Task description
A task description is a specification of the operation or objective to be performed, including relevant parameters, constraints, and intended outcome.
Data Analysis with LLMs
A textual specification of the operation that a model should perform.
- How it is used
- Directs a language model to perform an operation such as classifying a review, translating a question into SQL, or extracting properties.
- How deep it goes
- Shows task instructions in worked prompts and Python templates, without explaining model internals.
- Examples
- Uses laptop and movie-review classification, SQL translation, and property extraction from source text as examples.
- What you can do after
- Apply it via Python prompt templates that combine task instructions with a question, review, or source text.
Read how Data Analysis with LLMs explains task description →
Building LLM Applications with DSPy
A prompt component that states the overall goal the language model should accomplish.
- How it is used
- States the overall goal a language model should accomplish within a prompt.
- How deep it goes
- Defines the component and notes that its length varies with task complexity; leaves wording and ordering to task-specific confirmation.
- Examples
- Uses customer-message classification as a task goal.
- What you can do after
- Evaluate/understand the task description as the prompt component that states the overall goal.
Read how Building LLM Applications with DSPy explains task description →
Related concepts
- promptpart of · An input, typically containing instructions or context, supplied to an artificial-intelligence model to guide or elicit its output.
- templatepart of · A template is a reusable structure with fixed elements and variable parts for creating or generating similar content, outputs, or artifacts.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.