Enforcing guidance throughout the entire sampling process often proves counterproductive due to the model-fitting issue., where samples are generated to match the classifier's parameters rather than generalizing the expected condition. This work identifies and quantifies the problem, demonstrating that reducing or excluding guidance at numerous timesteps can mitigate this issue. By distributing the guidance densely in the early stages of the process, we observe a significant improvement in image quality and diversity while also reducing the required guidance timesteps by nearly 40%. This approach addresses a major challenge in applying guidance effectively to generative tasks. Consequently, our proposed method, termed Compress Guidance, allows for the exclusion of a substantial number of guidance timesteps while still surpassing baseline models in image quality. We validate our approach through benchmarks on label conditional and text-to-image generative tasks across various datasets and models.
View on arXiv