Guides

Learn about Cohere's UI in the playground overview. Read Intro to Large Language Models with Cohere for a brief visual overview to Large Language Models (LLMs) and some of their applications.

Text Classification

Text classification is one of the most useful applications of LLMs. The text classification with Embeddings article guides you through building a sentiment analysis text classifier if you have labeled data.

LLMs can also classify text using a small number of examples (few-shot classification). For these cases, see the question classification and sentimant analysis articles. These two articles use two different modes of the Choose Best endpoint.

Text Generation

LLMs can write coherent text like no other human technology before them could. This can be used for creative copy, but also for summarization and paraphrasing. We tune the inputs using prompt engineering techniques that get the model to produce useful outputs. Important text generations parameters include top-k and top-p.

Semantic Search

Learn how to use embeddings to build semantic search capabilities.

Finetuning

Customize Cohere models to fit your use case by finetuning our baseline models with your own data. Learn about finetuning generation models in addition to finetuning representation models.

Model Evaluation

Learn how the Likelihood endpoint can be a useful tool for model evaluation.

Concepts

Learn some of the key concepts involved in language generation and understanding.

• Tokens are words or parts of words that our models take as input or produce as output.
• Embeddings are lists of numbers that represent a word, token, or an entire piece of text. Embeddings capture information about the meaning and context of the words or sentences they represent.
• Temperature is a value that controls the outputs of a generation model by tuning the degree of randomness involved in picking output tokens.
• Likelihood is a measure of how “expected” each token is in a piece of text.