How to further quantize GGUF to Q4 format using llama.cpp?

[jasonsu123]

Dear all,

I am using a Windows environment.
Currently, I can successfully convert Safetensors files from Hugging Face into Q8 GGUF format using the following example:

python convert_hf_to_gguf.py D:\Ollama\TAIDE-LX-8B-Chat-Alpha1 --outfile D:\Ollama\TAIDE-LX-8B-Chat-Alpha1-q8_0.gguf --outtype q8_0

However, how can I quantize the model into Q1~Q7 or different versions like K_M and K_S?
I tried modifying the query based on the above example,
python convert_hf_to_gguf.py D:\Ollama\TAIDE-LX-8B-Chat-Alpha1 --outfile D:\Ollama\TAIDE-LX-8B-Chat-Alpha1-q4_0.gguf --outtype q4_0

but I encountered the following error:

convert_hf_to_gguf.py: error: argument --outtype: invalid choice: 'q4_0' (choose from 'f32', 'f16', 'bf16', 'q8_0', 'tq1_0', 'tq2_0', 'auto')

it seems like the ./quantize command is required for this,
./quantize ./models/Bailong-instruct-7B-f16.gguf ./models/Bailong-instruct-7B-v0.1-Q5_K_M.gguf q5_k_m

but since I am a Windows user, how should I modify the query to work in my environment?

Thank you!