llm-sandbox

OpenAI:

These scripts require a .env file with the following variables:

• MODEL: The name of the model to use, e.g. gpt-4
• API_KEY: Your OpenAI API key.
• OUTPUT_FOLDER: The folder to store output documents in. The scripts save YAML files with chat histories.
• PROMPT_FOLDER: oasb.py can read prompts from files in this directory.

OpenAI Scripts

• oasb.py: Simple question-and-answer session with ChatGPT--Sends one prompt at a time, with no history.
• oasbcc.py: Allows you set a system prompt, then have a conversation with ChatGPT.
• sentiment.py: Simply performs sentiment analysis on any entered prompt.
• dalle/dalle_app.py: A simple chatbot interface to generate images using DALL-E.

Kagi

Scripts to interact with Kagi's FastGPT and Summarizer APIs. These have their own requirements.txt file and need their own .env with the following variables:

• BASE_URL: Should be https://kagi.com/api/v0/fastgpt
• SUMMARIZER_URL: https://kagi.com/api/v0/summarize
• API_TOKEN: Your Kagi API key.
• DEFAULT_SUMMARIZER: The default summarizer "persona" to use, e.g. cecil
• SUMMARY_TYPE: Default summary type. Can be summary for prose, or takeaway for a list of key points.
• OUTPUT_FOLDER: Folder for output YAML files.

Kagi Scripts

• sb1.py: Simple interface to the FastGPT API.
• summarizer.py: Simple interface to the Summarizer API.