Brainformers: Trading Simplicity for Efficiency

• URL: http://arxiv.org/abs/2306.00008v2
• Date: Thu, 25 Apr 2024 05:46:01 GMT
• Title: Brainformers: Trading Simplicity for Efficiency
• Authors: Yanqi Zhou, Nan Du, Yanping Huang, Daiyi Peng, Chang Lan, Da Huang, Siamak Shakeri, David So, Andrew Dai, Yifeng Lu, Zhifeng Chen, Quoc Le, Claire Cui, James Laudon, Jeff Dean,
• Abstract summary: We develop a complex block, named Brainformer, that consists of a diverse sets of layers. Brainformer consistently outperforms the state-of-the-art dense and sparse Transformers. A Brainformer model with 8 billion activated parameters per token demonstrates 2x faster training convergence and 5x faster step time.
• Score: 39.53511089374572
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Transformers are central to recent successes in natural language processing and computer vision. Transformers have a mostly uniform backbone where layers alternate between feed-forward and self-attention in order to build a deep network. Here we investigate this design choice and find that more complex blocks that have different permutations of layer primitives can be more efficient. Using this insight, we develop a complex block, named Brainformer, that consists of a diverse sets of layers such as sparsely gated feed-forward layers, dense feed-forward layers, attention layers, and various forms of layer normalization and activation functions. Brainformer consistently outperforms the state-of-the-art dense and sparse Transformers, in terms of both quality and efficiency. A Brainformer model with 8 billion activated parameters per token demonstrates 2x faster training convergence and 5x faster step time compared to its GLaM counterpart. In downstream task evaluation, Brainformer also demonstrates a 3% higher SuperGLUE score with fine-tuning compared to GLaM with a similar number of activated parameters. Finally, Brainformer largely outperforms a Primer dense model derived with NAS with similar computation per token on fewshot evaluations.

Related papers

• UniMoD: Efficient Unified Multimodal Transformers with Mixture-of-Depths [17.68867710994329]
  UniMoD is a task-aware token pruning method that employs a separate router for each task to determine which tokens should be pruned. We apply our method to Show-o and Emu3, reducing training FLOPs by approximately 15% in Show-o and 40% in Emu3, while maintaining or improving performance on several benchmarks.
  arXiv  Detail & Related papers  (2025-02-10T13:52:52Z)
• Towards 3D Acceleration for low-power Mixture-of-Experts and Multi-Head Attention Spiking Transformers [5.1210823165448]
  Spiking Neural Networks(SNNs) provide a brain-inspired and event-driven mechanism that is believed to be critical to unlock energy-efficient deep learning. This paper introduces the first 3D hardware architecture and design methodology for Mixture-of-Experts and Multi-Head Attention spiking transformers.
  arXiv  Detail & Related papers  (2024-12-07T05:15:05Z)
• Mixture-of-Modules: Reinventing Transformers as Dynamic Assemblies of Modules [96.21649779507831]
  We propose a novel architecture dubbed mixture-of-modules (MoM) MoM is motivated by an intuition that any layer, regardless of its position, can be used to compute a token. We show that MoM provides not only a unified framework for Transformers but also a flexible and learnable approach for reducing redundancy.
  arXiv  Detail & Related papers  (2024-07-09T08:50:18Z)
• Hyper-Transformer for Amodal Completion [82.4118011026855]
  Amodal object completion is a complex task that involves predicting the invisible parts of an object based on visible segments and background information. We introduce a novel framework named the Hyper-Transformer Amodal Network (H-TAN) This framework utilizes a hyper transformer equipped with a dynamic convolution head to directly learn shape priors and accurately predict amodal masks.
  arXiv  Detail & Related papers  (2024-05-30T11:11:54Z)
• MoEUT: Mixture-of-Experts Universal Transformers [75.96744719516813]
  Universal Transformers (UTs) have advantages over standard Transformers in learning compositional generalizations. Layer-sharing drastically reduces the parameter count compared to the non-shared model with the same dimensionality. No previous work has succeeded in proposing a shared-layer Transformer design that is competitive in parameter count-dominated tasks such as language modeling.
  arXiv  Detail & Related papers  (2024-05-25T03:24:32Z)
• Sliceformer: Make Multi-head Attention as Simple as Sorting in Discriminative Tasks [32.33355192614434]
  We propose an effective and efficient surrogate of the Transformer, called Sliceformer. Our Sliceformer replaces the classic MHA mechanism with an extremely simple slicing-sorting'' operation. Our Sliceformer achieves comparable or better performance with lower memory cost and faster speed than the Transformer and its variants.
  arXiv  Detail & Related papers  (2023-10-26T14:43:07Z)
• Wide Attention Is The Way Forward For Transformers [9.252523881586054]
  We show that wide single layer Transformer models can compete with or outperform deeper ones in a variety of Natural Language Processing (NLP) tasks. Our results suggest that the critical direction for building better Transformers for NLP is their width, and that their depth is less relevant.
  arXiv  Detail & Related papers  (2022-10-02T21:49:54Z)
• nnFormer: Interleaved Transformer for Volumetric Segmentation [50.10441845967601]
  We introduce nnFormer, a powerful segmentation model with an interleaved architecture based on empirical combination of self-attention and convolution. nnFormer achieves tremendous improvements over previous transformer-based methods on two commonly used datasets Synapse and ACDC.
  arXiv  Detail & Related papers  (2021-09-07T17:08:24Z)
• GroupBERT: Enhanced Transformer Architecture with Efficient Grouped Structures [57.46093180685175]
  We demonstrate a set of modifications to the structure of a Transformer layer, producing a more efficient architecture. We add a convolutional module to complement the self-attention module, decoupling the learning of local and global interactions. We apply the resulting architecture to language representation learning and demonstrate its superior performance compared to BERT models of different scales.
  arXiv  Detail & Related papers  (2021-06-10T15:41:53Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.