Azure OpenAI Serivce

Azure OpenAI Service enables users to build enterprise-grade solutions with OpenAI models. With Azure OpenAI, users can summarize text, get code suggestions, generate images for a web site, and more. 

There are following categories of capabilities found in OpenAI models:

  • Generating natural language:
    • Text completion: generating and editing text, for example, summarizing complex text for different reading levels, suggesting alternative wording for sentences.
    • Embeddings: searching, classifying, and comparing text.
  • Generating code: translating code from one programming language into another, identifying and troubleshooting bugs in code, and more.
  • Generating images: generating images for publications from text descriptions and more.

GPT models for natural language generation

Generative pre-trained transformer (GPT) models are excellent at both understanding and creating natural language. Generative AI takes an input, or prompt, to return a natural language, visual, or code response. GPT tries to infer, or guess, the context of the user’s question based on the prompt.

GPT models are great at completing several natural language tasks, like summarizing text, classifying text, generating names of phrases, answering questions and suggesting content.

Codex models

Codex models are based off of GPT-3 and are optimized to understand and write code. These models have been trained on both natural language and billions of lines of code from public repositories. Codex is able to generate code from natural language instructions such as code comments, and can suggest ways to complete code functions.

This technology is used in GitHub Copilot which integrates the power of OpenAI Codex into a plugin for developer environments like Visual Studio Code.

DALL-E models

DELL-E (currently in preview by invitation, Apr 17, 2023) can help with following scnarios:

  • Image generation by providing text prompt with image description and style.
  • Editing an image, like removing or replacing objects.
  • Creating image variations

Reference Materials

%d bloggers like this: