convert: add dequant function for compressed_tensor (kimi-k2-thinking)#17064

ngxson wants to merge 6 commits intoggml-org:masterggml-org/llama.cpp:masterfrom

ngxson:xsn/convert_kimi_k2_quantngxson/llama.cpp:xsn/convert_kimi_k2_quantCopy head branch name to clipboard