llama.cpp
2763dc8b - ggml-quants : handle zero amax for MXFP4

Commit

64 days ago

ggml-quants : handle zero amax for MXFP4

References

compilade/gguf-py-mxfp4

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

Author

compilade

compilade

Parents

Loading