Use prompts to get completions from models
Once the model is deployed, you can test how it completes prompts. A prompt is the text portion of a request that is sent to the deployed model's completions endpoint. Responses are referred to as completions, which can come in form of text, code, or other formats.
Prompt types
Prompts can be grouped into types of requests based on task.
| Task type | Prompt example | Completion example |
|---|---|---|
| Classifying content | Tweet: I enjoyed the trip. Sentiment: |
Positive |
| Generating new content | List ways of traveling | 1. Bike 2. Car ... |
| Holding a conversation | A friendly AI assistant | See examples |
| Transformation (translation and symbol conversion) | English: Hello French: |
bonjour |
| Summarizing content | Provide a summary of the content {text} |
The content shares methods of machine learning. |
| Picking up where you left off | One way to grow tomatoes | is to plant seeds. |
| Giving factual responses | How many moons does Earth have? | One |
Completion quality
Several factors affect the quality of completions you'll get from a generative AI solution.
- The way a prompt is engineered. Learn more about prompt engineering here.
- The model parameters (covered next)
- The data the model is trained on, which can be adapted through model fine-tuning with customization
You have more control over the completions returned by training a custom model than through prompt engineering and parameter adjustment.
Making calls
You can start making calls to your deployed model via the REST API, Python, C#, or from the Studio. If your deployed model has a GPT-3.5 or GPT-4 model base, use the Chat completions documentation, which has different request endpoints and variables required than for other base models.