DeepSpeed (@DeepSpeedAI) / X

DeepSpeed

113 posts

DeepSpeed

@DeepSpeedAI

Official account for DeepSpeed, a library that enables unprecedented scale and speed for deep learning training + inference. 日本語 : @DeepSpeedAI_JP

Joined May 2020

Replies Articles Media

Pinned
DeepSpeed
@DeepSpeedAI
2h
Apologies for the short notice! We're holding two DeepSpeed developer meetings on Aug. 7 to discuss our Q3 roadmap goals: 9:00 a.m. PT — DeepCompile efficiency & robustness 10:00 a.m. PT — Expert Parallelism enhancements Roadmap: github.com/deepspeedai/De… Join Slack:
github.com
DeepSpeed - Q3 Roadmap · Issue #8104 · deepspeedai/DeepSpeed
This is the roadmap for DeepSpeed Q3 2026. Feedback is welcome — please leave comments on this issue or join the #2026q3-roadmap channel on the DeepSpeed Slack. Some of the items are the extension ...
DeepSpeed
@DeepSpeedAI
Jul 22
We will host our monthly Office Hour for DeepSpeed users on July 28th, 9AM PT. Feel free to join and discuss any topics ranging from issues to roadmaps. Zoom link: zoom-lfx.platform.linuxfoundation.org/meeting/939025…
DeepSpeed
@DeepSpeedAI
Jun 3
We now have native support for all ZeRO stages 1/2/3 for Muon Optimizers, providing superior performance on LLM pre-training and post-training. Feel free to try it out, kudos to @PKUWZP Guokai Ma, Peng Du and Chi for the contribution!
PyTorch
@PyTorch
Jun 3
DeepSpeed now supports the Muon Optimizer. Optimized specifically for internal 2D weights within neural networks, Muon is gaining traction for its significant memory savings and strong convergence metrics during LLM training. In our latest blog post, the DeepSpeed team shares a
DeepSpeed
@DeepSpeedAI
May 26
Reminder to join our DeepSpeed Office Hours on Tuesday, May 26 at 12:00 PM America/New_York. We'll cover general questions, Q2 roadmap progress, and requests for Q3. Everyone is welcome! Zoom: zoom-lfx.platform.linuxfoundation.org/meeting/939025…
DeepSpeed
@DeepSpeedAI
May 18
Article
AMD’s MORI now integrated into DeepSpeed with SDMA support
MORI is AMD's next-generation infrastructure for performance-critical AI workloads: Wide-EP, common communication kernels, end-to-end KVCache management and agentic system. ~10% training speedup by...