pamelafox

import json
import os
import requests
from datetime import datetime
import urllib.parse
from bs4 import BeautifulSoup
from rich import print
import openai
from pydantic import BaseModel
from azure.identity import AzureDeveloperCliCredential, get_bearer_token_provider

pamelafox

import os

import bs4
import requests
import rich
from openai import OpenAI
from pydantic import BaseModel

client = OpenAI(
    base_url="https://models.inference.ai.azure.com",

pamelafox

# print python version
import sys
print(sys.version)
import math

if True:
    wow = f"""this is a multiline f string with triple quotes {math.pi:0.2f} : 
1. Line 1
2. Line 3 {math.pi:0.2f}"""

pamelafox

Want the links in these tabs??

Get this gist from https://aka.ms/msbooth-pytorch

• https://learn.microsoft.com/en-us/training/paths/pytorch-fundamentals/
• https://github.com/microsoft/Generative-AI-For-Beginners
• https://learn.microsoft.com/en-us/azure/ai-studio/concepts/fine-tuning-overview
• https://learn.microsoft.com/en-us/azure/developer/python/

pamelafox

import os
import json

from dotenv import load_dotenv
from openai import AzureOpenAI
import requests

load_dotenv(".env", override=True)

pamelafox

-- Hybrid search with real time translation. Combining pgvector, postgis, azure_local_ai, azure_ai, Azure Language Services
WITH listings_cte AS (
    SELECT l.listing_id, name, listing_location, summary FROM listings l
    INNER JOIN calendar c ON l.listing_id = c.listing_id
    WHERE ST_DWithin(
        listing_location::geography,
        ST_GeographyFromText('POINT(-122.349358 47.620422)')
        ,3218 )
    AND c.date = '2016-01-13'
    AND c.available = 't'

pamelafox

import os

import psycopg2
import openai
import azure.identity
import numpy as np
from dotenv import load_dotenv
from pgvector.psycopg2 import register_vector

load_dotenv()

pamelafox

This is a comparison of calls made to either OpenAI gpt-35-turbo or Ollama Llama3.1 with a function calling request. The geneal goal of the function is to take a user query and turn it into a search query, along with optional price filter or brand filter. See the file below this one for the tools definition.

User question: "Best shoe for hiking"

gpt-35: 😊

{"search_query":"Best shoe for hiking"}

llama3.1: 😊

pamelafox

param location string
param aiHubName string
param aiProjectName string
param modelName string
param modelVersion string
param deploymentName string
param tags object = {}
param publicNetworkAccess string

param aiServicesName string

pamelafox

Evaluation:

• ai-rag-chat-evaluator: Repo with scripts to generate sample data based on Azure AI search index and evaluate responses from a chat app using GPT-4 model and custom metrics. Designed for use with azure-search-openai-demo but can be modified for use with other chat apps.
  • 📺 Quick video overview
  • 📺 Talk: Automated evaluations of LLM apps: Talk for Data Science Day. Slides also available.
  • 📝 Blog post: Evaluating a RAG Chat App - Approach, SDKs, Tools
• [📝 Blog post: Can your app say "I don't know?"](https://blog.pamelafox.org/2024/03/evaluating-rag-chat-apps-can-y