Pith. sign in

REVIEW 3 major objections 5 minor 6 cited by

1.58-bit FLUX

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that FLUX.1-dev, a leading text-to-image model, can be quantized to 1.58-bit weights—each linear-layer weight restricted to {-1, 0, +1}—while keeping 1024x1024 generation quality comparable to the original, cutting…

desk verdict A new and likely useful ternary-PTQ result for FLUX.1-dev, but the paper as written omits the quantization method, so the central claim is unverifiable from the text. read the letter →

arxiv 2412.18653 v1 pith:XHRWC26I submitted 2024-12-24 cs.CV cs.AIcs.LG

Chenglin Yang , Celong Liu , Xueqing Deng , Dongwon Kim , Xing Mei , Xiaohui Shen , Liang-Chieh Chen This is my paper · ORCID
classification cs.CVcs.AIcs.LG
keywords 1.58-bitquantizationternaryweightstext-to-imagegenerationFLUX.1-devdiffusiontransformerpost-trainingmodelcompressioninferenceefficiency
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that the state-of-the-art text-to-image model FLUX.1-dev can be quantized to 1.58-bit weights—each linear-layer weight restricted to -1, 0, or +1—while still producing 1024x1024 images whose quality is comparable to the full-precision model. This would matter because FLUX has 11.9 billion parameters in its transformer, making deployment on phones and edge devices impractical; the proposed compression cuts checkpoint storage 7.7x and inference memory over 5.1x. The method is post-training and requires no image data: only 7,232 text prompts are used to calibrate the quantization, plus a custom kernel that exploits ternary arithmetic to speed up inference. Evaluations on GenEval and T2I CompBench show small drops (0.64 vs 0.66 and 0.5812 vs 0.5876), which the paper interprets as comparable performance. If correct, this would be the first demonstration that a leading open text-to-image model survives near-total weight binarization without retraining from scratch.

What carries the argument

The central object is 1.58-bit quantization of the transformer's linear layers: weights are ternarized to {-1, 0, +1} with a scale factor, so a matrix product becomes sums of added or subtracted selected rows rather than general multiplications, and the weights pack into 2-bit signed integers. A custom kernel is the companion mechanism: it is built to fetch and multiply these packed ternary weights efficiently, which yields the measured memory and latency gains. The third piece is the calibration procedure, a set of 7,232 text-only prompts on which the model's own outputs supply the supervision for choosing quantization parameters, giving the method its image-data-free property.

What would settle it

Take a fixed evaluation set, compute the 1.58-bit model using one 7,232-prompt calibration set and then a disjoint 7,232-prompt set from the same distribution; if the two quantized models' GenEval scores differ by more than the reported gap to FLUX (0.02), the result depends on the particular calibration prompts rather than on the ternary representation itself.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that 99.5% of the FLUX.1-dev vision transformer's 11.9B parameters can be replaced by ternary values in {-1, 0, +1} without mixed-precision schemes or image-data calibration, and the resulting model remains competitive on standard text-to-image benchmarks. Weights are stored as 2-bit signed integers, compressing from 16-bit precision and accounting for the 7.7x storage reduction; a custom linear kernel then executes low-bit operations to reduce inference memory by more than 5.1x and to lower latency, especially on weaker GPUs such as L20 and A10. The authors report GenEval scores of 0.64 for the quantized model versus 0.66 for FLUX, and T2I CompBench averages of 0.5812 versus 0.5876, with visual comparisons generated from the same latent noise shown as qualitatively close. The quantization is calibrated from 7,232 text prompts drawn from Parti-1k and the T2I CompBench training split, using the model's own forward behavior on those prompts as supervision and no images.

Load-bearing premise

The load-bearing premise is that 7,232 text-only prompts, with no images, give enough signal to set the quantization scales and rounding choices so the model's behavior across all prompts stays close to the full-precision model.

Editorial extensions

If this is right

  • A 7.7x storage reduction means the FLUX checkpoint fits in roughly one-eighth the space, easing distribution and loading on devices with limited flash memory.
  • A 5.1x inference-memory reduction makes the model runnable on GPUs that previously ran out of memory, as shown by the A10 result where FLUX runs out of memory and 1.58-bit FLUX completes generation.
  • Latency improves most on lower-end GPUs (13.2% on L20, and A10 becomes usable), while high-end GPUs see small gains because this version quantizes weights but not activations.
  • The benchmark numbers imply that everyday compositional prompts, counts, colors, and spatial relations survive the compression nearly intact, with the largest reported GenEval gap in the two-object category (0.81 to 0.77).
  • Because calibration needs only text prompts, the same image-data-free workflow could in principle be repeated for other publicly available diffusion transformers without collecting paired image datasets.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the calibration signal is genuinely model-side rather than data-side, the same recipe may transfer to other diffusion transformers: any text-to-image model with public weights could be ternarized using only text prompts, provided its forward pass can serve as the self-supervision source.
  • Because activation quantization is absent, the current latency gains understate what a full low-bit pipeline could deliver; quantizing activations and the time-step embeddings would attack the remaining speed bottleneck, which the paper itself flags as future work.
  • The reported benchmark gaps are averages, and a per-prompt stress test on fine-grained details such as small text, faces, or high-frequency textures might reveal where ternary weights fail despite the comparable aggregate scores—the paper's own limitation section notes that high-resolution detail still lags.
  • A practical extension would be measuring how quality depends on calibration size: how few prompts suffice before GenEval drops measurably, which would tell deployment engineers how much calibration data the method really needs.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes 1.58-bit FLUX, a post-training quantization method that restricts the linear-layer weights of FLUX.1-dev to the values {-1, 0, +1}, covering 99.5% of the vision transformer's 11.9B parameters. Calibration is claimed to be image-data-free, using 7,232 text prompts from Parti-1k and the T2I CompBench training split. The paper reports comparable GenEval and T2I CompBench scores, a 7.7x storage reduction, a 5.1x inference-memory reduction, and modest latency improvements, supported by a custom 1.58-bit kernel. The main scientific claim is that an 11.9B-parameter text-to-image model can be ternarized without image data while retaining generation quality.

