t5-base-grammar-correction

t5-base-grammar-correction

Rewrites English text with fewer grammar mistakes, using a T5 model trained on the JFLEG dataset. Made by vennify for writers, students and developers who want a grammar fixing step in Python.

Made by vennify on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: t5-base-grammar-correction is a T5 model from vennify that takes a piece of English text and returns a revised copy with fewer grammatical errors. It was trained with the Happy Transformer library on a dataset called JFLEG. You call it from Python and get the corrected text back.

Who it suits: It suits writers, students and developers who want automatic grammar fixes and can run a short Python script.

Industry | Who uses it | How they use it | Example
Education | Students and teachers | Clean up grammar in essays and exercises | Run a sentence with doubled words through the model
Publishing | Editors and bloggers | Fix small slips in drafts before a human review | Pass a draft in one sentence at a time
Software | Developers | Add a grammar fixing step to a text tool | Call the model from a script and show the revised text

People who write in languages other than English can skip it, since the record lists English only.

Where to use it: The model runs from Python on a computer you control.

• Windows, macOS and Linux, with Python
• Python scripts that use the Happy Transformer library
• Python projects built on the transformers library, which the record lists

Use case | Who | What they get back
Proofreading a sentence | Writers | A revised sentence with the errors fixed
Batch clean-up of text | Developers | Revised text for each input a script passes in

How to set it up: You need Python with pip, and a computer that can download the model files.

1. Install the Hugging Face hub library with pip.
2. Download the model files with the hf command.
3. Install the Happy Transformer library with pip.

pip install -U huggingface_hub
hf download vennify/t5-base-grammar-correction
pip install happytransformer

How to use it: In Python, load the T5 model through Happy Transformer, set the generation settings and put the prefix grammar: and a space before each input. Then call generate_text and print the text of the result.

from happytransformer import HappyTextToText, TTSettings
happy_tt = HappyTextToText("T5", "vennify/t5-base-grammar-correction")
args = TTSettings(num_beams=5, min_length=1)
result = happy_tt.generate_text("grammar: This sentences has has bads grammar.", args=args)
print(result.text)

Tool | What it does
HappyTextToText | Loads the T5 model for text to text jobs
TTSettings | Holds the generation settings, here 5 beams and a minimum length of 1
generate_text | Returns the revised text for the input you give it

Example: Give the model a sentence with doubled words and a wrong plural, such as the one in the card's snippet. The card's own comment shows the result as a clean sentence reading "This sentence has bad grammar."

For newcomers and experts: If you are new to this, install the library, paste the snippet and swap the sentence inside the quotes for your own. Keep the grammar: prefix at the start of the text.

If you know the field, this is T5 base fine-tuned for sentence revision on JFLEG, run as a text to text job with beam search. It returns a rewritten sentence and not a list of marked errors.

Add the grammar: prefix to your text
Load T5 with Happy Transformer
The model generates with 5 beams
The revised text comes back in result.text

Good to know: The model returns a revised version of your text, so read the output before you use it.

• License: cc-by-nc-sa-4.0 as the record states, which marks it non-commercial
• Languages: English only in the record
• Other versions: the record counts 4 quantized copies and 2 fine-tuned ones made by other people
• Cost to run: none named
• Data it reaches: the text you pass in, plus the model files you download from Hugging Face
• Not for: paid products, so check the license terms first

On Hugging Face, 13 November 2026

• Library: transformers
• Downloads in the last 30 days: 127,257
• Likes: 187
• License: cc-by-nc-sa-4.0
• Languages: en
• Quantized versions: 4
• Fine-tunes: 2
• Tags: transformers, pytorch, t5, text2text-generation, grammar

Links

• Model: https://huggingface.co/vennify/t5-base-grammar-correction
• Files: https://huggingface.co/vennify/t5-base-grammar-correction/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download vennify/t5-base-grammar-correction

Then follow the usage in its model card. Nothing in this kit runs by itself.