Papers › PTQ4DiT: Post-training Quantization for Diffusion Transformers

PTQ4DiT: Post-training Quantization for Diffusion Transformers

25 May 2024arXiv:2405.16005archive 2025-07-28

Junyi Wu, Haoxuan Wang, Yuzhang Shang, Mubarak Shah, Yan Yan

The recent introduction of Diffusion Transformers (DiTs) has demonstrated exceptional capabilities in image generation by using a different backbone architecture, departing from traditional U-Nets and embracing the scalable nature of transformers. Despite their advanced capabilities, the wide deployment of DiTs, particularly for real-time applications, is currently hampered by considerable computational demands at the inference stage. Post-training Quantization (PTQ) has emerged as a fast and data-efficient solution that can significantly reduce computation and memory footprint by using low-bit weights and activations. However, its applicability to DiTs has not yet been explored and faces non-trivial difficulties due to the unique design of DiTs. In this paper, we propose PTQ4DiT, a specifically designed PTQ method for DiTs. We discover two primary quantization challenges inherent in DiTs, notably the presence of salient channels with extreme magnitudes and the temporal variability in distributions of salient activation over multiple timesteps. To tackle these challenges, we propose Channel-wise Salience Balancing (CSB) and Spearmen's ρ-guided Salience Calibration (SSC). CSB leverages the complementarity property of channel magnitudes to redistribute the extremes, alleviating quantization errors for both activations and weights. SSC extends this approach by dynamically adjusting the balanced salience to capture the temporal variations in activation. Additionally, to eliminate extra computational costs caused by PTQ4DiT during inference, we design an offline re-parameterization strategy for DiTs. Experiments demonstrate that our PTQ4DiT successfully quantizes DiTs to 8-bit precision (W8A8) while preserving comparable generation ability and further enables effective quantization to 4-bit weight precision (W4A8) for the first time.

PaperPDFCodeCode Syntology ranMethod as code

In Syntology View this paper on Syntology, its page in Syntology's graph. That page adds which of its functions are also present in other papers' code Syntology holds, with dates, and its dated run record as JSON and a README badge. It also lists the repositories linked to the paper, the abstract and the calls for agents.

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

For agents, on Syntology's MCP service (how to connect):

Code

Syntology 21 code samples: 19 from 1 repository linked to this paper, 2 identical to code harvested from another paper's repository. Across all 21: 9 ran (of which 6 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 2 where Syntology's instrument failed) · 12 unverified. Where Syntology's instrument failed: 1 where our contract draft was wrong; 1 where the contract check could not run; failures of Syntology's instrument, not of the code.

By repository: official repository: 19 samples from 1 repository, 7 ran (7 with no instrument failure; 0 where Syntology's instrument failed); 2 identical to code first harvested elsewhere. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

Levels written between 2026-09-22 and 2026-09-24 (UTC): the dates Syntology's graph wrote each sample's verification level, not when the code executed. 1 of the 9 samples that ran has no recorded date.

adreamwu/ptq4dit officialmentioned in papermentioned on GitHubpytorchSyntology: 19 samples, 7 ran (0 instrument-failed) · 12 unverified 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

21 samples harvested: 9 ran (of which 6 constructed an object rather than computing a result; 7 with no instrument failure: 0 honoured, 0 violated, 7 with no contract checked; 2 where Syntology's instrument failed) · 12 unverified. “Where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Syntology's record for this page has not changed since , the first build that kept a record date for it; when this build read Syntology's graph is in the build record. That date is not when the samples ran.

1ran · our draft was wrong
1ran; the contract check could not run
7ran
12unverified

Licence: 21 of the 21 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 adreamwu/ptq4dit. 2 samples are listed last, in a group without a repository. “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. A line beginning “fixture:” under a sample that ran gives the arguments Syntology's fixture called the sample with, each as name:type/rank/element kind (? where none was recorded), and the type of what the fixture's run returned, exactly as the harness recorded them; the arguments and the returned type were recorded by two separate runs of the same fixture. No output value is shown, and a returned type says nothing about whether the output was right. “Level recorded” gives the UTC date when Syntology's graph wrote the sample's verification level, not when the code executed; where the level was written before Syntology's graph kept that date, the row says “level recorded, date not recorded”. For 1 sample that ran, the graph records a signature this page does not show: for a class the fixture built without calling, the harness called the object with inputs it made up itself, so the recorded signature is the harness's probe, not the fixture's arguments. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself. “Our draft was wrong” and “the contract check could not run” are failures of Syntology's instrument, not of the code: the sample ran on Syntology's synthesized input, and then the drafted contract did not describe it, or the check itself raised an error (RAISES), lacked a dependency (DEP_MISSING) or ran out of time (TIMEOUT); hover the label for which. A metamorphic tier, where shown, is evidence of the output's shape, not that the sample implements the method.