Significance. If substantiated, this result would be significant for efficient deployment of large text-to-image models: a 7.7x checkpoint reduction and a 5.1x inference-memory reduction at small benchmark-score drops would be a practical step toward on-device generation. The idea of calibrating ternary quantization with text prompts alone is valuable and worth testing. The manuscript does not, however, provide a reproducible method: no quantization formula, no scale computation, no rounding rule, and no kernel arithmetic are given, and the empirical evaluation lacks variance estimates and has a potential calibration/evaluation overlap. The authors do include an explicit limitations section conceding fine-detail quality loss at high resolution, which is commendable but does not compensate for the missing methodological detail.

major comments (3)
  1. [Section 3.1, 'Quantization'] The central method is not described. After stating that weights of all linear layers in FluxTransformerBlock and FluxSingleTransformerBlock are reduced to 1.58 bits, the paper does not specify the ternary assignment rule (e.g., absmean scaling, magnitude thresholding, or optimization-based rounding), whether per-tensor or per-channel scale factors are used, how the 7,232 calibration prompts are processed (text embeddings only, random latents, generated images, which timesteps), or how biases, normalization layers, embeddings, and the output projection are handled. The custom kernel's arithmetic, including how ternary weights and scale factors are combined, is also absent. Without this specification, Tables 1 and 2 cannot be reproduced and the claim that the weights are truly restricted to {-1, 0, +1} cannot be audited. This omission is load-bearing because the 'first successful approach' is the paper's core contribution.
  2. [Tables 1 and 2; Section 3.1, 'Evaluation'] The evaluation does not support the strength of the 'comparable performance' claim. The reported differences are small (GenEval overall 0.66 to 0.64; T2I CompBench average 0.5876 to 0.5812), but no error bars, confidence intervals, or significance tests are provided, and the number of independent runs and random seeds are not stated. Moreover, the calibration set uses the training split of T2I CompBench while the evaluation is on the validation split of the same benchmark. This overlap means the measured performance may reflect calibration-distribution familiarity rather than general image-data-free quantization. The paper should report results with a calibration set that excludes the evaluation benchmark, or otherwise quantify sensitivity to the calibration distribution.
  3. [Figure 2 and Table 3] The efficiency claims are not sufficiently specified to be audited. The paper does not state how storage is measured (checkpoint file size versus raw tensor bytes), how inference memory is measured (peak CUDA memory, resident set size, or reported by a profiling tool), or whether the reported latency includes kernel launch overhead, CPU-GPU transfer, and the same sampling parameters (steps, guidance, batch size) for both models. The 7.7x storage reduction is close to the theoretical 8x from 16-bit to 2-bit storage, so it is plausible, but the 5.1x inference-memory reduction needs a clear accounting of which tensors are stored at 2 bits and which remain full precision. Without this methodology, the efficiency numbers cannot be independently verified.
minor comments (5)
  1. [Abstract and Section 1] The term '1.58-bit' is used without a definition; the paper should clarify that this follows the BitNet b1.58 nomenclature and that the actual storage format is 2-bit signed integers, since a reader may otherwise expect a non-integer bit width in the checkpoint format.
  2. [Tables 1 and 2] The rows '1.58-bit FLUX (w/o kernel)' and '1.58-bit FLUX' differ slightly, yet the kernel is described as an implementation of the same arithmetic. The authors state this demonstrates implementation accuracy, but the observed differences (e.g., GenEval Counting 0.69 vs. 0.68) are not explained; the paper should state whether these are due to numerical non-associativity, nondeterminism, or an actual behavioral difference.
  3. [References [26] and [41]] Reference [26] (Ideogram) and reference [41] (Midjourney) both point to the same URL, https://updates.midjourney.com/version-6-1/. The Ideogram reference should be corrected or removed.
  4. [Figure 2] The x-axis labels 'm-nG' are difficult to parse; defining the convention (GPU type m with maximum memory n gigabytes) directly in the caption would improve readability.
  5. [Section 4, 'Limitations on visual qualities'] The limitations section concedes that 1.58-bit FLUX 'still lags behind the original FLUX model in rendering fine details at very high resolutions.' This important caveat should also appear in the abstract or introduction, since the abstract currently claims 'maintaining comparable performance' without this qualification.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: benchmark scores are measurements; the missing quantization specification is a reproducibility concern, not circularity.

