# microsoft/MInference

<https://repo.masters-masters.com/en/repos/microsoft/MInference>

[Home](https://repo.masters-masters.com/en) › [microsoft](https://repo.masters-masters.com/en/owners/microsoft) › [MInference](https://repo.masters-masters.com/en/repos/microsoft/MInference)

microsoft/MInference has 1,231 stars on GitHub. Its RepoMasters card: Beta tier. \[NeurIPS'24 Spotlight, ICLR'25, ICML'25\] To speed up Long-context LLMs'…

| Field | Value |
| --- | --- |
| Owner | [microsoft](https://repo.masters-masters.com/en/owners/microsoft) |
| Stars | 1,231 |
| Forks | 84 |
| Language | [Python](https://repo.masters-masters.com/en/languages/python) |
| License | MIT |
| Created | 2024 |
| Tier | [Beta (700 to 1,999 ★)](https://repo.masters-masters.com/en/ranks/box_office) |
| View on GitHub | [microsoft/MInference](https://github.com/microsoft/MInference) |

## Repository

\[NeurIPS'24 Spotlight, ICLR'25, ICML'25\] To speed up Long-context LLMs' inference, approximate and dynamic sparse calculate the attention, which reduces inference latency by up to 10x for pre-filling on an A100 while maintaining accuracy.

The tier for repos with 700 to 1,999 ★. Published odds per card pulled: 6%. Rarity: Rare.

## More repos by microsoft

- [microsoft/vscode · ★ 194,462 · TypeScript · LTS](https://repo.masters-masters.com/en/repos/microsoft/vscode)
- [microsoft/markitdown · ★ 193,456 · Python · LTS](https://repo.masters-masters.com/en/repos/microsoft/markitdown)
- [microsoft/PowerToys · ★ 140,398 · C · LTS](https://repo.masters-masters.com/en/repos/microsoft/PowerToys)
- [microsoft/generative-ai-for-beginners · ★ 122,788 · Jupyter Notebook · LTS](https://repo.masters-masters.com/en/repos/microsoft/generative-ai-for-beginners)
- [microsoft/TypeScript · ★ 111,714 · Go · LTS](https://repo.masters-masters.com/en/repos/microsoft/TypeScript)
- [microsoft/terminal · ★ 105,311 · C++ · LTS](https://repo.masters-masters.com/en/repos/microsoft/terminal)

## Nearby in Python

- [taranis-ai/taranis-ai · ★ 1,231 · Python · Beta](https://repo.masters-masters.com/en/repos/taranis-ai/taranis-ai)
- [LifeArchiveProject/HistoryHub · ★ 1,231 · Python · Beta](https://repo.masters-masters.com/en/repos/LifeArchiveProject/HistoryHub)
- [refreshdotdev/web-eval-agent · ★ 1,231 · Python · Beta](https://repo.masters-masters.com/en/repos/refreshdotdev/web-eval-agent)
- [asciidisco/plugin.video.netflix · ★ 1,231 · Python · Beta](https://repo.masters-masters.com/en/repos/asciidisco/plugin.video.netflix)
- [python-adaptive/adaptive · ★ 1,231 · Python · Beta](https://repo.masters-masters.com/en/repos/python-adaptive/adaptive)
- [qubvel/classification\_models · ★ 1,231 · Python · Beta](https://repo.masters-masters.com/en/repos/qubvel/classification_models)

---
Data: GitHub · Illustration: the project’s public image, pixelated by RepoMasters.