DeepSeek-FlashMLA

---

Reposity forkato manualmente ( con VsCode ) dal reposity deepSeek

ovvero ho copiato manulamente il reposity con VsCode - Insiders

---

FlashMLA

FlashMLA is an efficient MLA decoding kernel for Hopper GPUs, optimized for variable-length sequences serving.

Currently released:

• BF16, FP16
• Paged kvcache with block size of 64

Quick start

Install

python setup.py install

Benchmark

python tests/test_flash_mla.py

Achieving up to 3000 GB/s in memory-bound configuration and 580 TFLOPS in computation-bound configuration on H800 SXM5, using CUDA 12.8.

Usage

from flash_mla import get_mla_metadata, flash_mla_with_kvcache

tile_scheduler_metadata, num_splits = get_mla_metadata(cache_seqlens, s_q * h_q // h_kv, h_kv)

for i in range(num_layers):
    ...
    o_i, lse_i = flash_mla_with_kvcache(
        q_i, kvcache_i, block_table, cache_seqlens, dv,
        tile_scheduler_metadata, num_splits, causal=True,
    )
    ...

Requirements

• Hopper GPUs
• CUDA 12.3 and above
  • But we highly recommend 12.8 or above for the best performance
• PyTorch 2.0 and above

Acknowledgement

FlashMLA is inspired by FlashAttention 2&3 and cutlass projects.

Community Support

MetaX

For MetaX GPUs, visit the official website: MetaX.

The corresponding FlashMLA version can be found at: MetaX-MACA/FlashMLA

Moore Threads

For the Moore Threads GPU, visit the official website: Moore Threads.

The corresponding FlashMLA version is available on GitHub: MooreThreads/MT-flashMLA.

Hygon DCU

For the Hygon DCU, visit the official website: Hygon Developer.

The corresponding FlashMLA version is available here: OpenDAS/MLAttention.

Intellifusion

For the Intellifusion NNP, visit the official website: Intellifusion.

The corresponding FlashMLA version is available on Gitee: Intellifusion/tyllm.

Iluvatar Corex

For Iluvatar Corex GPUs, visit the official website: Iluvatar Corex.

The corresponding FlashMLA version is available on GitHub: Deep-Spark/FlashMLA

AMD Instinct

For AMD Instinct GPUs, visit the official website: AMD Instinct.

The corresponding FlashMLA version can be found at: AITER/MLA

Citation

@misc{flashmla2025,
      title={FlashMLA: Efficient MLA decoding kernels},
      author={Jiashi Li},
      year={2025},
      publisher = {GitHub},
      howpublished = {\url{https://github.com/deepseek-ai/FlashMLA}},
}

---

IDE in cui modificarlo :

• Visual Studio Code - insiders

• Visual studio 2022

• Visual Studio code

  • il Fork di VsCode ovvero Cursor
  • il Fork di VsCode ovvero VSCodium

• PyCharm

• Sublime Text

• Notepad++

• Atom

• Eclipse

• IntelliJ IDEA

• NetBeans

• Android Studio

• Xcode( solo per Mac )

• WebStorm

  • PhpStorm
  • DataGrip
  • RubyMine

• AppCode

• CLion

• DataGrip

• Rider

• GoLand

• PyCharm Edu( solo per Linux , non cambia nulla se è su un'altro OS )

• Brackets

• Vim( solo per Linux , ma anche per il OS di Apple [ Mac os ] )

• Emacs

• nano

• Geany

• Bluefish

• Kate

• Code::Blocks

• Anjuta

• KDevelop

• Lazarus

• MonoDevelop

• CodeLite

• JCreator

• DrJava

• Notepadqq

• Gedit

• Kate

• KWrite ( solo ed escusivamente su Linux )

• SciTE

• Kile

• WinEdt

• LyX

• TeXShop

  • TeXworks

  • TeXstudio

• TeXnicCenter

  • TeXmaker

    • TeXpen

• TeXlipse

  • TeXmacs

  • TeXShade

    • TeX2pag(e)

    • TeX2RTF

      • TeX2HTML ( in questo caso tutti gli TeX saranno in ritardo di 1-8 giorni )