full rationale

No circular step is exhibited. The paper's central claim is an empirical result: after post-training ternary quantization, FLUX's generation quality remains comparable as measured on GenEval and T2I CompBench, and its efficiency improves as measured by storage, memory, and latency. These values are measurements, not quantities derived from the quantization definition by construction. The 7.7x storage reduction follows arithmetically from the stated 99.5% coverage of ternary weights stored as 2-bit signed integers, rather than from the benchmark scores. Calibration uses prompts from Parti-1k and the CompBench training split, while the headline CompBench numbers are on the validation split and GenEval is a separate external benchmark, so the evaluation is held-out rather than statistically forced by the calibration input. No load-bearing self-citation appears; the BitNet b1.58 reference supplies the ternary value scheme by analogy, but the FLUX-specific results are independently measured. The paper's failure to specify the actual quantization algorithm is a serious reproducibility and correctness concern, but an omitted method is not a circular derivation: there is no equation, fitted constant, or defined quantity that is renamed as a prediction. Therefore the circularity score is 0.

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The paper introduces no new entities. The core unstated assumptions are that ternarization with unspecified per-layer scales preserves FLUX's behavior and that a text-only calibration set is a sufficient proxy for image-level quality. The quantization scales, if any, are free parameters never reported.

free parameters (1)
  • per-layer quantization scales = not reported
    Every ternary PTQ method needs at least one scale per layer to map real weights into {-1,0,+1}; the paper never states how these are computed. They are fitted on the 7,232-prompt calibration set, so they are free parameters with undisclosed values.
assumptions (3)
  • domain assumption FLUX.1-dev's transformer weights can be ternarized with no retraining and no image data while preserving output distribution.
    Core premise of the paper; no derivation or ablation in the text supports it.
  • domain assumption Text-prompt self-supervision from the model itself provides sufficient calibration signal.
    The abstract and Section 3.1 assert image-data-free calibration but never define the self-supervision mechanism, such as hidden activations or losses.
  • domain assumption Benchmark scores from single runs are stable enough to support 'comparable' claims.
    No error bars, seeds, or significance tests are reported in Tables 1 and 2.

how reviews work

0 comments
Cite this review

Pith. "Pith review of 1.58-bit FLUX." pith.science (2026). https://pith.science/paper/XHRWC26I

@misc{pith2026241218653,
  author       = {Pith},
  title        = {Pith review of: 1.58-bit FLUX},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XHRWC26I}},
  note         = {Machine review of arXiv:2412.18653}
}
read the original abstract

We present 1.58-bit FLUX, the first successful approach to quantizing the state-of-the-art text-to-image generation model, FLUX.1-dev, using 1.58-bit weights (i.e., values in {-1, 0, +1}) while maintaining comparable performance for generating 1024 x 1024 images. Notably, our quantization method operates without access to image data, relying solely on self-supervision from the FLUX.1-dev model. Additionally, we develop a custom kernel optimized for 1.58-bit operations, achieving a 7.7x reduction in model storage, a 5.1x reduction in inference memory, and improved inference latency. Extensive evaluations on the GenEval and T2I Compbench benchmarks demonstrate the effectiveness of 1.58-bit FLUX in maintaining generation quality while significantly enhancing computational efficiency.

Figures

Figures reproduced from arXiv: 2412.18653 by the authors.

Figure 1
Figure 1. Visual comparisons between FLUX and 1.58-bit FLUX. 1.58-bit FLUX demonstrates comparable generation quality to FLUX while employing 1.58-bit quantization, where 99.5% of the 11.9B parameters in the vision transformer are constrained to the values +1, -1, or 0. For consistency, all images in each comparison are generated using the same latent noise input. 1.58-bit FLUX utilizes a custom 1.58-bit kernel. Additional vi… view at source ↗
Figure 2
Figure 2. Efficiency measurements on the vision transformer component of FLUX and 1.58-bit FLUX. The measurements are based on generating a single image with 50 inference steps. (a) 1.58-bit FLUX reduces checkpoint storage by 7.7× compared to FLUX. (b) 1.58-bit FLUX achieves a 5.1× reduction in inference memory usage across various GPU types. The x-axis labels, m-nG, represent GPU type m with a maximum memory capacity of n Gi… view at source ↗
Figure 3
Figure 3. Visual comparisons between FLUX and 1.58-bit FLUX on GenEval dataset. 1.58-bit FLUX demonstrates comparable generation quality to FLUX while employing 1.58-bit quantization, where 99.5% of the 11.9B parameters in the vision transformer are constrained to the values +1, -1, or 0. For consistency, all images in each comparison are generated using the same latent noise input. 1.58-bit FLUX utilizes a custom 1.58-bit ke… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Visual comparisons between FLUX and 1.58-bit FLUX on the validation split of T2I CompBench. 1.58-bit FLUX demon￾strates comparable generation quality to FLUX while employing 1.58-bit quantization, where 99.5% of the 11.9B parameters in the vision transformer are constr…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 6 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Investigation of Accuracy and Bias in Face Recognition Trained with Synthetic Data

    cs.CV 2025-07 conditional novelty 6.0 of 10

    Balanced synthetic face data from Stable Diffusion v3.5 reduces racial bias in face recognition models but does not yet match real-data accuracy on hard benchmarks.

