From 58ef4afa49cc0691ae637307a49cbd66e3370117 Mon Sep 17 00:00:00 2001 From: sakell Date: Mon, 16 Sep 2024 09:52:16 +0300 Subject: [PATCH] Added Intel Gaudi to Accelerator Setup Guide --- docs/_tutorials/accelerator-setup-guide.md | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/docs/_tutorials/accelerator-setup-guide.md b/docs/_tutorials/accelerator-setup-guide.md index 36d33f31b5b3..ffe91626e37b 100644 --- a/docs/_tutorials/accelerator-setup-guide.md +++ b/docs/_tutorials/accelerator-setup-guide.md @@ -9,6 +9,7 @@ tags: getting-started - [Intel Architecture (IA) CPU](#intel-architecture-ia-cpu) - [Intel XPU](#intel-xpu) - [Huawei Ascend NPU](#huawei-ascend-npu) +- [Intel Gaudi](#intel-gaudi) # Introduction DeepSpeed supports different accelerators from different companies. Setup steps to run DeepSpeed on certain accelerators might be different. This guide allows user to lookup setup instructions for the accelerator family and hardware they are using. @@ -245,3 +246,10 @@ accelerator: npu ## Multi-card parallel training using Huawei Ascend NPU To perform model training across multiple Huawei Ascend NPU cards using DeepSpeed, see the examples provided in [DeepSpeed Examples](https://github.com/microsoft/DeepSpeedExamples/blob/master/training/cifar/cifar10_deepspeed.py). + +# Intel Gaudi +PyTorch models can be run on Intel® Gaudi® AI accelerator using DeepSpeed. Refer to the following user guides to start using DeepSpeed with Intel Gaudi: +* [Getting Started with DeepSpeed](https://docs.habana.ai/en/latest/PyTorch/DeepSpeed/Getting_Started_with_DeepSpeed/Getting_Started_with_DeepSpeed.html#getting-started-with-deepspeed) +* [DeepSpeed User Guide for Training](https://docs.habana.ai/en/latest/PyTorch/DeepSpeed/DeepSpeed_User_Guide/DeepSpeed_User_Guide.html#deepspeed-user-guide) +* [Optimizing Large Language Models](https://docs.habana.ai/en/latest/PyTorch/DeepSpeed/Optimizing_LLM.html#llms-opt) +* [Inference Using DeepSpeed](https://docs.habana.ai/en/latest/PyTorch/DeepSpeed/Inference_Using_DeepSpeed.html#deepspeed-inference-user-guide)