Automatically Record QueryBot Calls with PromptRecorder

In this tutorial, we will learn how to use the PromptRecorder class to automatically record calls made to the QueryBot. The PromptRecorder class is designed to record prompts and responses, making it a perfect fit for logging interactions with the QueryBot.

Prerequisites

Before we begin, make sure you have the following Python libraries installed:

• pandas
• panel

You can install them using pip:

pip install pandas panel

Step 1: Import the necessary classes

First, we need to import the PromptRecorder and QueryBot classes from their respective source files. You can do this by adding the following lines at the beginning of your script:

from llamabot.recorder import PromptRecorder, autorecord
from llamabot.bot.querybot import QueryBot

Step 2: Initialize the QueryBot

Next, we need to create an instance of the QueryBot class. You can do this by providing the necessary parameters, such as the system message, model name, and document paths. For example:

system_message = "You are a helpful assistant that can answer questions based on the provided documents."
model_name = "gpt-4"
doc_paths = ["document1.txt", "document2.txt"]

query_bot = QueryBot(system_message, model_name=model_name, document_paths=doc_paths)

Step 3: Use the PromptRecorder context manager

Now that we have an instance of the QueryBot, we can use the PromptRecorder context manager to automatically record the prompts and responses. To do this, simply wrap your interactions with the QueryBot inside a with statement, like this:

with PromptRecorder() as recorder:
    # Interact with the QueryBot here

Step 4: Interact with the QueryBot

Inside the with statement, you can now interact with the QueryBot by calling it with your queries. For example:

with PromptRecorder() as recorder:
    query = "What is the main idea of document1?"
    response = query_bot(query)
    print(response.content)

    query = "How does document2 support the main idea?"
    response = query_bot(query)
    print(response.content)

The PromptRecorder will automatically record the prompts and responses for each interaction with the QueryBot.

Step 5: Access the recorded data

After you have finished interacting with the QueryBot, you can access the recorded data using the PromptRecorder instance. For example, you can print the recorded data as a pandas DataFrame:

print(recorder.dataframe())

Or, you can display the recorded data as an interactive panel:

recorder.panel().show()

Complete Example

Here's the complete example that demonstrates how to use the PromptRecorder to automatically record QueryBot calls:

from llamabot.recorder import PromptRecorder, autorecord
from llamabot.bot.querybot import QueryBot

system_message = "You are a helpful assistant that can answer questions based on the provided documents."
model_name = "gpt-4"
doc_paths = ["document1.txt", "document2.txt"]

query_bot = QueryBot(system_message, model_name=model_name, document_paths=doc_paths)

with PromptRecorder() as recorder:
    query = "What is the main idea of document1?"
    response = query_bot(query)
    print(response.content)

    query = "How does document2 support the main idea?"
    response = query_bot(query)
    print(response.content)

print(recorder.dataframe())
recorder.panel().show()

That's it! You now know how to use the PromptRecorder class to automatically record calls made to the QueryBot. This can be a useful tool for logging and analyzing interactions with your chatbot.