How to use this model in Local H2O installation

#2
by late7 - opened

Hi,
I'm quite new to h2o. Yet, I like to use this model on H2O running on my local PC. H2O runs file with models packaged and delivered as .bin files. I simply can't find instructions how to use model that are not by default downloadable in .bin format.
So, question is it possible to use this model in local H2O installation? If, so all advice highly appreciated.

If you want to use the models in Huggingface then you need to do it as following:
check src/prompter.py for --prompt_type

use ---base_model= what you get from huggingface
`` python
python generate.py --base_model='h2oai/h2ogpt-gm-oasst1-en-2048-falcon-40b-v2' --prompt_type=prompt_answer --cli=True --langchain_mode=UserData

python
python generate.py --base_model='JosephusCheung/Guanaco' --prompt_type=instruct_simple --cli=True --langchain_mode=UserData --user_path=./SOURCE_DOCUMENT
``

This comment has been hidden

Sign up or log in to comment