"WizardLM 13B Uncensored" conversion to GGML

[pugzly]

I just noticed "WizardLM 13B Uncensored" appeared in the wild
https://huggingface.co/ehartford/WizardLM-13B-Uncensored

But I'm still trying to work out the correct process of conversion for "pytorch_model.bin" in to GGML
So I figured I'll check with guys around, if somebody here already done it and has all the right steps at hand?

(while I continue reading through all docs and experiment)

EDIT:
Thanks to Geen-SKY, it was as simple as:
python3 convert.py models/WizardLM-13B-Uncensored/pytorch_model.bin
and
./quantize ./models/WizardLM-13B-Uncensored/pytorch_model.bin ./models/WizardLM-13B-Uncensored/ggml-model-q4_0.bin q4_0

[Green-Sky]

Simply run the convert.py pointing to the directory you put the files.
then run the qunatize executable for the desired quantization.
done.

[Green-Sky]

more details in the readme https://github.com/ggerganov/llama.cpp#prepare-data--run

[pugzly]

ah yes, of course, it had to be something painfully simple..
I didn't realize I had to provide full path to bin file for conversion. Thanks a lot!

For anyone else passing by, I only needed run the following:
python3 convert.py models/WizardLM-13B-Uncensored/pytorch_model.bin
./quantize ./models/WizardLM-13B-Uncensored/pytorch_model.bin ./models/WizardLM-13B-Uncensored/ggml-model-q4_0.bin q4_0

[Green-Sky]

using just python3 convert.py models/WizardLM-13B-Uncensored/ was not enough?

[pugzly]

no, it was throwing me error, that there was no model in that directory, until I gave it full path to the file