  2. Retrieval Visual Contrastive Decoding to Mitigate Object Hallucinations in Large Vision-Language Models

    cs.CV 2025-05 conditional novelty 6.0 of 10

    RVCD uses YOLO detections and retrieved single-concept AI images to adjust LVLM logits at decode time, cutting CHAIR hallucination rates by roughly half versus prior contrastive decoding baselines.

  3. Grouping First, Attending Smartly: Training-Free Acceleration for Diffusion Transformers

    cs.CV 2025-05 conditional novelty 6.0 of 10

    GRAT accelerates pretrained diffusion transformers by grouping tokens and restricting each group's attention to neighboring blocks or criss-cross rows and columns, achieving large speedups with near-full-attention quality.

  4. Democratizing Text-to-Image Masked Generative Models with Compact Text-Aware One-Dimensional Tokens

    cs.CV 2025-01 conditional novelty 6.0 of 10

    A text-aware 1D tokenizer and a masked generative model, trained entirely on public data, reach FID and GenEval scores comparable to larger private-data diffusion and autoregressive models.

  5. ID-Card Synthetic Generation: Toward a Simulated Bona fide Dataset

    cs.CV 2025-08 conditional novelty 5.0 of 10

    Synthetic ID cards made with Stable Diffusion and hybrid face/text compositing are recognized as bona fide by a commercial PAD system and could augment scarce real training data.

  6. DFVEdit: Conditional Delta Flow Vector for Zero-shot Video Editing

    cs.CV 2025-06 conditional novelty 4.0 of 10

    DFVEdit edits videos by iteratively subtracting a conditional delta flow vector, the difference between the model's predictions under the target and source prompts, from the latent representation of the source video.

Reference graph

Works this paper leans on

63 extracted references · 25 canonical work pages · cited by 6 Pith papers

  1. [1]

    https://news.adobe.com/news/news- details/2024/adobe-introduces-firefly-image-3-foundation- model-to-take-creative-exploration-and-ideation-to-new- heights

    Adobe. https://news.adobe.com/news/news- details/2024/adobe-introduces-firefly-image-3-foundation- model-to-take-creative-exploration-and-ideation-to-new- heights. 2024. 2

    work page 2024

  2. [2]

    https://www.recraft.ai/blog/recraft-introduces- a-revolutionary-ai-model-that-thinks-in-design-language

    Recraft AI. https://www.recraft.ai/blog/recraft-introduces- a-revolutionary-ai-model-that-thinks-in-design-language

    work page

  3. [3]

    Improving image generation with better captions

    James Betker, Gabriel Goh, Li Jing, Tim Brooks, Jianfeng Wang, Linjie Li, Long Ouyang, Juntang Zhuang, Joyce Lee, Yufei Guo, et al. Improving image generation with better captions. Computer Science. https://cdn. openai. com/papers/dall-e-3. pdf, 2(3):8, 2023. 2

    2023

  4. [4]

    Zeroq: A novel zero shot quantization framework

    Yaohui Cai, Zhewei Yao, Zhen Dong, Amir Gholami, Michael W Mahoney, and Kurt Keutzer. Zeroq: A novel zero shot quantization framework. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13169–13178, 2020. 3

    work page 2020

  5. [5]

    Low-bitwidth floating point quantization for efficient high- quality diffusion models

    Cheng Chen, Christina Giannoula, and Andreas Moshovos. Low-bitwidth floating point quantization for efficient high- quality diffusion models. arXiv preprint arXiv:2408.06995,

    arXiv

  6. [6]

    Pixart- alpha: Fast training of diffusion transformer for photorealistic text-to-image synthesis

    Junsong Chen, Jincheng Yu, Chongjian Ge, Lewei Yao, Enze Xie, Yue Wu, Zhongdao Wang, James Kwok, Ping Luo, Huchuan Lu, et al. Pixart- alpha: Fast training of diffusion transformer for photorealistic text-to-image synthesis. arXiv preprint arXiv:2310.00426, 2023. 3

    arXiv

  7. [7]

    Q-dit: Ac- curate post-training quantization for diffusion transformers

    Lei Chen, Yuan Meng, Chen Tang, Xinzhu Ma, Jingyan Jiang, Xin Wang, Zhi Wang, and Wenwu Zhu. Q-dit: Ac- curate post-training quantization for diffusion transformers. arXiv preprint arXiv:2406.17343, 2024. 3

    arXiv 2024

  8. [8]

    Prefixquant: Static quantization beats dy- namic through prefixed outliers in llms

    Mengzhao Chen, Yi Liu, Jiahao Wang, Yi Bin, Wenqi Shao, and Ping Luo. Prefixquant: Static quantization beats dy- namic through prefixed outliers in llms. arXiv preprint arXiv:2410.05265, 2024. 3

    arXiv 2024

Show all 63 references
  1. [9]

