Papers › Score identity Distillation: Exponentially Fast Distillation of Pretrained Diffusion...

Score identity Distillation: Exponentially Fast Distillation of Pretrained Diffusion Models for One-Step Generation

5 Apr 2024arXiv:2404.04057archive 2025-07-28

Mingyuan Zhou, Huangjie Zheng, Zhendong Wang, Mingzhang Yin, Hai Huang

We introduce Score identity Distillation (SiD), an innovative data-free method that distills the generative capabilities of pretrained diffusion models into a single-step generator. SiD not only facilitates an exponentially fast reduction in Fr\'echet inception distance (FID) during distillation but also approaches or even exceeds the FID performance of the original teacher diffusion models. By reformulating forward diffusion processes as semi-implicit distributions, we leverage three score-related identities to create an innovative loss mechanism. This mechanism achieves rapid FID reduction by training the generator using its own synthesized images, eliminating the need for real data or reverse-diffusion-based generation, all accomplished within significantly shortened generation time. Upon evaluation across four benchmark datasets, the SiD algorithm demonstrates high iteration efficiency during distillation and surpasses competing distillation approaches, whether they are one-step or few-step, data-free, or dependent on training data, in terms of generation quality. This achievement not only redefines the benchmarks for efficiency and effectiveness in diffusion distillation but also in the broader field of diffusion-based generation. The PyTorch implementation is available at https://github.com/mingyuanzhou/SiD

PaperPDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2404.04057")

Code

Syntology Ran 9 of 15 code samples harvested from 1 repository linked to this paper; 6 have no recorded run. Of those that ran: 3 ran · honoured contract; 1 ran · our draft was wrong; 5 ran with no contract checked.

By repository: official repository: 15 samples from 1 repository, 9 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

mingyuanzhou/sid officialmentioned in papermentioned on GitHubpytorchApache-2.0 report
mingyuanzhou/sid-lsg mentioned on GitHubpytorchApache-2.0 report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

15 samples harvested; 9 ran; 3 honoured the contract we drafted; 6 have no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

3ran · honoured contract
1ran · our draft was wrong
5ran
6unverified

Licence: 0 of the 15 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from mingyuanzhou/SiD. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

calculate_fid_from_inception_stats mingyuanzhou/SiD/sid_generate.py official repository ran · honoured contract Apache-2.0 (permissive) · 68b5e143931956cf · report
compute_distances mingyuanzhou/SiD/metrics/sid_precision_recall.py official repository ran Apache-2.0 (permissive) · 790b320b90341326 · report
file_ext mingyuanzhou/SiD/dataset_tool.py official repository ran · our draft was wrong fingerprinted Apache-2.0 (permissive) · a2b45afa097b55b6 · report
is_valid_metric mingyuanzhou/SiD/metrics/sid_metric_main.py official repository ran Apache-2.0 (permissive) · 147acc22e758c8e5 · report
maybe_min mingyuanzhou/SiD/dataset_tool.py official repository ran · honoured contract Apache-2.0 (permissive) · dd1dc700e87f36cc · report
parse_int_list mingyuanzhou/SiD/sid_metrics.py official repository ran · honoured contract Apache-2.0 (permissive) · cacd4f6ec202d9b4 · report
parse_tuple mingyuanzhou/SiD/dataset_tool.py official repository ran Apache-2.0 (permissive) · 07b21b35ecb4815c · report
register_metric mingyuanzhou/SiD/metrics/sid_metric_main.py official repository ran Apache-2.0 (permissive) · e574c21e7f007b61 · report
setup_snapshot_image_grid mingyuanzhou/SiD/training/sid_training_loop.py official repository ran Apache-2.0 (permissive) · 41d0a46d745163d4 · report
compute_fid mingyuanzhou/SiD/metrics/sid_frechet_inception_distance.py official repository unverified Apache-2.0 (permissive) · a7058b1ccaa13516 · report
compute_is mingyuanzhou/SiD/metrics/sid_inception_score.py official repository unverified Apache-2.0 (permissive) · 47c7a1249e324f74 · report
compute_kid mingyuanzhou/SiD/metrics/sid_kernel_inception_distance.py official repository unverified Apache-2.0 (permissive) · 5dd3f075a53b2c58 · report
compute_pr mingyuanzhou/SiD/metrics/sid_precision_recall.py official repository unverified Apache-2.0 (permissive) · b657531c701b1786 · report
get_feature_detector_name mingyuanzhou/SiD/metrics/sid_metric_utils.py official repository unverified Apache-2.0 (permissive) · c127dc46d6966547 · report
sid_sampler mingyuanzhou/SiD/sid_generate.py official repository unverified Apache-2.0 (permissive) · 4c9c2d2ca6c238d8 · report

Tasks

Image Generation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Image Generation AFHQ-v2 64x64 SiD FID 1.711 #2 of 2 Archive leaderboard report
Image Generation AFHQ-v2 64x64 SiD NFE 1 #2 of 2 Archive leaderboard report
Image Generation CIFAR-10 SiD FID 1.71 #8 of 78 Archive leaderboard report
Image Generation CIFAR-10 SiD NFE 1 #8 of 78 Archive leaderboard report
Image Generation FFHQ 64x64 SiD FID 1.55 #2 of 2 Archive leaderboard report
Image Generation FFHQ 64x64 SiD NFE 1 #2 of 2 Archive leaderboard report
Image Generation ImageNet 64x64 SiD FID 1.524 #13 of 65 Archive leaderboard report
Image Generation ImageNet 64x64 SiD NFE 1 #13 of 65 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Diffusion

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections