The post contains a request for the "links you shared", and the comments discuss using GPT4ALL, a language model. Community members provide suggestions on how to use GPT4ALL, including loading the model and tokenizer, and mention a GitHub repository with a notebook for integrating GPT4ALL. However, there is no explicitly marked answer in the comments.
If you don't have access to the pipeline, you'll have to load the model and tokenizsr yourself, tokenize the text, call the model, and return the newly generated text
So if you have a model that you can already test giving inputs to, you likely have this all written already somewhere
