Illustration drawn by RepoMasters from the repo’s ID and its language color. Credits
PrototypeCommon

itsnamgyu/block-transformer

Block Transformer: Global-to-Local Language Modeling for Fast Inference (NeurIPS 2024)

Owner
itsnamgyu (personal account)
Stars
Stars: 166
Forks
Forks: 9
Language
Python
License
MIT
Created
2024
Topics
kv-cachekv-cache-compressionllmllm-architecturellm-inference
Tier
Prototype (100 to 249 ★) What this tier means
View on GitHub

More repos by itsnamgyu

All 2 repos by itsnamgyu

Nearby in Python

All Python repos