PrototypeCommon
JT-Ushio/MLAfication
Bringing Multi-Head Latent Attention to pretrained LLMs and VLMs: MHA2MLA (ACL 2025), MHA2MLA-VLM (AAAI 2026), and MLAfication.
- Owner
- JT-Ushio (personal account)
- Stars
- Stars: 215
- Forks
- Forks: 26
- Language
- Python
- License
- Apache-2.0
- Created
- 2024
- Topics
- economical-key-value-cacheefficient-attention-architectures
- Tier
- Prototype (100 to 249 ★) What this tier means
Nearby in Python
All Python repos- reflection-oss/mirrorStars: 215PythonPrototype
- christos-antono/NPCTalkStars: 215PythonPrototype
- jouaaaaaa/mcd-self-butlerStars: 215PythonPrototype
- dziegler/django-cssStars: 215PythonPrototype
- Freeseer/freeseerStars: 215PythonPrototype
- adityarathod/pyandocStars: 215PythonPrototype