LLM chatbot and writing assistant using HuggingFace and the Matrix protocol.
🌟TorchON is an adaptable retrieval augmented application that provides question answering over documents, GitHub repositories, and websites. It takes data, creates synthetic data, and uses that synthetic data to optimize the prompts of the Torchon application. The application recompiles itself every run in a unique and adapted way to the user query. You can then decide to publish this application locally, privately, or share it with selected people.