    Efficientqat: Efficient quantization-aware training for large language models.arXiv preprint arXiv:2407.11062, 2024

    Mengzhao Chen, Wenqi Shao, Peng Xu, Jiahao Wang, Peng Gao, Kaipeng Zhang, and Ping Luo. Efficientqat: Efficient quantization-aware training for large language models.arXiv preprint arXiv:2407.11062, 2024. 3

    2024 arXiv

  2. [10]

    Low-bit quantization of neural networks for efficient infer- ence

    Yoni Choukroun, Eli Kravchik, Fan Yang, and Pavel Kisilev. Low-bit quantization of neural networks for efficient infer- ence. In 2019 IEEE/CVF International Conference on Com- puter Vision Workshop (ICCVW) , pages 3009–3018. IEEE,

    2019

  3. [11]

    Vq4dit: Efficient post-training vec- tor quantization for diffusion transformers

    Juncan Deng, Shuaiting Li, Zeyu Wang, Hong Gu, Kedong Xu, and Kejie Huang. Vq4dit: Efficient post-training vec- tor quantization for diffusion transformers. arXiv preprint arXiv:2408.17131, 2024. 3

    2024 arXiv

  4. [12]

    Tim Dettmers, Mike Lewis, Younes Belkada, and Luke Zettlemoyer. Gpt3. int8 (): 8-bit matrix multiplication for transformers at scale. Advances in Neural Information Pro- cessing Systems, 35:30318–30332, 2022. 3

    2022

  5. [13]

    Hawq: Hessian aware quantization of neural networks with mixed-precision

    Zhen Dong, Zhewei Yao, Amir Gholami, Michael W Ma- honey, and Kurt Keutzer. Hawq: Hessian aware quantization of neural networks with mixed-precision. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, pages 293–302, 2019. 3

    2019

  6. [14]

    Scaling recti- fied flow transformers for high-resolution image synthesis

    Patrick Esser, Sumith Kulal, Andreas Blattmann, Rahim Entezari, Jonas M ¨uller, Harry Saini, Yam Levi, Dominik Lorenz, Axel Sauer, Frederic Boesel, et al. Scaling recti- fied flow transformers for high-resolution image synthesis. In Forty-first International Conference on Mach...

    2024

  7. [15]

    Learned step size quantization

    Steven K Esser, Jeffrey L McKinstry, Deepika Bablani, Rathinakumar Appuswamy, and Dharmendra S Modha. Learned step size quantization. arXiv preprint arXiv:1902.08153, 2019. 3

    1902 arXiv

  8. [16]

    Gptq: Accurate post-training quantization for generative pre-trained transformers

    Elias Frantar, Saleh Ashkboos, Torsten Hoefler, and Dan Alistarh. Gptq: Accurate post-training quantization for generative pre-trained transformers. arXiv preprint arXiv:2210.17323, 2022. 3 6

    2022 arXiv

  9. [17]

    Geneval: An object-focused framework for evaluating text- to-image alignment

    Dhruba Ghosh, Hannaneh Hajishirzi, and Ludwig Schmidt. Geneval: An object-focused framework for evaluating text- to-image alignment. Advances in Neural Information Pro- cessing Systems, 36, 2024. 2, 4

    2024

  10. [18]

    Differ- entiable soft quantization: Bridging full-precision and low- bit neural networks

    Ruihao Gong, Xianglong Liu, Shenghu Jiang, Tianxiang Li, Peng Hu, Jiazhen Lin, Fengwei Yu, and Junjie Yan. Differ- entiable soft quantization: Bridging full-precision and low- bit neural networks. In Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision, pa...

  11. [19]

    Deep com- pression: Compressing deep neural networks with pruning, trained quantization and huffman coding

    Song Han, Huizi Mao, and William J Dally. Deep com- pression: Compressing deep neural networks with pruning, trained quantization and huffman coding. arXiv preprint arXiv:1510.00149, 2015. 3

    2015 arXiv

  12. [20]

    Efficientdm: Efficient quantization-aware fine-tuning of low-bit diffusion models

    Yefei He, Jing Liu, Weijia Wu, Hong Zhou, and Bo- han Zhuang. Efficientdm: Efficient quantization-aware fine-tuning of low-bit diffusion models. arXiv preprint arXiv:2310.03270, 2023. 3

    2023 arXiv

  13. [21]

    Bivit: Extremely com- pressed binary vision transformers

    Yefei He, Zhenyu Lou, Luoming Zhang, Jing Liu, Weijia Wu, Hong Zhou, and Bohan Zhuang. Bivit: Extremely com- pressed binary vision transformers. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 5651–5663, 2023. 3

    2023

  14. [22]

    Ptqd: Accurate post-training quantization for diffusion models

    Yefei He, Luping Liu, Jing Liu, Weijia Wu, Hong Zhou, and Bohan Zhuang. Ptqd: Accurate post-training quantization for diffusion models. Advances in Neural Information Pro- cessing Systems, 36, 2024. 3

