [secure@microsoft.com](mailto:secure@microsoft.com). If possible, encrypt your message with our PGP key; please download it from the the [Microsoft Security Response Center PGP Key page](https://www.microsoft.com/en-us/msrc/pgp-key-msrc). If you prefer to submit without logging in, send email to [secure@microsoft.com](mailto:secure@microsoft.com). If possible, encrypt your message with our PGP key; please download it from the the [Microsoft Security Response Center PGP Key page](https://www.microsoft.com/en-us/msrc/pgp-key-msrc).
You should receive a response within 24 hours. If for some reason you do not, please follow up via email to ensure we received your original message. Additional information can be found at [microsoft.com/msrc](https://www.microsoft.com/msrc). You should receive a response within 24 hours. If for some reason you do not, please follow up via email to ensure we received your original message. Additional information can be found at [microsoft.com/msrc](https://www.microsoft.com/msrc).
Please include the requested information listed below (as much as you can provide) to help us better understand the nature and scope of the possible issue: Please include the requested information listed below (as much as you can provide) to help us better understand the nature and scope of the possible issue:
* Type of issue (e.g. buffer overflow, SQL injection, cross-site scripting, etc.) * Type of issue (e.g. buffer overflow, SQL injection, cross-site scripting, etc.)
* Full paths of source file(s) related to the manifestation of the issue * Full paths of source file(s) related to the manifestation of the issue
* The location of the affected source code (tag/branch/commit or direct URL) * The location of the affected source code (tag/branch/commit or direct URL)
* Any special configuration required to reproduce the issue * Any special configuration required to reproduce the issue
* Step-by-step instructions to reproduce the issue * Step-by-step instructions to reproduce the issue
* Proof-of-concept or exploit code (if possible) * Proof-of-concept or exploit code (if possible)
* Impact of the issue, including how an attacker might exploit the issue * Impact of the issue, including how an attacker might exploit the issue
This information will help us triage your report more quickly. This information will help us triage your report more quickly.
If you are reporting for a bug bounty, more complete reports can contribute to a higher bounty award. Please visit our [Microsoft Bug Bounty Program](https://microsoft.com/msrc/bounty) page for more details about our active programs. If you are reporting for a bug bounty, more complete reports can contribute to a higher bounty award. Please visit our [Microsoft Bug Bounty Program](https://microsoft.com/msrc/bounty) page for more details about our active programs.
## Preferred Languages ## Preferred Languages
We prefer all communications to be in English. We prefer all communications to be in English.
## Policy ## Policy
Microsoft follows the principle of [Coordinated Vulnerability Disclosure](https://www.microsoft.com/en-us/msrc/cvd). Microsoft follows the principle of [Coordinated Vulnerability Disclosure](https://www.microsoft.com/en-us/msrc/cvd).
<!-- END MICROSOFT SECURITY.MD BLOCK --> <!-- END MICROSOFT SECURITY.MD BLOCK -->
azure/attach.sh 100755 → 100644
View file @ 7d1a83a9
File mode changed from 100755 to 100644
azure/azure_ssh.sh 100755 → 100644
View file @ 7d1a83a9
File mode changed from 100755 to 100644
azure/build_docker_image.sh 100755 → 100644
View file @ 7d1a83a9
File mode changed from 100755 to 100644
azure/create_vms.sh 100755 → 100644
View file @ 7d1a83a9
File mode changed from 100755 to 100644
azure/setup_docker.sh 100755 → 100644
View file @ 7d1a83a9
File mode changed from 100755 to 100644
azure/setup_vms.sh 100755 → 100644
View file @ 7d1a83a9
File mode changed from 100755 to 100644
azure/shutdown_vms.sh 100755 → 100644
View file @ 7d1a83a9
File mode changed from 100755 to 100644
azure/start_container.sh 100755 → 100644
View file @ 7d1a83a9
File mode changed from 100755 to 100644
bin/deepspeed deleted 120000 → 0
View file @ ab5534fc
ds
\ No newline at end of file
bin/deepspeed 0 → 100644
View file @ 7d1a83a9
#!/usr/bin/env python3
from deepspeed.launcher.runner import main
if __name__ == '__main__':
main()
bin/deepspeed.pt deleted 120000 → 0
View file @ ab5534fc
ds
\ No newline at end of file
bin/deepspeed.pt 0 → 100644
View file @ 7d1a83a9
#!/usr/bin/env python3
from deepspeed.launcher.runner import main
if __name__ == '__main__':
main()
bin/ds 100755 → 100644
View file @ 7d1a83a9
#!/usr/bin/env python #!/usr/bin/env python3
from deepspeed.launcher.runner import main from deepspeed.launcher.runner import main
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment