SenseFlow: Scaling Distribution Matching for Flow-based Text-to-Image Distillation

Xingtong Ge^1,2, Xin Zhang², Tongda Xu³, Yi Zhang⁴, Xinjie Zhang¹, Yan Wang³, Jun Zhang¹

¹HKUST, ²SenseTime Research, ³Tsinghua University, ⁴CUHK MMLab

Abstract

The Distribution Matching Distillation (DMD) has been successfully applied to text-to-image diffusion models such as Stable Diffusion (SD) 1.5. However, vanilla DMD suffers from convergence difficulties on large-scale flow-based text-to-image models, such as SD 3.5 and FLUX. In this paper, we first analyze the issues when applying vanilla DMD on large-scale models. Then, to overcome the scalability challenge, we propose implicit distribution alignment (IDA) to regularize the distance between the generator and fake distribution. Furthermore, we propose intra-segment guidance (ISG) to relocate the timestep importance distribution from the teacher model. With IDA alone, DMD converges for SD 3.5; employing both IDA and ISG, DMD converges for SD 3.5 and FLUX.1 dev. Along with other improvements such as scaled up discriminator models, our final model, dubbed SenseFlow, achieves superior performance in distillation for both diffusion based text-to-image models such as SDXL, and flow-matching models such as SD 3.5 Large and FLUX. The source code will be released.

🌟Table 1: Quantitative Results on COCO-5K Dataset

Bold = best, Underline = second best. All results on 4-step generation.

Stable Diffusion XL Comparison

Method	NFE	FID-T	Patch FID-T	CLIP	HPSv2	Pick	ImageReward
SDXL	80	--	--	0.3293	0.2930	22.67	0.8719
LCM-SDXL	4	18.47	30.63	0.3230	0.2824	22.22	0.5693
PCM-SDXL	4	14.38	17.77	0.3242	0.2920	22.54	0.6926
Flash-SDXL	4	17.97	23.24	0.3216	0.2830	22.17	0.4295
SDXL-Lightning	4	13.67	16.57	0.3214	0.2931	22.80	0.7799
Hyper-SDXL	4	13.71	17.49	0.3254	0.3000	22.98	0.9777
DMD2-SDXL	4	15.04	18.72	0.3277	0.2963	22.98	0.9324
Ours-SDXL	4	17.76	21.01	0.3248	0.3010	23.17	0.9951

Stable Diffusion 3.5 Comparison

Method	NFE	FID-T	Patch FID-T	CLIP	HPSv2	Pick	ImageReward
SD 3.5 Large	100	--	--	0.3310	0.2993	22.98	1.1629
SD 3.5 Large Turbo	4	13.58	22.88	0.3262	0.2909	22.89	1.0116
Ours-SD 3.5	4	13.38	17.48	0.3286	0.3016	23.01	1.1713
Ours-SD 3.5 (Euler)	4	15.24	20.26	0.3287	0.3008	22.90	1.2062

FLUX Comparison

Method	NFE	FID-T	Patch FID-T	CLIP	HPSv2	Pick	ImageReward
FLUX.1 dev	50	--	--	0.3202	0.3000	23.18	1.1170
FLUX.1 dev	25	--	--	0.3207	0.2986	23.14	1.1063
FLUX.1-schnell	4	--	--	0.3264	0.2962	22.77	1.0755
Hyper-FLUX	4	11.24	23.47	0.3238	0.2963	23.09	1.0983
FLUX-Turbo-Alpha	4	11.22	24.52	0.3218	0.2907	22.89	1.0106
Ours-FLUX	4	15.64	19.60	0.3167	0.2997	23.13	1.0921
Ours-FLUX (Euler)	4	16.50	20.29	0.3171	0.3008	23.26	1.1424

---

🌟Table 2: 4-Step Results on T2I-CompBench

Bold, Underline = best and second best (per teacher).

Method	Color	Shape	Texture	Spatial	Non-spatial	Complex-3-in-1
LCM-SDXL	0.5997	0.4015	0.4958	0.1672	0.3010	0.3364
SDXL-Lightning	0.5758	0.4492	0.5154	0.2124	0.3098	0.3517
Hyper-SDXL	0.6435	0.4732	0.5581	0.2213	0.3104	0.3301
PCM-SDXL	0.5591	0.4142	0.4693	0.2013	0.3099	0.3234
DMD2-SDXL	0.5811	0.4477	0.5175	0.2124	0.3098	0.3301
Ours-SDXL	0.6867	0.4828	0.5989	0.2224	0.3100	0.3594
SD 3.5 Large Turbo	0.7050	0.5443	0.6512	0.2839	0.3130	0.3520
Ours-SD 3.5	0.7657	0.6069	0.7427	0.2970	0.3177	0.3916
Ours-SD 3.5 (Euler)	0.7711	0.6149	0.7543	0.2857	0.3182	0.3968
FLUX.1-schnell	0.7317	0.5649	0.6919	0.2626	0.3122	0.3669
Hyper-FLUX	0.7465	0.5023	0.6153	0.2945	0.3116	0.3766
FLUX-Turbo-Alpha	0.7406	0.4873	0.6024	0.2501	0.3094	0.3688
Ours-FLUX	0.7284	0.5055	0.6031	0.2451	0.3028	0.3652
Ours-FLUX (Euler)	0.7363	0.5120	0.6112	0.2521	0.3028	0.3697

🌟 1024 x 1024 examples of our 4-step generator distilled on SD 3.5 Large

🌟 1024 x 1024 examples of our 4-step generator distilled on SDXL