shanaka95/gemma-4-2b-finetuned-grammar2
The shanaka95/gemma-4-2b-finetuned-grammar2 model is a 5.1 billion parameter Gemma-4-E2B-it base model, developed by Shanaka Anuradha and fine-tuned on the Grammarly CoEdIT dataset. With a 32768 token context length, it is specifically optimized for text editing tasks such as grammatical error correction, paraphrasing, simplification, and style transfer. This specialized model excels at following natural language instructions to revise and improve existing text.
Loading preview...
Model Overview
This model, shanaka95/gemma-4-2b-finetuned-grammar2, is a specialized text-editing assistant developed by Shanaka Anuradha. It is built upon the Gemma-4-E2B-it base model and fine-tuned using the Grammarly CoEdIT dataset. The model leverages Unsloth for efficient training, achieving 2x faster training times and reduced memory usage. Its primary function is to take an instruction and a source text, then generate a revised or altered version of that text based on the command.
Key Capabilities
This model is explicitly trained for various text revision tasks, including:
- Grammatical Error Correction (GEC): Correcting grammatical mistakes in sentences.
- Text Simplification: Making text easier to understand.
- Paraphrasing: Rewriting sentences in different ways.
- Formality Transfer: Adjusting the formality level of text.
- Coherence & Flow Improvement: Enhancing the readability and structure of paragraphs.
- Neutralization: Making sentences more neutral in tone.
Intended Use Cases
This model is ideal for applications requiring precise, instruction-driven text transformations. It excels at structural and stylistic modifications of existing text. It is not designed for open-ended chatbot conversations, long-form creative writing, or factual question answering. Its strength lies in its ability to follow natural language commands for targeted text editing.