A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training

If you like SpeeD, please give us a star ⭐ on GitHub for the latest update.

Paper | Project Page | Hugging Face

This repository contains the code and implementation details for the research paper titled "A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training." In this paper, SpeeD, a novel speed-up method for diffusion model training, is introduced.

Authors

• Kai Wang², Yukun Zhou^1,2, Mingjia Shi², Zhihang Yuan³, Yuzhang Shang⁴, Xiaojiang Peng*¹, Hanwang Zhang⁵, Yang You²
• ¹Shenzhen Technology University, ²National University of Singapore, ³Infinigence-AI, ⁴Illinois Institute of Technology, and ⁵Nanyang Technological University [Kai, Yukun, and Mingjia contribute equally to this work. We will update this repo asap.]

😮 Highlights

Our method, which is easily compatible, can accelerate the training of diffusion model.

✒️ Motivation

Inspired by the following observation on the time step, we propose the re-sampling + re-weighting strategy as shown below.

To take a closer look at time steps, we find that the time steps could be divided into three areas: acceleration, decceleration and convergence areas. Samples of the corresponding time step in the convergence region are of limited benefit to training, while these time steps take up the most. Empirically, the training losses of these samples are quite low compare to the ones of the other two areas. Re-sampling: Suppress the attendance of the time step in convergence areas. Weighting: The faster changing time steps in the diffusion process are given more weight.

🔆 Method

We use the sampling and weighting strategy which are simple and easily compatible to achieve the acceleration. The following is the core code SpeeD/speed/diffusion/iddpm/speed.py ,

class SpeeDiffusion(SpacedDiffusion):
    def __init__(self, faster, **kwargs):
        super().__init__(**kwargs)
        self.faster = faster
        if faster:
            grad = np.gradient(self.sqrt_one_minus_alphas_cumprod)

            # set the meaningful steps in diffusion, which is more important in inference
            self.meaningful_steps = np.argmax(grad < 1e-4) + 1

            # p2 weighting from: Perception Prioritized Training of Diffusion Models
            self.p2_gamma = 1
            self.p2_k = 1
            self.snr = 1.0 / (1 - self.alphas_cumprod) - 1
            sqrt_one_minus_alphas_bar = torch.from_numpy(self.sqrt_one_minus_alphas_cumprod)
            # sample more meaningful step
            p = torch.tanh(1e6 * (torch.gradient(sqrt_one_minus_alphas_bar)[0] - 1e-4)) + 1.5
            self.p = F.normalize(p, p=1, dim=0)
            self.weights = self._weights()
        else:
            self.meaningful_steps = self.num_timesteps

    def _weights(self):
        # process where all noise to noisy image with content has more weighting in training
        # the weights act on the mse loss
        weights =  1 / (self.p2_k + self.snr) ** self.p2_gamma
        weights = weights
        return weights

    # get the weights and sampling t in training diffusion
    def t_sample(self, n, device):
        if self.faster:
            t = torch.multinomial(self.p, n // 2 + 1, replacement=True).to(device)
            # dual sampling, which can balance the step multiple task training
            dual_t = torch.where(t < self.meaningful_steps, self.meaningful_steps - t, t - self.meaningful_steps)
            t = torch.cat([t, dual_t], dim=0)[:n]
            weights = self.weights
        else:
            # if
            t = torch.randint(0, self.num_timesteps, (n,), device=device)
            weights = None

        return t, weights

You can enable our acceleration module with diffusion.faster=True.

# config file
diffusion:
    timestep_respacing: '250'
    faster: true  #enabl module for training acceleration

🛠️ Requirements and Installation

This code base does not use hardware acceleration technology, experimental environment is not complicated.

You can create a new conda environment:

conda env create -f environment.yml
conda activate speed

or install the necessary package by:

pip install -r requirements.txt

If necessary, we will provide more methods (e.g., docker) to facilitate the configuration of the experimental environment.

🗝️ Implementation

We provide a complete process for generating tasks including training, inference and test. The current code is only compatible with class-conditional image generation tasks. We will be compatible with more generation tasks about diffusion in the future.

We refactor the facebookresearch/DiT code and loaded the configs using OmegaConf . The configuration file loading rule is recursive for easier argument modification. Simply put, the file in the latter path will override the previous setting of base.yaml.

You can modify the experiment setting by modifying the config file and the command line. More details about the reading of config are written in configs/README.md.

For each experiment, you must provide two arguments by command,

-c: config path;
-p: phase including ['train', 'inference', 'sample'].

Train & inference

For example, class-conditional image generation task with 256x256 ImageNet dataset and DiT-XL/2 models.

# Training: training diffusion and saving checkpoints
torchrun --nproc_per_node=8 main.py -c configs/image/imagenet_256/base.yaml -p train
# inference: generating samples for testing
torchrun --nproc_per_node=8 main.py -c configs/image/imagenet_256/base.yaml -p inference
# sample: sample some images for visualization
python main.py -c configs/image/imagenet_256/base.yaml -p sample

How to do ablation?

You can modify the experiment setting by modifying the config file and the command line. More details about the reading of config are written in configs/README.md.

For example, change the classifier-free guidance scale in sampling by command line:

python main.py -c configs/image/imagenet_256/base.yaml -p sample guidance_scale=1.5

Test

Test the generation tasks require the results of inference. The more details about testing in evaluations.

👍 Acknowledgement

We are grateful for the following exceptional work and generous contribution to open source.

• DiT: Scalable Diffusion Models with Transformers.
• Open-Sora : Open-Sora: Democratizing Efficient Video Production for All
• OpenDiT: An acceleration for DiT training. We adopt valuable acceleration strategies for training progress from OpenDiT.

🔒 License

The majority of this project is released under the Apache 2.0 license as found in the LICENSE file.

✏️Citation

If you find our code useful in your research, please consider giving a star ⭐ and citation 📝.

@software{speed,
  author = {Yukun Zhou, Kai Wang, Hanwang Zhang, Yang You and Xiaojiang Peng},
  title = {SpeeD: A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training},
  month = {March},
  year = {2024},
  url = {https://github.com/kaiwang960112/SpeeD}
}

@article{wang2024closer,
      title={A Closer Look at Time Steps is Worthy of Triple Speed-Up for Diffusion Model Training}, 
      author={Kai Wang, Yukun Zhou, Mingjia Shi, Zhihang Yuan, Yuzhang Shang, Xiaojiang Peng, Hanwang Zhang and Yang You},
      year={2024},
      journal={arXiv preprint arXiv:2405.17403},
}