Hey everyone,
I’m looking for some help on where to start with writing an ai that is able to have small-talk and is able to have conversations about random topics,
so it should not be specialized on a single topic and more like a assistant you can have random conversations with (similar to something like Jarvis).
The problem i have is that i don’t rly know how and where to start.
If anyone could help me with that it would be greatly appreciated.
Have you been through the huggingface tutorials and guides yet?
i read through the NLP Course
This is a full featured chat application that uses multiple different models from huggingface. GitHub - huggingface/chat-ui: Open source codebase powering the HuggingChat app
At the other end is something as simple as using a pipeline
like so
from transformers import pipeline, Conversation
converse = pipeline("conversational")
conversation = Conversation("how are you")
converse([conversation])
See
ah ty, im gonna try some stuff out with that,
btw, is it possible to somehow train the ai to answer or talk in a certain way, so basically giving it personality trades?
To start, I recommend using a pre-trained language model like GPT-3 via OpenAI’s API or Hugging Face’s transformers for natural conversations. Use Python and libraries like spaCy or NLTK for NLP tasks. Begin by setting up a basic conversational loop and refine it based on test interactions. Once comfortable, you can expand functionality with personalized responses and context tracking.