GrammarErrorCorrectionTemplate

class torchtune.data.GrammarErrorCorrectionTemplate

Prompt template for grammar correction datasets.

Correct this to standard English: <YOUR SENTENCE HERE>
---
Corrected:

classmethod format(sample: Mapping[str, Any], column_map: Optional[Dict[str, str]] = None) → str

Generate prompt from sentence.

Parameters:

• sample (Mapping[str, Any]) – a single data sample with sentence

• column_map (Optional[Dict[str, str]]) – a mapping from the expected placeholder names in the template to the column names in the sample. If None, assume these are identical.

Examples

>>> # Simple sentence
>>> GrammarErrorCorrectionTemplate.format(sample={"sentence": "The quik brown fox jumps the lazy dog"})
Correct this to standard English: The quik brown fox jumps the lazy dog
---
Corrected:

>>> # Sentence with column map where the 'sentence' key is actually named 'input' in the given sample
>>> GrammarErrorCorrectionTemplate.format(
...     sample={"input": "The quik brown fox jumps the lazy dog"},
...     column_map={"sentence": "input"}
... )
Correct this to standard English: The quik brown fox jumps the lazy dog
---
Corrected:

Returns:

The formatted prompt