CoChat

The DOI matches its Crossref record; 65 of the 68 references it cites were found in a registry or an index.

DeepNet: Scaling Transformers to 1,000 Layers

Open access
Wang, Hongyu, Ma, Shuming, Dong, Li, Huang, Shaohan, Zhang, Dongdong, Wei, Furu · 2024 · IEEE Transactions on Pattern Analysis and Machine Intelligence

Abstract

In this paper, we propose a simple yet effective method to stabilize extremely deep Transformers. Specifically, we introduce a new normalization function (DeepNorm) to modify the residual connection in Transformer, accompanying with theoretically derived initialization. In-depth theoretical analysis shows that model updates can be bounded in a stable way. The proposed method combines the best of two worlds, i.e., good performance of Post-LN and stable training of Pre-LN, makingDeepNorma preferred alternative. We successfully scale Transformers up to 1,000 layers (i.e., 2,500 attention and feed-forward network sublayers) without difficulty, which is one order of magnitude deeper than previous deep Transformers. Extensive experiments demonstrate thatDeepNethas superior performance across various benchmarks, including machine translation, language modeling (i.e., BERT, GPT) and vision pre-training (i.e., BEiT). Remarkably, on a multilingual benchmark with 7,482 translation directions, our 200-layer model with 3.2B parameters significantly outperforms the 48-layer state-of-the-art model with 12B parameters by 5 BLEU points, which indicates a promising scaling direction. Our code is available athttps://aka.ms/torchscale.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1109/tpami.2024.3386927

Cites

65 of 65 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 68 references

Research works printed but not linked: 2

  1. Wang, Magneto: A foundation transformer, 36077
  2. Dosovitskiy, An image is worth 16x16 words: Transformers for image recognition at scale

Cited by

See all 118

Metadata

Authors
Wang, H., Ma, S., Dong, L., Huang, S., Zhang, D., & Wei, F.
Affiliations
  • Chinese Academy of Sciences (CN)
  • Microsoft Research Asia (China) (CN)
Venue
IEEE Transactions on Pattern Analysis and Machine Intelligence
Published
2024-10
Volume
46
Issue
10
Pages
6761-6774
ISSN
  • 0162-8828 (ISSN-L)
  • 2160-9292
  • 1939-3539
OpenAlex
W4394698525
Cited by
113
References
68
Citations by year
2026: 21 · 2025: 27 · 2024: 23
FWCI
10.48
Type
journal article
Language
en
Open access
Locations
  • IEEE Transactions on Pattern Analysis and Machine Intelligence (publishedVersion, cc-by)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
  • Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
Keywords
layer normalization, initialization, training stability, model scaling, multilingual machine translation, BEiT, language modeling

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: DeepNet: Scaling Transformers to 1,000 Layers. https://app.cochat.ai/papers/10.1109/tpami.2024.3386927
The paper
Wang, H., Ma, S., Dong, L., Huang, S., Zhang, D., & Wei, F. (2024). DeepNet: Scaling Transformers to 1,000 Layers. IEEE Transactions on Pattern Analysis and Machine Intelligence. https://doi.org/10.1109/tpami.2024.3386927

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.