Improving Consistency Models with Generator-Augmented Flows
Authors: Thibaut Issenhuth, Sangchul Lee, Ludovic Dos Santos, Jean-Yves Franceschi, Chansoo Kim, Alain Rakotomamonjy
ICML 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Our approach based on a joint learning strategy leads to faster convergence and improves the performance compared to the base model and OT-based approaches on image generation benchmarks. ... Results across multiple datasets and metrics are presented in Table 1, and visual examples are shown in Figure 8 in Appendix. |
| Researcher Affiliation | Collaboration | 1Criteo AI Lab, Paris, France 2AI, Information and Reasoning (AI/R) Laboratory, Korea Institute of Science and Technology 3AI and Robot Department, University of Science and Technology, Korea 4LITIS, Univ Rouen-Normandie. Correspondence to: Thibaut Issenhuth, Chansoo Kim <EMAIL; EMAIL>. |
| Pseudocode | Yes | Algorithm 1 Training of consistency models with generator-augmented trajectories |
| Open Source Code | Yes | The code is available at: github.com/thibautissenhuth/consistencyGC. |
| Open Datasets | Yes | Our experiments are done on the following datasets: CIFAR10 (Krizhevsky, 2009), Image Net (Deng et al., 2009), Celeb A (Liu et al., 2015) and LSUN Church (Yu et al., 2015). |
| Dataset Splits | Yes | For the three different metrics, we use the standard practice (e.g. (Song and Dhariwal, 2024)) of FID which is to compare sets of 50 000 generated versus training images. Confidence intervals reported in Table 1 are averaged on five runs by sampling new sets of training images, and new sets of generated images from the same model. |
| Hardware Specification | Yes | As mentioned in the paper, the image dataset experiments have been conducted on NVIDIA A100 40GB GPUs. |
| Software Dependencies | No | The code is based on the Py Torch library (Paszke et al., 2019). We use the lion optimizer (Chen et al., 2023) implemented from https://github.com/lucidrains/lion-pytorch. The paper mentions software components (PyTorch, Lion optimizer) but does not provide specific version numbers for these, only citations to the papers introducing them. |
| Experiment Setup | Yes | Table 4. Hyperparameters for CIFAR-10. ... Table 5. Hyperparameters for Celeb A and LSUN Church. ... Table 6. Hyperparameters for Image Net-1k. |