The DOI matches its Crossref record; 9 of the 9 references it cites were found in a registry or an index.
Understanding the Difficulty of Training Transformers
Open accessAbstract
Transformers have proved effective in many NLP tasks.However, their training requires non-trivial efforts regarding carefully designing cutting-edge optimizers and learning rate schedulers (e.g., conventional SGD fails to train Transformers effectively).Our objective here is to understand what complicates Transformer training from both empirical and theoretical perspectives.Our analysis reveals that unbalanced gradients are not the root cause of the instability of training.Instead, we identify an amplification effect that influences training substantially-for each layer in a multi-layer Transformer model, heavy dependency on its residual branch makes training unstable, since it amplifies small parameter perturbations (e.g., parameter updates) and results in significant disturbances in the model output.Yet we observe that a light dependency limits the model potential and leads to inferior trained models.Inspired by our analysis, we propose Admin (Adaptive model initialization) to stabilize the early stage's training and unleash its full potential in the late stage.Extensive experiments show that Admin is more stable, converges faster, and leads to better performance 1 .W (V 2 ) h , where f s is the row-wise softmax function and W (•) h are parameters.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
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
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1423 · Cited by 54,325
Rethinking the Inception Architecture for Computer Vision
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., & Wojna, Z. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.308 · Cited by 32,138
Veličković, P., Cucurull, G., Casanova, A., Romero, A., Liò, P., & Bengio, Y. (2017) · arXiv · doi:10.48550/arxiv.1710.10903 · Cited by 12,107
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., & Liu, P. J. (2019) · arXiv · doi:10.48550/arxiv.1910.10683 · Cited by 10,929
Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour
Goyal, P., Dollár, P., Girshick, R., Noordhuis, P., Wesolowski, L., Kyrola, A., Tulloch, A., Jia, Y., & He, K. (2017) · arXiv · doi:10.48550/arxiv.1706.02677 · Cited by 2,294
Parmar, N., Vaswani, A., Uszkoreit, J., Kaiser, Ł., Shazeer, N., Ku, A., & Tran, D. (2018) · arXiv · doi:10.48550/arxiv.1802.05751 · Cited by 1,085
Stand-Alone Self-Attention in Vision Models
Ramachandran, P., Parmar, N., Vaswani, A., Bello, I., Levskaya, A., & Shlens, J. (2019) · arXiv · doi:10.48550/arxiv.1906.05909 · Cited by 977
Cited by
Zero-shot text-to-image generation
Ramesh, A., Pavlov, M., Goh, G., Gray, S., Voss, C., Radford, A., Chen, M., & Sutskever, I. (2021) · arXiv · doi:10.48550/arxiv.2102.12092 · Cited by 2,417
Lin, T., Wang, Y., Liu, X., & Qiu, X. (2022) · AI Open · doi:10.1016/j.aiopen.2022.10.001 · Cited by 1,422
Self-citationDiffusion policy: Visuomotor policy learning via action diffusion
Chi, C., Xu, Z., Feng, S., Cousineau, E., Du, Y., Burchfiel, B., Tedrake, R., & Song, S. (2025) · The International Journal of Robotics Research · doi:10.1177/02783649241273668 · Cited by 613
“The difficulty of transformer training [25] is not unique to Diffusion Policy and could potentially be resolved in the future with improved transformer training techniques or increased data scale.”
BackgroundDiffusion Policy: Visuomotor Policy Learning via Action Diffusion
Chi, C., Feng, S., Du, Y., Xu, Z., Cousineau, E., Burchfiel, B., & Song, S. (2023) · Robotics: Science and Systems XIX · doi:10.15607/rss.2023.xix.026 · Cited by 530
ActionFormer: Localizing Moments of Actions with Transformers
Zhang, C.-L., Wu, J., & Li, Y. (2022) · Lecture Notes in Computer Science · doi:10.1007/978-3-031-19772-7_29 · Cited by 435
Metadata
- Authors
- Liu, L., Liu, X., Gao, J., Chen, W., & Han, J.
- Venue
- Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
- Published
- 2020
- Pages
- 5747-5763
- OpenAlex
- W3103334733
- Cited by
- 170
- References
- 9
- FWCI
- 11.39
- Type
- conference paper
- Language
- en
- Open access
- hybrid
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: Understanding the Difficulty of Training Transformers. https://app.cochat.ai/papers/10.18653/v1/2020.emnlp-main.463
- The paper
- Liu, L., Liu, X., Gao, J., Chen, W., & Han, J. (2020). Understanding the Difficulty of Training Transformers. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). https://doi.org/10.18653/v1/2020.emnlp-main.463
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.