    2024

  15. [23]

    T2i-compbench: A comprehensive bench- mark for open-world compositional text-to-image genera- tion

    Kaiyi Huang, Kaiyue Sun, Enze Xie, Zhenguo Li, and Xihui Liu. T2i-compbench: A comprehensive bench- mark for open-world compositional text-to-image genera- tion. Advances in Neural Information Processing Systems , 36:78723–78747, 2023. 2, 3, 4

    2023

  16. [24]

    Tfmq-dm: Temporal feature maintenance quantization for diffusion models

    Yushi Huang, Ruihao Gong, Jing Liu, Tianlong Chen, and Xianglong Liu. Tfmq-dm: Temporal feature maintenance quantization for diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 7362–7371, 2024. 3

    2024

  17. [25]

    Binarized neural networks

    Itay Hubara, Matthieu Courbariaux, Daniel Soudry, Ran El- Yaniv, and Yoshua Bengio. Binarized neural networks. Ad- vances in Neural Information Processing Systems, 29, 2016. 3

    2016

  18. [26]

    https://updates.midjourney.com/version-6-1/

    Ideogram. https://updates.midjourney.com/version-6-1/

  19. [27]

    Quantization and training of neural networks for efficient integer-arithmetic-only inference

    Benoit Jacob, Skirmantas Kligys, Bo Chen, Menglong Zhu, Matthew Tang, Andrew Howard, Hartwig Adam, and Dmitry Kalenichenko. Quantization and training of neural networks for efficient integer-arithmetic-only inference. In Proceed- ings of the IEEE Conference on Computer Vision ...

    2018

  20. [28]

    Squeezellm: Dense-and-sparse quantization

    Sehoon Kim, Coleman Hooper, Amir Gholami, Zhen Dong, Xiuyu Li, Sheng Shen, Michael W Mahoney, and Kurt Keutzer. Squeezellm: Dense-and-sparse quantization. arXiv preprint arXiv:2306.07629, 2023. 3

    2023 arXiv

  21. [29]

    https://blackforestlabs.ai/announcements/

    Black Forest Labs. https://blackforestlabs.ai/announcements/

  22. [30]

    https://github.com/black-forest-labs/flux

    Black Forest Labs. https://github.com/black-forest-labs/flux

  23. [31]

    Playground v2

    Daiqing Li, Aleks Kamko, Ehsan Akhgari, Ali Sabet, Lin- miao Xu, and Suhail Doshi. Playground v2. 5: Three in- sights towards enhancing aesthetic quality in text-to-image generation. arXiv preprint arXiv:2402.17245, 2024. 3

    2024 arXiv

  24. [32]

    Svdquant: Absorbing outliers by low- rank components for 4-bit diffusion models

    Muyang Li*, Yujun Lin*, Zhekai Zhang*, Tianle Cai, Xi- uyu Li, Junxian Guo, Enze Xie, Chenlin Meng, Jun-Yan Zhu, and Song Han. Svdquant: Absorbing outliers by low- rank components for 4-bit diffusion models. arXiv preprint arXiv:2411.05007, 2024. 3

    2024

  25. [33]

    Q-diffusion: Quantizing diffusion models

    Xiuyu Li, Yijiang Liu, Long Lian, Huanrui Yang, Zhen Dong, Daniel Kang, Shanghang Zhang, and Kurt Keutzer. Q-diffusion: Quantizing diffusion models. In Proceedings of the IEEE/CVF International Conference on Computer Vi- sion, pages 17535–17545, 2023. 3

    2023

  26. [34]

    Awq: Activation-aware weight quantization for on-device llm compression and acceleration

    Ji Lin, Jiaming Tang, Haotian Tang, Shang Yang, Wei-Ming Chen, Wei-Chen Wang, Guangxuan Xiao, Xingyu Dang, Chuang Gan, and Song Han. Awq: Activation-aware weight quantization for on-device llm compression and acceleration. Proceedings of Machine Learning and Systems , 6:87–100,

  27. [35]

    Qserve: W4a8kv4 quantization and system co-design for efficient llm serving

    Yujun Lin, Haotian Tang, Shang Yang, Zhekai Zhang, Guangxuan Xiao, Chuang Gan, and Song Han. Qserve: W4a8kv4 quantization and system co-design for efficient llm serving. arXiv preprint arXiv:2405.04532, 2024. 3

    2024 arXiv

  28. [36]

    Playground v3: Im- proving text-to-image alignment with deep-fusion large lan- guage models

    Bingchen Liu, Ehsan Akhgari, Alexander Visheratin, Aleks Kamko, Linmiao Xu, Shivam Shrirao, Chase Lambert, Joao Souza, Suhail Doshi, and Daiqing Li. Playground v3: Im- proving text-to-image alignment with deep-fusion large lan- guage models. arXiv preprint arXiv:2409.10695, 2024. 2

    2024 arXiv

  29. [37]

    Hq-dit: Efficient diffu- sion transformer with fp4 hybrid quantization.arXiv preprint arXiv:2405.19751, 2024

