social-gpt 0.0.10

Social gpt

Social GPT

This package helps you to build embeddings based on someone social profile. It will scrape the data from the social media and build the embeddings based on the data. The embeddings can be used for further analysis. Once scraped, you can query it with the use of openai.

Installation

Install using following command: pipenv install social-gpt

Setup

Update the example.env and change it's name to .env. Add relevant information in the .env file.

Creating embeddings

As of now, only youtube is supported. We will be bringing more social media platforms soon. To create embeddings, run the following command:

from social_gpt.ingestion.ingestion import SocialIngestion
ingestion = SocialIngestion(channel_id, 'youtube')
index_id = ingestion.create_embeddings()

Store this index_id for querying the embeddings.

Querying embeddings

from social_gpt.query.query_helper import QueryHelper
q = QueryHelper(index_id)
response = q.query_embeddings('What is the best way to learn python?')
print(response)

That's it!