davidmezzetti

🤖 Agent Team 🤖

This example shows how to build an Agent having multiple agents as tools. This "Agent Team" works together to answer questions.

A more basic example can be found here

Install

pip install git+https://github.com/neuml/txtai

davidmezzetti

🚀 Agents with txtai

We're thrilled to share a preview version of txtai agents. Inspired by the simplicity of frameworks like OpenAI Swarm, txtai agents are built on top of the Transformers Agent framework. This supports all LLMs txtai supports (Hugging Face, llama.cpp, OpenAI + Claude + AWS Bedrock via LiteLLM).

The following example shows how to create an agent with txtai. Agents will be available in the upcoming txtai 8.0 release (available now in the txtai GitHub repo now - follow #804 - feedback welcome).

Install

davidmezzetti

from txtai import Embeddings

embeddings = Embeddings(
  method="model2vec",
  path="minishlab/M2V_base_output"
)

embeddings.index(["test"])
embeddings.search("test")
# [(0, 1.0)]

davidmezzetti

from txtai import Embeddings
embeddings = Embeddings(
  method="sentence-transformers",
  vectors={
    "backend": "onnx", 
    "model_kwargs": {"file_name": "model_qint8_avx512.onnx"}
  }
)
embeddings.index(["test"])
embeddings.search("test")

davidmezzetti

from txtai import Embeddings, RAG

# Load embeddings index
embeddings = Embeddings()
embeddings.load(provider="huggingface-hub", container="neuml/txtai-wikipedia")

template = """
Answer the following question using the provided context.

Question:

davidmezzetti

import re

from txtai import Embeddings, LLM

# Prompt courtesy of the following link: https://github.com/codelion/optillm/blob/main/cot_reflection.py
def cot(system, user):
    system = f"""
        {system}

        You are an AI assistant that uses a Chain of Thought (CoT) approach with reflection to answer queries. Follow these steps:

davidmezzetti

import os

from txtai import Embeddings, RAG

# For demo only. Set via environment.
os.environ["AWS_ACCESS_KEY_ID"] = ""
os.environ["AWS_SECRET_ACCESS_KEY"] = ""
os.environ["AWS_REGION_NAME"] = ""
os.environ["ANN_URL"] = "postgresql+psycopg2://postgres:[email protected]/postgres"
os.environ["CLIENT_URL"] = "postgresql+psycopg2://postgres:[email protected]/postgres"

davidmezzetti

import json
import re

import yfinance as yf

from txtai import Embeddings
from txtai.pipeline import Textractor

def djia():
    """

davidmezzetti

YAML-driven Embeddings Service with Docker

Did you know that most of the functionality in txtai can be run with configuration? That's right, txtai can dynamically load Embeddings, LLM, RAG and other pipelines with YAML configuration.

Check out this example that loads an Embeddings database via Docker with a couple of lines of YAML config. The example then runs a graph search via the API and plots the results with Sigma.js.

Start Docker

config/config.yml

davidmezzetti

# pip install yfinance txtai

import yfinance as yf
import txtai

def prompt(ticker):
  return f"""
Answer the following question using only the context below. Only include information
specifically discussed.