Uncensored AI Agent using llama api

Questions
1

Intro

I have a feeling what I’m trying to do is not possible, but hey - I might learn something in the process.

Objective

I want to create an AI Agent that is a) uncensored and b) is compatible with HTTP tools. The use case is a HR agent that returns private information about users (DOB/Drivers license, etc.) from a variety of sources (i.e. tools)

Describe the problem/error/question

I want to use an uncensored llama model (e.g. llama2-uncensored) as the chat model. I am using llama API to serve the endpoints. For the life of me, I cannot get the thing to connect. Here’s what I’ve tried so far:

  1. Configure the llama chat model using the base URL https://api.llama-api.com
  2. Configure the OpenAI model pointing at the base URL https://api.llama-api.com

I have been successful in running tinyllama in a basic LLM Chain, but there’s two issues here:
a) Tinyllama does not support tools in a AI Agent
b) My VPS isn’t that grunty and I’d rather just buy credits for a server with a GPU (i.e. https://llama-api.com)

I can also successfully run the HTTP tool using native chat models using OpenAI and Anthropic… but it censors the results, so doesn’t fit for my use case.

What is the error message (if any)?

  1. Problem in node ‘AI Agent‘ / Cannot read properties of undefined (reading ‘content’)

Please share your workflow

Information on your n8n setup

  • **n8n version:1.60.1
  • Database (default: SQLite):
  • **Running n8n via (Docker, npm, n8n cloud, desktop app):docker via easypanel
  • **Operating system:Linux Ubuntu
1 Like
4

Hi @Christo ,

Does this happen with other API endpoints as well or just this one?

Could you also share the full error message? :slight_smile:

5

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.