PR #8838 gguf-py : simplify support for quant types

gguf-py : use classes for quants

compilade committed 1 year ago

convert_hf : simplify internal quantization type selection

compilade committed 1 year ago

gguf-py : fix flake8 lint

compilade committed 1 year ago

gguf-py : fix BF16 numpy view type

compilade committed 1 year ago

gguf-py : remove LlamaFileTypeMap

compilade committed 1 year ago

llama.cpp gguf-py : simplify support for quant types #8838 Merged