Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2511.11446 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911266068496384 |
|---|---|
| author | Amin, Farhana Afroz, Sabiha Gharami, Kanchon Moghadampanah, Mona Nikolopoulos, Dimitrios S. |
| author_facet | Amin, Farhana Afroz, Sabiha Gharami, Kanchon Moghadampanah, Mona Nikolopoulos, Dimitrios S. |
| contents | Diffusion models produce high quality images but inference is costly due to many denoising steps and heavy matrix operations. We present DiffPro, a post-training, hardware-faithful framework that works with the exact integer kernels used in deployment and jointly tunes timesteps and per-layer precision in Diffusion Transformers (DiTs) to reduce latency and memory without any training. DiffPro combines three parts: a manifold-aware sensitivity metric to allocate weight bits, dynamic activation quantization to stabilize activations across timesteps, and a budgeted timestep selector guided by teacher-student drift. In experiments DiffPro achieves up to 6.25x model compression, fifty percent fewer timesteps, and 2.8x faster inference with Delta FID <= 10 on standard benchmarks, demonstrating practical efficiency gains. DiffPro unifies step reduction and precision planning into a single budgeted deployable plan for real-time energy-aware diffusion inference. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_11446 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DiffPro: Joint Timestep and Layer-Wise Precision Optimization for Efficient Diffusion Inference Amin, Farhana Afroz, Sabiha Gharami, Kanchon Moghadampanah, Mona Nikolopoulos, Dimitrios S. Machine Learning Diffusion models produce high quality images but inference is costly due to many denoising steps and heavy matrix operations. We present DiffPro, a post-training, hardware-faithful framework that works with the exact integer kernels used in deployment and jointly tunes timesteps and per-layer precision in Diffusion Transformers (DiTs) to reduce latency and memory without any training. DiffPro combines three parts: a manifold-aware sensitivity metric to allocate weight bits, dynamic activation quantization to stabilize activations across timesteps, and a budgeted timestep selector guided by teacher-student drift. In experiments DiffPro achieves up to 6.25x model compression, fifty percent fewer timesteps, and 2.8x faster inference with Delta FID <= 10 on standard benchmarks, demonstrating practical efficiency gains. DiffPro unifies step reduction and precision planning into a single budgeted deployable plan for real-time energy-aware diffusion inference. |
| title | DiffPro: Joint Timestep and Layer-Wise Precision Optimization for Efficient Diffusion Inference |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2511.11446 |