Themata.AI
Follow coverage
Themata.AI
DeepSeek v4Reinforcement learningOpen source AIHybrid architectures

Report · lmsys.org

DeepSeek-V4 on Day 0: From Fast Inference to Verified RL with SGLang and Miles

DeepSeek-V4 on Day 0: From Fast Inference to Verified RL with SGLang and Miles - LMSYS Blog

lmsys.org

Source published April 25, 2026

Summary: 1 min read

🔥🔥🔥🔥🔥

48/100

Summary

DeepSeek-V4 is now supported for both inference and reinforcement learning (RL) training from Day 0. SGLang and Miles provide the first open-source stack designed for DeepSeek-V4’s hybrid sparse-attention architecture and manifold-constrained hyper-connections, utilizing FP4 expert weights.

Key Takeaways

  • DeepSeek-V4 features Day-0 support for both inference and reinforcement learning training, utilizing SGLang and Miles as the first open-source stack for its launch.
  • The model incorporates hybrid sparse-attention, manifold-constrained hyper-connections (mHC), and FP4 expert weights for enhanced performance and efficiency.
  • DeepSeek-V4 supports a 1M-token context window by combining sliding window attention with two compression mechanisms, allowing for efficient processing of large inputs.
  • The ShadowRadix caching mechanism enables coherent management of multiple heterogeneous key-value pools, optimizing the handling of hybrid attention across layers.
Read original article

Coverage & discussions

Hacker News · lmsys.org

DeepSeek-V4 on Day 0: From Fast Inference to Verified RL with SGLang and Miles ↗

Community discussions

Hacker News discussion ↗

Historical count: 1 · 1 saved in sample · Historical sample

Awaiting analysis

Related Articles

deepseek-ai/DeepSeek-V4.1-Flash-1017 · Hugging Face

DeepSeek-v4.1-Flash-1017

Oct 17, 2026

GitHub - ryanzhou/deepseek-v4-flash-mi300x

DeepSeek V4 Flash on a Single AMD MI300X

Aug 4, 2026

Bringing up DeepSeek-V4-Flash on AMD MI300X

Bringing Up DeepSeek-V4-Flash on AMD MI300X

Jun 2, 2026

GitHub - deepseek-ai/DeepInfer: DeepInfer: High-performance, large-scale inference framework for DeepSeek models

DeepSeek just open-sourced its inference engine

Oct 20, 2026

DeepSeek V4—almost on the frontier, a fraction of the price

DeepSeek V4–almost on the frontier, a fraction of the price

May 1, 2026