Machine Learning (ML) is the process of training a model to perform a task. ML involves providing a dataset of examples of the task, and then training a model on that dataset. The model is then able to perform the task.
Large Language Models (LLMs) are advanced natural language processing models designed to understand and generate human-like text. Trained on massive amounts of data, LLMs capture intricate language patterns and structures, allowing them to perform various tasks, such as text summarization, translation, question-answering and even holding a conversation. Examples of LLMs include GPT-4, Cohere, Bard, and PaLM and more. These models leverage the power of deep learning and the Transformer architecture to achieve state-of-the-art performance in many language-related tasks. By fine-tuning LLMs on specific domains or tasks, they can be adapted to a wide range of applications, making them a valuable resource in the field of artificial intelligence and natural language processing.
A prompt is the primary way of interacting with AI systems. It is the input that the bot uses as the basis for it's response. Usually a prompt takes the form of a piece of text, but it can technically be any form of digitized data.
Prompt Engineering is the process of creating a prompt that will generate a desired response from a bot. Prompt engineering involves 2 step cycle:
- Compile a new prompt
- Analyze the response to the prompt
Fine tuning is the process of updating an existing bot with new data. Fine tuning a bot involves providing a dataset of text, and then fine tuning the bot on that dataset.
Embeddings are a set of vectors that represent a piece of text. Embeddings are used to represent a piece of text in a way that is easier for a model to understand.