    Wenxuan Liu and Sai Qian Zhang. Hq-dit: Efficient diffu- sion transformer with fp4 hybrid quantization.arXiv preprint arXiv:2405.19751, 2024. 3

    2024 arXiv

  30. [38]

    En- hanced distribution alignment for post-training quantization of diffusion models

    Xuewen Liu, Zhikai Li, Junrui Xiao, and Qingyi Gu. En- hanced distribution alignment for post-training quantization of diffusion models. arXiv preprint arXiv:2401.04585, 2024. 3

    2024 arXiv

  31. [39]

    Llm-qat: Data-free quantization aware training for large language models.arXiv preprint arXiv:2305.17888, 2023

    Zechun Liu, Barlas Oguz, Changsheng Zhao, Ernie Chang, Pierre Stock, Yashar Mehdad, Yangyang Shi, Raghuraman Krishnamoorthi, and Vikas Chandra. Llm-qat: Data-free quantization aware training for large language models.arXiv preprint arXiv:2305.17888, 2023. 3

    2023 arXiv

  32. [40]

    The era of 1-bit llms: All large language models are in 1.58 bits

    Shuming Ma, Hongyu Wang, Lingxiao Ma, Lei Wang, Wen- hui Wang, Shaohan Huang, Li Dong, Ruiping Wang, Jilong Xue, and Furu Wei. The era of 1-bit llms: All large language models are in 1.58 bits. arXiv preprint arXiv:2402.17764 ,

    arXiv

  33. [41]

    https://updates.midjourney.com/version-6-1/

    Midjourney. https://updates.midjourney.com/version-6-1/

  34. [42]

    Up or down? adap- tive rounding for post-training quantization

    Markus Nagel, Rana Ali Amjad, Mart Van Baalen, Chris- tos Louizos, and Tijmen Blankevoort. Up or down? adap- tive rounding for post-training quantization. In International Conference on Machine Learning, pages 7197–7206. PMLR,

  35. [43]

    A white paper on neural network quantization

    Markus Nagel, Marios Fournarakis, Rana Ali Amjad, Yely- sei Bondarenko, Mart Van Baalen, and Tijmen Blankevoort. A white paper on neural network quantization. arXiv preprint arXiv:2106.08295, 2021. 3 7

    2021 arXiv

  36. [44]

    Overcoming oscillations in quantization-aware training

    Markus Nagel, Marios Fournarakis, Yelysei Bondarenko, and Tijmen Blankevoort. Overcoming oscillations in quantization-aware training. In International Conference on Machine Learning, pages 16318–16330. PMLR, 2022. 3

    2022

  37. [45]

    Sdxl: Improving latent diffusion mod- els for high-resolution image synthesis

    Dustin Podell, Zion English, Kyle Lacey, Andreas Blattmann, Tim Dockhorn, Jonas M ¨uller, Joe Penna, and Robin Rombach. Sdxl: Improving latent diffusion mod- els for high-resolution image synthesis. arXiv preprint arXiv:2307.01952, 2023. 3

    2023 arXiv

  38. [46]

    Post-training quantization on diffusion models

    Yuzhang Shang, Zhihang Yuan, Bin Xie, Bingzhe Wu, and Yan Yan. Post-training quantization on diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 1972–1981, 2023. 3

    1972

  39. [47]

    Omniquant: Omnidirectionally calibrated quantization for large language models

    Wenqi Shao, Mengzhao Chen, Zhaoyang Zhang, Peng Xu, Lirui Zhao, Zhiqian Li, Kaipeng Zhang, Peng Gao, Yu Qiao, and Ping Luo. Omniquant: Omnidirectionally calibrated quantization for large language models. arXiv preprint arXiv:2308.13137, 2023. 3

    2023 arXiv

  40. [48]

    Bitsfusion: 1.99 bits weight quantization of diffusion model

    Yang Sui, Yanyu Li, Anil Kag, Yerlan Idelbayev, Junli Cao, Ju Hu, Dhritiman Sagar, Bo Yuan, Sergey Tulyakov, and Jian Ren. Bitsfusion: 1.99 bits weight quantization of diffusion model. arXiv preprint arXiv:2406.04333, 2024. 3

    2024 arXiv

  41. [49]

    Post-training quan- tization with progressive calibration and activation relaxing for text-to-image diffusion models

    Siao Tang, Xin Wang, Hong Chen, Chaoyu Guan, Zewen Wu, Yansong Tang, and Wenwu Zhu. Post-training quan- tization with progressive calibration and activation relaxing for text-to-image diffusion models. In European Conference on Computer Vision, pages 404–420. Springer, 2025. 3

    2025

  42. [50]

    Towards accurate post-training quantization for diffusion models

    Changyuan Wang, Ziwei Wang, Xiuwei Xu, Yansong Tang, Jie Zhou, and Jiwen Lu. Towards accurate post-training quantization for diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 16026–16035, 2024. 3

    2024

  43. [51]

    Bitnet: Scaling 1-bit trans- formers for large language models

    H Wang, S Ma, L Dong, S Huang, H Wang, L Ma, F Yang, R Wang, Y Wu, and F Wei. Bitnet: Scaling 1-bit trans- formers for large language models. arxiv. arXiv preprint arXiv:2310.11453, 2023. 3

