TeEFusion: Blending Text Embeddings to Distill Classifier-Free Guidance
This work addresses efficiency issues in text-to-image generation for AI practitioners, offering a significant speed-up with minimal quality loss, though it is incremental as it builds on existing distillation and guidance techniques.
The paper tackles the high inference cost of classifier-free guidance in text-to-image synthesis by introducing TeEFusion, a distillation method that fuses text embeddings to incorporate guidance magnitude, enabling a student model to achieve up to 6x faster inference speeds while maintaining comparable image quality to the teacher model.
Recent advances in text-to-image synthesis largely benefit from sophisticated sampling strategies and classifier-free guidance (CFG) to ensure high-quality generation. However, CFG's reliance on two forward passes, especially when combined with intricate sampling algorithms, results in prohibitively high inference costs. To address this, we introduce TeEFusion (Text Embeddings Fusion), a novel and efficient distillation method that directly incorporates the guidance magnitude into the text embeddings and distills the teacher model's complex sampling strategy. By simply fusing conditional and unconditional text embeddings using linear operations, TeEFusion reconstructs the desired guidance without adding extra parameters, simultaneously enabling the student model to learn from the teacher's output produced via its sophisticated sampling approach. Extensive experiments on state-of-the-art models such as SD3 demonstrate that our method allows the student to closely mimic the teacher's performance with a far simpler and more efficient sampling strategy. Consequently, the student model achieves inference speeds up to 6$\times$ faster than the teacher model, while maintaining image quality at levels comparable to those obtained through the teacher's complex sampling approach. The code is publicly available at https://github.com/AIDC-AI/TeEFusion.