Each sample ends with its code_sha256. A sample's code_sha256 is Syntology's identity for that exact code. It is printed in full (16 hexadecimal characters for every sample in this build), and get_code(code_sha256="…") takes it exactly as printed. An agent calls it on Syntology's MCP service (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 are grouped by repository: a repository marked official first, then by how directly the repository is tied to the paper, then in the archive's order; never by how many ran. 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 six 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; the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar); this paper's copy was not located, so its licence is not recorded; or this paper's copy is cleared, but the stored copy of the same code that Syntology serves is not. 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.

A function or class name links to its page when the name occurs in the harvested code of at least 5 papers; the name is a string, not an identity. A census tag appears when Syntology's census ran implementations with that name on one shared input per bucket: it gives how many were compared, from how many papers, and how many distinct outputs they produced, and names this sample's bucket when this sample was one of them.

adreamwu/ptq4dit official repository 19 samples, 7 ran (0 instrument-failed) · 12 unverified
Attention in 344 papers
census: 13 implementations from 12 papers, 10 distinct outputs across 4 buckets, one shared input per bucket
quant/layer_recon.py ran · metamorphic tier: deterministic no licence file found · pointer only · 623291d109e82c34 · report fixture: returned Attention · level recorded 2026-09-22
DataSaverHook quant/layer_recon.py ran no licence file found · pointer only · 1d9158f89f879a34 · report fixture: returned DataSaverHook · level recorded 2026-09-22
LinearTempDecay quant/layer_recon.py ran no licence file found · pointer only · 4c33b681b95218bb · report fixture: returned LinearTempDecay · level recorded 2026-09-22
Mlp in 102 papers quant/layer_recon.py ran · metamorphic tier: deterministic no licence file found · pointer only · 2cfea2de6e135773 · report fixture: returned Mlp · level recorded 2026-09-22
StopForwardException quant/layer_recon.py ran · metamorphic tier: deterministic no licence file found · pointer only · 13900bea8f3b93a1 · report fixture: returned StopForwardException · level recorded 2026-09-22
StraightThrough quant/layer_recon.py ran fingerprinted no licence file found · pointer only · 382e4d6d6d5d10b0 · report fixture: signature input:Tensor/4/float · returned Tensor · level recorded 2026-09-24
UniformAffineQuantizer in 11 papers quant/layer_recon.py ran · metamorphic tier: deterministic fingerprinted no licence file found · pointer only · f237f023cdb441a9 · report fixture: returned UniformAffineQuantizer · level recorded 2026-09-22
AdaRoundQuantizer quant/layer_recon.py unverified no licence file found · pointer only · e18ddf8148f56942 · report
BaseQuantBlock quant/layer_recon.py unverified no licence file found · pointer only · 7eb71652f7925d3a · report
DiTBlock in 17 papers quant/layer_recon.py unverified no licence file found · pointer only · 31126f616a1300f1 · report
GetLayerInpOut quant/layer_recon.py unverified no licence file found · pointer only · 985822f943599470 · report
LossFunction in 8 papers quant/layer_recon.py unverified no licence file found · pointer only · cc1d937a9352676a · report
QuantDiTBlock quant/layer_recon.py unverified no licence file found · pointer only · dacb0f6316a59775 · report
QuantFinalLayer quant/layer_recon.py unverified no licence file found · pointer only · 7c760bf5741cff4d · report
QuantModel quant/layer_recon.py unverified no licence file found · pointer only · 7e5d23a4a2b80dbd · report
QuantModule quant/layer_recon.py unverified no licence file found · pointer only · 384bed9ea057ffc6 · report
get_specials quant/layer_recon.py unverified no licence file found · pointer only · 190cc4466b584805 · report
layer_reconstruction quant/layer_recon.py unverified no licence file found · pointer only · 422a5c8d1d390767 · report
save_inp_oup_data quant/layer_recon.py unverified no licence file found · pointer only · 5d18164308d717eb · report
Repository not located 2 samples, 2 ran (2 instrument-failed) · 0 unverified
2 samples are identical to code Syntology harvested from another paper's repository; this paper's own copy was not located, so no repository is named and no licence is recorded.
round_ste in 18 papers
census: 3 implementations from 6 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32
repository not located ran · our draft was wrong fingerprinted licence of this copy not recorded · 675da8641c602b03 · report fixture: signature x:Tensor/2/float · returned Tensor · level recorded, date not recorded
lp_loss in 20 papers
census: 6 implementations from 10 papers, 1 distinct output on one shared input this sample: bucket 1 of 1, arg 1: rank 2, kind float, dtype float32 · arg 2: rank 2, kind float, dtype float32
repository not located ran; the contract check could not run (RAISES) · metamorphic tier: invariant fingerprinted licence of this copy not recorded · 08c6ea7576220a13 · report fixture: signature pred:Tensor/2/float, tgt:Tensor/2/float, p:float/0/float, reduction:str/0/str · returned Tensor · level recorded 2026-09-22

