facebookresearch

facebookresearch/any4

Quantize transformers to any learned arbitrary 4-bit numeric format

10 stars 0 forks 0 watchers Python

2 Open Issues Need Help Last updated: Jul 5, 2025

Open Issues Need Help

View All on GitHub

TODOs 12 days ago

good first issue

facebookresearch/any4

10

Quantize transformers to any learned arbitrary 4-bit numeric format

Python

Create Modules for NF4, FP4, MX4 about 2 months ago

enhancement good first issue

facebookresearch/any4

10

Quantize transformers to any learned arbitrary 4-bit numeric format

Python