Continuous-Time Distribution Matching for Few-Step Diffusion Distillation
Algorithm Overview •
Results •
Inference •
Training •
Evaluation •
Citation
## Algorithm Overview
**Overview of Continuous-Time Distribution Matching (CDM).** **Top:** Our approach employs a dynamic continuous time schedule during backward simulation, sampling intermediate anchors uniformly from (0, 1]. **Bottom Left:** CFG augmentation (CA) and distribution matching (DM) operate on this dynamic schedule to align text-image conditions and data distributions at on-trajectory anchors. **Bottom Right:** To address inter-anchor inconsistency, the proposed CDM objective explicitly extrapolates off-trajectory latents using the predicted velocity.
## 4-NFE Generation Results
### SD3-Medium
### LongCat
---
## Inference
```bash
# Clone this repository
git clone https://github.com/byliutao/cdm.git
cd cdm
# [Optional] Use HuggingFace mirror if huggingface.co is not accessible
export HF_ENDPOINT="https://hf-mirror.com"
export HF_TOKEN="hf_xxx"
# Create and activate the inference environment
conda create -n cdm_infer python=3.10
conda activate cdm_infer
pip install -r config/requirements_infer.txt
# Run inference
python scripts/infer/sd3_m.py # SD3-Medium
python scripts/infer/longcat.py # LongCat
```
## Training
```bash
# Create and activate the training environment
conda create -n cdm_train python=3.10
conda activate cdm_train
pip install -r config/requirements_train.txt
pip install flash-attn==2.7.4.post1 --no-build-isolation # May take 1-2 hours
# Launch training with FSDP2
accelerate launch --config_file config/accelerate_fsdp2.yaml \
--num_processes 8 -m scripts.train \
--config config/config.py:sd3 # SD3-Medium
accelerate launch --config_file config/accelerate_fsdp2.yaml \
--num_processes 8 -m scripts.train \
--config config/config.py:longcat # LongCat
```
## Evaluation
Evaluation is split into two phases: **image generation** and **metric computation**.
### Step 1 — Export a checkpoint to a pipeline
```bash
conda activate cdm_train
python -m scripts.save \
--experiment_dir "logs/experiments/sd3/test" \
--output_dir "logs/pipelines/test" \
--checkpoint_steps "2000"
```
### Step 2 — Generate images
```bash
accelerate launch --num_processes 8 -m scripts.eval \
--phase generate \
--model_path "logs/pipelines/test/checkpoint-2000" \
--eval_metrics imagereward clipscore pickscore hpsv2 hpsv3 aesthetic ocr dpgbench \
--output_dir "logs/evaluations/test" \
--base_model sd3 \
--save_images
```
### Step 3 — Compute metrics
```bash
# Create a separate environment for evaluation dependencies
conda create -n cdm_eval python=3.10
conda activate cdm_eval
pip install -r config/requirements_eval.txt
pip install image-reward --no-deps
pip install fairseq --no-deps
# NOTE: If running on multiple GPUs, download checkpoints on 1 GPU first.
# For FID evaluation, place COCO 2014 val images under: dataset/coco2014val_10k/images
accelerate launch --num_processes 8 -m scripts.eval \
--phase evaluate \
--eval_metrics imagereward clipscore pickscore hpsv2 hpsv3 aesthetic ocr dpgbench \
--output_dir "logs/evaluations/test"
```
## License
This project is licensed under the MIT License — see the [LICENSE](LICENSE) file for details.
## Citation
If our work assists your research, please consider giving us a star ⭐ or citing us:
```bibtex
@misc{liu2026continuoustimedistributionmatchingfewstep,
title={Continuous-Time Distribution Matching for Few-Step Diffusion Distillation},
author={Tao Liu and Hao Yan and Mengting Chen and Taihang Hu and Zhengrong Yue and Zihao Pan and Jinsong Lan and Xiaoyong Zhu and Ming-Ming Cheng and Bo Zheng and Yaxing Wang},
year={2026},
eprint={2605.06376},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2605.06376},
}
```