Method as code Syntology

What Syntology's graph records as this paper's claimed method, lifted into code: each assembly names its bricks (functions and classes), each with its own role and its own verification level. An assembly is not a check that the code implements the method correctly or reproduces the paper's results. Levels are each brick's own verification level in Syntology's graph (level 0: no recorded run), listed brick by brick and never combined into one level; “ran” means the brick executed on a synthesized input, not that its output is correct, and the instrument figure counts failures of Syntology's instrument, not of the code. Where a model read the code against the paper's claim, each assembly's label verdict is one model's judgement, kept in this page's JSON twin; it is not calibrated, so no card prints it.

1 assembly on this paper: 1 harvested, 0 generated.

harvested Assembly for PTQ4DiT (graph method: PTQ4DiT). 23 bricks; individually on a fixture: 11 ran (of which 7 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 3 where Syntology's instrument failed) · 12 unverified. The assembly has not been run end-to-end.

  1. layer_reconstruction entry · level 0 · unverified · 422a5c8d1d390767
  2. AdaRoundQuantizer helper · level 0 · unverified · e18ddf8148f56942
  3. LossFunction loss · level 0 · unverified · cc1d937a9352676a
  4. QuantModel helper · level 0 · unverified · 7e5d23a4a2b80dbd
  5. QuantModule model · level 0 · unverified · 384bed9ea057ffc6
  6. StraightThrough helper · level 1 · ran · 382e4d6d6d5d10b0
  7. save_inp_oup_data helper · level 0 · unverified · 5d18164308d717eb
  8. UniformAffineQuantizer model · level 1 · ran · f237f023cdb441a9
  9. round_ste update_rule · level 2 · ran · our draft was wrong · 675da8641c602b03
  10. LinearTempDecay helper · level 1 · ran · 4c33b681b95218bb
  11. lp_loss metric · level 1 · ran; the contract check could not run (RAISES) · 08c6ea7576220a13
  12. BaseQuantBlock helper · level 0 · unverified · 7eb71652f7925d3a
  13. DiTBlock helper · level 0 · unverified · 31126f616a1300f1
  14. FinalLayer helper · level 1 · ran · 790db4e566777808
  15. QuantDiTBlock helper · level 0 · unverified · dacb0f6316a59775
  16. QuantFinalLayer helper · level 0 · unverified · 7c760bf5741cff4d
  17. get_specials helper · level 0 · unverified · 190cc4466b584805
  18. GetLayerInpOut helper · level 0 · unverified · 985822f943599470
  19. Attention helper · level 1 · ran · 623291d109e82c34
  20. Mlp helper · level 1 · ran · 2cfea2de6e135773
  21. modulate helper · level 1 · ran · our draft was wrong · 03310bba324ae4fb
  22. DataSaverHook helper · level 1 · ran · 1d9158f89f879a34
  23. StopForwardException helper · level 1 · ran · 13900bea8f3b93a1

Each brick ends with its code_sha256, Syntology's identity for that exact code; click one to copy the get_code call that returns it. For agents, on Syntology's MCP service: assembly(arxiv_id="2405.16005") returns these assemblies with each brick's role and wiring (how to connect).

Tasks

Image GenerationQuantization

Results from the paper archive 2025-07-28

No leaderboard rows for this paper in the archive.

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