ctm-lsun-bedroom

Consistency Trajectory Model (CTM) – LSUN Bedroom (256x256)

This repository hosts the official model checkpoint for Consistency Trajectory Models (CTM) trained on the LSUN Bedroom dataset.

---

Overview

Consistency Trajectory Models (CTM) unify score-based diffusion models and Consistency Models (CM) into a single framework. CTM trains a single neural network capable of:

• Generating high-quality samples in a single forward pass (1-step generation).
• Performing unrestricted long-jump traversals along the Probability Flow ODE trajectory for flexible quality-vs-speed trade-offs.
• Accessing score functions directly to support likelihood evaluation and controllable/conditional generation methods.

Abstract

Consistency Models (CM) (Song et al., 2023) accelerate score-based diffusion model sampling at the cost of sample quality but lack a natural way to trade-off quality for speed. To address this limitation, we propose Consistency Trajectory Model (CTM), a generalization encompassing CM and score-based models as special cases. CTM trains a single neural network that can -- in a single forward pass -- output scores (i.e., gradients of log-density) and enables unrestricted traversal between any initial and final time along the Probability Flow Ordinary Differential Equation (ODE) in a diffusion process. CTM enables the efficient combination of adversarial training and denoising score matching loss to enhance performance and achieves new state-of-the-art FIDs for single-step diffusion model sampling on CIFAR-10 (FID 1.73) and ImageNet at 64x64 resolution (FID 1.92). CTM also enables a new family of sampling schemes, both deterministic and stochastic, involving long jumps along the ODE solution trajectories. It consistently improves sample quality as computational budgets increase, avoiding the degradation seen in CM. Furthermore, unlike CM, CTM's access to the score function can streamline the adoption of established controllable/conditional generation methods from the diffusion community. This access also enables the computation of likelihood.

---

Citation details:

If you find this method and/or code useful, please consider citing:

@article{kim2023consistency,
  title={Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusion},
  author={Kim, Dongjun and Lai, Chieh-Hsin and Liao, Wei-Hsiang and Murata, Naoki and Takida, Yuhta and Uesaka, Toshimitsu and He, Yutong and Mitsufuji, Yuki and Ermon, Stefano},
  journal={arXiv preprint arXiv:2310.02279},
  year={2023}

Credits & Attribution

Teacher model checkpoint is a re-hosting of the official weights released by OpenAI for their Consistency Models research.

• Original Paper: Consistency Models (Song et al., 2023)
• Original Repository: openai/consistency_models
• Original Checkpoint Source: OpenAI Public Blob Storage
• License: MIT License

Citation

If you use this model in your research or project, please cite the original paper:

@article{song2023consistency,
  title={Consistency Models},
  author={Song, Yang and Dhariwal, Prafulla and Chen, Mark and Sutskever, Ilya},
  journal={arXiv preprint arXiv:2303.01469},
  year={2023}
}

<!-- **Where to send questions or comments about the paper and model:** Questions and comments about CTM can be sent via the [Github repository](https://github.com/sony/ctm) of the project or [email protected]. -->