PrototypeCommon
microsoft/RetrievalAttention
[VLDB 26, NeurIPS 25] Scalable long-context LLM decoding that leverages sparsity—by treating the KV cache as a vector storage system.
- Owner
- microsoft (organization)
- Stars
- Stars: 153
- Forks
- Forks: 36
- Language
- Python
- License
- MIT
- Created
- 2024
- Tier
- Prototype (100 to 249 ★) What this tier means
More repos by microsoft
All 1,923 repos by microsoft- microsoft/vscodeStars: 194,462TypeScriptLTS
- microsoft/markitdownStars: 193,456PythonLTS
- microsoft/PowerToysStars: 140,398CLTS
- microsoft/generative-ai-for-beginnersStars: 122,788Jupyter NotebookLTS
- microsoft/TypeScriptStars: 111,714GoLTS
- microsoft/terminalStars: 105,311C++LTS
Nearby in Python
All Python repos- bertel2020/HA-HousekeeperStars: 153PythonPrototype
- silvermoong/stocking-texture-toolStars: 153PythonPrototype
- semrincaballero/Polymarket-Prediction-BotStars: 153PythonPrototype
- heynemann/pynqStars: 153PythonPrototype
- deepthawtz/fakerStars: 153PythonPrototype
- softlayer/softlayer-pythonStars: 153PythonPrototype