PrototypeCommon
facebookresearch/diffq
DiffQ performs differentiable quantization using pseudo quantization noise. It can automatically tune the number of bits used per weight or group of weights, in order to achieve a given trade-off between model size and accuracy.
Archived: this repository is read-only
- Owner
- facebookresearch (organization)
- Stars
- Stars: 238
- Forks
- Forks: 17
- Language
- Python
- License
- None
- Created
- 2021
- Tier
- Prototype (100 to 249 ★) What this tier means
More repos by facebookresearch
All 640 repos by facebookresearch- facebookresearch/segment-anythingStars: 55,084Jupyter NotebookLTS
- facebookresearch/faissStars: 41,242C++LTS
- facebookresearch/detectron2Stars: 34,958PythonLTS
- facebookresearch/fairseqStars: 32,212PythonLTS
- facebookresearch/fastTextStars: 26,513HTMLLTS
- facebookresearch/DetectronStars: 26,345PythonLTS
Nearby in Python
All Python repos- RoboProbe/RoboProbeStars: 238PythonPrototype
- Yzmblog/DMADStars: 238PythonPrototype
- NoizAI/WorldSonusStars: 238PythonPrototype
- bikalims/bika.lims.legacyStars: 238PythonPrototype
- jazzband/django-mongonautStars: 238PythonPrototype
- charettes/django-mutantStars: 238PythonPrototype