GitHub - andrewgcodes/lightspeedGPT: Use GPT4 and GPT3.5 on inputs of unlimited size. Uses multithreading to process multiple chunks in parallel. Useful for tasks like Named Entity Recognition, information extraction on large books, datasets, etc.

GitHub - andrewgcodes/lightspeedGPT: Use GPT4 and GPT3.5 on inputs of unlimited size. Uses multithreading to process multiple chunks in parallel. Useful for tasks like Named Entity Recognition, information extraction on large books, datasets, etc.

🚀 Discover lightspeedGPT - a powerful tool leveraging GPT4 and GPT3.5 for tasks like Named Entity Recognition & info extraction on large texts! 🤖📚 Process unlimited input chunks simultaneously with multithreading. Perfect for AI enthusiasts and data wizards! 🌟 #AI #tools #github

  • lightspeedGPT is a tool that uses multithreading to process multiple chunks of input in parallel, enabling tasks like Named Entity Recognition and information extraction on large texts.
  • It splits input into manageable pieces to avoid OpenAI token limits, sending each chunk separately for processing simultaneously.
  • The script retries with exponential backoff if OpenAI rate limit is reached, giving up after three failures.
  • Prerequisites include Python 3.6+, an OpenAI API key, and basic command-line interface knowledge.
  • Usage involves setting the model, chunk size, tokens, temperature, and prompt via command-line arguments.
  • An example command: `python main.py -i input.txt -o output.txt -l log.txt -m 'gpt-3.5-turbo' -c 500 -t 200 -v 0.5 -p 'Translate English to French:'`
  • Inspired by infiniteGPT, lightspeedGPT is designed for processing large text bodies efficiently using OpenAI models.
  • Repository by andrewgcodes intends to facilitate tasks like summarization, translation, and information extraction from extensive inputs.