LLM Toolkit for AI Applications - Mirascope

---

---

AI Nuggets

Mirascope Installation and Usage Instructions

Installation

1. Install Mirascope with default settings (includes OpenAI library):

   pip install mirascope  

2. Install Mirascope with additional models (Anthropic, GRO, LogFire, etc.):
   • For specific models:

     pip install mirascope[anthropic]  
     pip install mirascope[gro]  
     pip install mirascope[logfire]  

   • To install all extras at once:

     pip install mirascope[all]  

Setting Up API Key

1. Set your OpenAI API key in your environment. Obtain the API key from platform.openai.com (requires a paid account).

Writing Function Calls with Mirascope

1. Import necessary modules, such as literal from typing.
2. Create a wrapper for OpenAI function calls using Mirascope.
3. Define functions for your specific use case, such as get_current_weather.
4. Define a Python class to encapsulate the tool use with OpenAI.
5. Instantiate the class and call the desired function.

Running the Code

1. Ensure that the OpenAI API key is set and Mirascope is installed.
2. Navigate to the directory containing your script.
3. Run the script to see the output of the function calls.

Example Code Snippet

from typing import Literal  
  
# Define a wrapper for OpenAI function calls  
# ...  
  
# Define a function for a specific use case  
def get_current_weather(location: Literal['Tokyo', 'San Francisco', 'Paris']):  
    # Function implementation  
    # ...  
  
# Define a class to encapsulate tool use with OpenAI  
class MyApplication:  
    # Class implementation  
    # ...  
  
# Instantiate the class and call the function  
app = MyApplication()  
app.some_method()  

Running the Function

python my_script.py  

Additional Information

• The GitHub repository link for Mirascope is provided in the video’s description.
• The video demonstrates the ease of using Mirascope for function calling with OpenAI.
• Mirascope is an evolving project, and more model support is expected to be added.

Video Description

• The video description contains the link to the Mirascope GitHub repository.
• The video encourages subscribing to the channel and sharing the content.

For more detailed instructions and examples, please refer to the Mirascope GitHub repository provided in the video’s description.