ContextualOpenAI

If you want to make your OpenAI generate context-based responses and copy the necessary files in this repository, follow these simple steps

1. Navigate to the directory where you want to copy the files and enable context-based responses
2. Copy and paste the following script: (For linux/unix users)

  git clone --depth 1 --filter=blob:none --sparse https://github.com/gan0412/ContextualOpenAI.git temp_ContextualOpenAI
  cd temp_ContextualOpenAI
  git sparse-checkout set context.py chistory.py
  mv context.py chistory.py ..
  cd ..
  rm -rf temp_ContextualOpenAI
  echo "Files have been copied to the current directory."

(For powershell users)

  git clone --depth 1 --filter=blob:none --sparse https://github.com/gan0412/ContextualOpenAI.git temp_ContextualOpenAI
  Set-Location temp_ContextualOpenAI
  git sparse-checkout set context.py chistory.py
  Move-Item context.py -Destination ..
  Move-Item chistory.py -Destination ..
  Set-Location ..
  Remove-Item temp_ContextualOpenAI -Recurse -Force
  Write-Host "Files have been copied to the current directory."

---

How to Use context.py Functions for Context-Based OpenAI Responses

The core functionality for enabling context-based responses is provided by two functions:

• contextualize_prompt(prompt: str, max_tokens_limit: int) -> List[Message]
• clear_history()

1. Generating Contextual Responses

Use contextualize_prompt to prepare your prompt with relevant conversation history for the OpenAI API. The parameters are:

• user_input: the client's current prompt
• max_tokens_limit: maximum token length of the context window

Use Case: contextualize_prompt should be called every time you want to generate a context-aware response.

Example Usage:

from context import contextualize_prompt

# Your current user prompt
user_input = "Tell me more about the last topic we discussed."

# Set a max token limit for the context window (e.g., 800)
context_messages = contextualize_prompt(user_input, max_tokens_limit=800)

# Send this message list to OpenAI's chat/completions endpoint
# For example, with openai-python:
import openai

response = openai.ChatCompletion.create(
    model="gpt-3.5-turbo",
    messages=context_messages
)

print(response['choices'][0]['message']['content'])

• The function automatically loads previous user messages from history.json, manages the token window, and attaches relevance scores to historical messages.

2. Clearing Conversation History

Use clear_history to reset or clear the stored conversation history (e.g., at the start of a new session).

Example Usage:

from chistory import clear_history

# This will clear history.json and start fresh
clear_history()

---

Notes

• Only user messages are persisted; system and assistant messages are handled internally for context.
• The token limit ensures that the total context fits within your OpenAI model's window.
• Make sure to have the required dependencies installed (sentence-transformers, tiktoken, scikit-learn).