Build a Chatbot by Fine-Tuning Llama 2
Use Llama 2 and supervised fine tuning (SFT) to build a Chatbot. Use a pre-trained Large Language Model (LLM) in Domino and fine-tune the model for a specific task. We will convert this model using CTranslate2 to optimize its throughput and deploy it as a model API and app in Domino.
Fine-tuning a pre-trained LLM is a commonly used technique for solving NLP problems with machine learning. This is a typical transfer learning task where the final model is realized through a number of training phases.