    2023 arXiv

  44. [52]

    1-bit ai infra: Part 1.1, fast and lossless bitnet b1

    Jinheng Wang, Hansong Zhou, Ting Song, Shaoguang Mao, Shuming Ma, Hongyu Wang, Yan Xia, and Furu Wei. 1-bit ai infra: Part 1.1, fast and lossless bitnet b1. 58 inference on cpus. arXiv preprint arXiv:2410.16144, 2024. 3

    2024 arXiv

  45. [53]

    Haq: Hardware-aware automated quantization with mixed precision

    Kuan Wang, Zhijian Liu, Yujun Lin, Ji Lin, and Song Han. Haq: Hardware-aware automated quantization with mixed precision. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 8612– 8620, 2019. 3

    2019

  46. [54]

    Outlier suppression: Pushing the limit of low-bit transformer language models

    Xiuying Wei, Yunchen Zhang, Xiangguo Zhang, Ruihao Gong, Shanghang Zhang, Qi Zhang, Fengwei Yu, and Xian- glong Liu. Outlier suppression: Pushing the limit of low-bit transformer language models. Advances in Neural Informa- tion Processing Systems, 35:17402–17414, 2022. 3

    2022

  47. [55]

    Ptq4dit: Post-training quantization for diffu- sion transformers

    Junyi Wu, Haoxuan Wang, Yuzhang Shang, Mubarak Shah, and Yan Yan. Ptq4dit: Post-training quantization for diffu- sion transformers. arXiv preprint arXiv:2405.16005, 2024. 3

    2024 arXiv

  48. [56]

    Smoothquant: Accurate and effi- cient post-training quantization for large language models

    Guangxuan Xiao, Ji Lin, Mickael Seznec, Hao Wu, Julien Demouth, and Song Han. Smoothquant: Accurate and effi- cient post-training quantization for large language models. In International Conference on Machine Learning , pages 38087–38099. PMLR, 2023. 3

    2023

  49. [57]

    Efficient quantization strategies for latent diffusion models

    Yuewei Yang, Xiaoliang Dai, Jialiang Wang, Peizhao Zhang, and Hongbo Zhang. Efficient quantization strategies for latent diffusion models. arXiv preprint arXiv:2312.05431 ,

    arXiv

  50. [58]

    Timestep-aware cor- rection for quantized diffusion models

    Yuzhe Yao, Feng Tian, Jun Chen, Haonan Lin, Guang Dai, Yong Liu, and Jingdong Wang. Timestep-aware cor- rection for quantized diffusion models. arXiv preprint arXiv:2407.03917, 2024. 3

    2024 arXiv

  51. [59]

    Zeroquant: Ef- ficient and affordable post-training quantization for large- scale transformers

    Zhewei Yao, Reza Yazdani Aminabadi, Minjia Zhang, Xi- aoxia Wu, Conglong Li, and Yuxiong He. Zeroquant: Ef- ficient and affordable post-training quantization for large- scale transformers. Advances in Neural Information Process- ing Systems, 35:27168–27183, 2022. 3

    2022

  52. [60]

    Scaling autoregres- sive models for content-rich text-to-image generation

    Jiahui Yu, Yuanzhong Xu, Jing Yu Koh, Thang Luong, Gun- jan Baid, Zirui Wang, Vijay Vasudevan, Alexander Ku, Yin- fei Yang, Burcu Karagol Ayan, et al. Scaling autoregres- sive models for content-rich text-to-image generation. arXiv preprint arXiv:2206.10789, 2(3):5, 2022. 3

    2022 arXiv

  53. [61]

    Vidit-q: Efficient and accurate quantization of diffusion transformers for im- age and video generation

    Tianchen Zhao, Tongcheng Fang, Enshu Liu, Wan Rui, Widyadewi Soedarmadji, Shiyao Li, Zinan Lin, Guohao Dai, Shengen Yan, Huazhong Yang, et al. Vidit-q: Efficient and accurate quantization of diffusion transformers for im- age and video generation. arXiv preprint arXiv:2406.02540,

    arXiv

  54. [62]

    Mixdq: Memory-efficient few-step text-to-image dif- fusion models with metric-decoupled mixed precision quan- tization

    Tianchen Zhao, Xuefei Ning, Tongcheng Fang, Enshu Liu, Guyue Huang, Zinan Lin, Shengen Yan, Guohao Dai, and Yu Wang. Mixdq: Memory-efficient few-step text-to-image dif- fusion models with metric-decoupled mixed precision quan- tization. arXiv preprint arXiv:2405.17873, 2024. 3

    2024 arXiv

  55. [63]

    Atom: Low-bit quantization for efficient and accurate llm serving

    Yilong Zhao, Chien-Yu Lin, Kan Zhu, Zihao Ye, Lequn Chen, Size Zheng, Luis Ceze, Arvind Krishnamurthy, Tianqi Chen, and Baris Kasikci. Atom: Low-bit quantization for efficient and accurate llm serving. Proceedings of Machine Learning and Systems, 6:196–209, 2024. 3 8

    2024

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.