llama.cpp
gguf-py : add Numpy MXFP4 de/quantization support
#15111

Merged

gguf-py : add Numpy MXFP4 de/quantization support #15111

compilade merged 2 commits into master from compilade/gguf-py-mxfp4

gguf-py : add MXFP4 de/quantization support

141cab13

compilade added python

compilade added Tensor Encoding Scheme

compilade commented on 2025-08-06

CISC approved these changes on 2025-08-06

ggml-quants : handle zero amax for MXFP4

2763dc8b

github-actions added ggml

slaren approved these changes on 2025-08-07

compilade merged e54d41be into master 34 days ago

Reviewers

slaren

CISC

Assignees

No one assigned

Labels

python ggml Tensor Encoding Scheme

Milestone

No milestone