TinyStyler: Efficient Few-Shot Text Style Transfer with Authorship Embeddings
About
The goal of text style transfer is to transform the style of texts while preserving their original meaning, often with only a few examples of the target style. Existing style transfer methods generally rely on the few-shot capabilities of large language models or on complex controllable text generation approaches that are inefficient and underperform on fluency metrics. We introduce TinyStyler, a lightweight but effective approach, which leverages a small language model (800M params) and pre-trained authorship embeddings to perform efficient, few-shot text style transfer. We evaluate on the challenging task of authorship style transfer and find TinyStyler outperforms strong approaches such as GPT-4. We also evaluate TinyStyler's ability to perform text attribute style transfer (formal $\leftrightarrow$ informal) with automatic and human evaluations and find that the approach outperforms recent controllable text generation methods. Our model has been made publicly available at https://huggingface.co/tinystyler/tinystyler .
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Author Style Transfer | Project Gutenberg 100 source–target author pairs (test) | Toward16 | 10 | |
| Style-controlled text generation | Style mimicry evaluation dataset Section 7 (test) | L2 Distance (Our Embedding)49.97 | 6 | |
| Style-Guided Text Generation | Human Writing Styles | L2 Distance (Our Embedding)54.69 | 6 | |
| Machine-Text Detection Evasion (Text Transformation) | Reddit, Amazon, and Blogs Average | Edit Distance212.6 | 5 |