DeepSpeed: Accelerating large-scale model inference and training
![DeepSpeed: Accelerating large-scale model inference and training](https://www.microsoft.com/en-us/research/uploads/prod/2021/05/1400x788_deepspeed_no_logo_still-1-scaled.jpg)
miro.medium.com/v2/resize:fit:1400/0*7l0yGZjkm3dyx
![](https://miro.medium.com/v2/resize:fit:1400/1*DafLIAEn1yQAxOSWmEb2YA.png)
miro.medium.com/v2/resize:fit:1400/1*DafLIAEn1yQAx
![](https://ar5iv.labs.arxiv.org/html/2201.05596/assets/x8.png)
2201.05596] DeepSpeed-MoE: Advancing Mixture-of-Experts Inference and Training to Power Next-Generation AI Scale
![](https://economics.illinois.edu/sites/default/files/2023-02/Econometrics%20lab%20logo_0.png)
Machine Learning and Inference Laboratory - Photos from Conferences, thomas mitchell machine learning
![](https://preview.redd.it/67d33gbsw7871.jpg?vthumb=1&s=59b98a202a06ad02ba18a17426c48e177b5f3dc5)
N] Improvement on model's inference from DeepSpeed team. [D] How
GitHub - Naagar/a_paper_a_day: I am trying a new initiative - a paper a day. This repository will hold all those papers and related summaries and notes.
![](https://user-images.githubusercontent.com/58739961/187154444-fce76639-ac8d-429b-9354-c6fac64b7ef8.jpg)
GitHub - microsoft/DeepSpeed: DeepSpeed is a deep learning
![](https://pbs.twimg.com/profile_images/1384358208257413123/mtW_tqvr_400x400.jpg)
Samyam Rajbhandari (@samyamrb) / X
![](https://cdn.pathfactory.com/assets/10412/contents/155796/thumbnails/600x/cwe00000-connect-with-experts_4x3.jpg)
Toward INT8 Inference: Deploying Quantization-Aware Trained
![](https://arxiv.org/html/2402.16363v3/x15.png)
LLM Inference Unveiled: Survey and Roofline Model Insights
![](https://www.microsoft.com/en-us/research/uploads/prod/2023/06/DeepSpeedZero-BlogHeroFeature-1400x788-1-1024x576.png)
DeepSpeed ZeRO++: A leap in speed for LLM and chat model training
![](https://d3i71xaburhd42.cloudfront.net/d1a6b3a5efde3783b53f822dc8dd00aaac934b95/2-Figure1-1.png)
SpecInfer: Accelerating Generative LLM Serving with Speculative