Pith. sign in

REVIEW 4 cited by

Understanding the Difficulty of Training Transformers

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2004.08249 v3 pith:PVRUE43M submitted 2020-04-17 cs.LG cs.CLstat.ML

Understanding the Difficulty of Training Transformers

classification cs.LG cs.CLstat.ML
keywords trainingmodeltextbftransformersadminanalysisdependencyleads
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
abstract

Transformers have proved effective in many NLP tasks. However, their training requires non-trivial efforts regarding designing cutting-edge optimizers and learning rate schedulers carefully (e.g., conventional SGD fails to train Transformers effectively). Our objective here is to understand $\textit{what complicates Transformer training}$ 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 ($\textbf{Ad}$aptive $\textbf{m}$odel $\textbf{in}$itialization) to stabilize 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. Implementations are released at: https://github.com/LiyuanLucasLiu/Transforemr-Clinic.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 4 Pith papers

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

  1. Revealing Modular Gradient Noise Imbalance in LLMs: Calibrating Adam via Signal-to-Noise Ratio

    cs.LG 2026-05 unverdicted novelty 6.0

    MoLS scales Adam updates using module-level SNR estimates to correct gradient noise imbalance and improve LLM training convergence and generalization.

  2. H$_2$O: Heavy-Hitter Oracle for Efficient Generative Inference of Large Language Models

    cs.LG 2023-06 unverdicted novelty 6.0

    H2O evicts non-heavy-hitter tokens from the KV cache using a dynamic submodular policy, retaining recent and frequent-co-occurrence tokens to reduce memory while preserving accuracy.

  3. Improving End-to-End Speech Recognition for Dysarthric Speech through In-Domain Data Augmentation

    eess.AS 2026-06 unverdicted novelty 4.0

    Severity-specific fine-tuning of Wav2Vec2 with SRM, PM, FM, and VTLP augmentations yields relative WER reductions of 15-30% on dysarthric speech across severity levels.

  4. Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities

    cs.CL 2025-07 unverdicted novelty 4.0

    Gemini 2.5 Pro and Flash models are presented as achieving frontier performance in reasoning, coding, and long-context multimodal tasks while spanning a cost-capability Pareto curve.