Revisiting Funnel Transformers for Modern LLM Architectures with Comprehensive Ablations in Training and Inference Configurations

• URL: http://arxiv.org/abs/2504.02877v1
• Date: Wed, 02 Apr 2025 02:09:17 GMT
• Title: Revisiting Funnel Transformers for Modern LLM Architectures with Comprehensive Ablations in Training and Inference Configurations
• Authors: DongHyun Choi, Lucas Spangher, Chris Hidey, Peter Grabowski, Ramy Eskander,
• Abstract summary: We investigate the impact of funneling in contemporary Gemma2 Transformer architectures.<n>We find that funneling creates information bottlenecks that propagate through deeper network layers, particularly in larger models.<n>Our findings highlight key trade-offs between computational efficiency and model accuracy.
• Score: 1.6255281211429766
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Transformer-based Large Language Models, which suffer from high computational costs, advance so quickly that techniques proposed to streamline earlier iterations are not guaranteed to benefit more modern models. Building upon the Funnel Transformer proposed by Dai and Le (2020), which progressively compresses intermediate representations, we investigate the impact of funneling in contemporary Gemma2 Transformer architectures. We systematically evaluate various funnel configurations and recovery methods, comparing: (1) standard pretraining to funnel-aware pretraining strategies, (2) the impact of funnel-aware fine-tuning, and (3) the type of sequence recovery operation. Our results demonstrate that funneling creates information bottlenecks that propagate through deeper network layers, particularly in larger models (e.g., Gemma 7B), leading to at times unmanageable performance lost. However, carefully selecting the funneling layer and employing effective recovery strategies, can substantially mitigate performance losses, achieving up to a 44\% reduction in latency. Our findings highlight key trade-offs between computational efficiency and model accuracy, providing practical guidance for deploying funnel-based approaches in large-scale natural language applications.

Related papers

• BEExformer: A Fast Inferencing Transformer Architecture via Binarization with Multiple Early Exits [2.7651063843287718]
  Large Language Models (LLMs) based on transformers achieve cutting-edge results on a variety of applications.<n>Among various efficiency considerations, model binarization and Early Exit (EE) are common effective solutions.<n>We propose Binarized Early Exit Transformer (BEExformer), the first-ever selective learning transformer architecture.
  arXiv  Detail & Related papers  (2024-12-06T17:58:14Z)
• Efficient Diffusion as Low Light Enhancer [63.789138528062225]
  Reflectance-Aware Trajectory Refinement (RATR) is a simple yet effective module to refine the teacher trajectory using the reflectance component of images. textbfReflectance-aware textbfDiffusion with textbfDistilled textbfTrajectory (textbfReDDiT) is an efficient and flexible distillation framework tailored for Low-Light Image Enhancement (LLIE)
  arXiv  Detail & Related papers  (2024-10-16T08:07:18Z)
• Investigating Low-Rank Training in Transformer Language Models: Efficiency and Scaling Analysis [16.253898272659242]
  This study focuses on Transformer-based LLMs, specifically applying low-rank parametrization to feedforward networks (FFNs) Experiments on the large RefinedWeb dataset show that low-rank parametrization is both efficient (e.g., 2.6$times$ FFN speed-up with 32% parameters) and effective during training. Motivated by this finding, we develop the wide and structured networks surpassing the current medium-sized and large-sized Transformer in perplexity and throughput performance.
  arXiv  Detail & Related papers  (2024-07-13T10:08:55Z)
• SHERL: Synthesizing High Accuracy and Efficient Memory for Resource-Limited Transfer Learning [63.93193829913252]
  We propose an innovative METL strategy called SHERL for resource-limited scenarios. In the early route, intermediate outputs are consolidated via an anti-redundancy operation. In the late route, utilizing minimal late pre-trained layers could alleviate the peak demand on memory overhead.
  arXiv  Detail & Related papers  (2024-07-10T10:22:35Z)
• DiJiang: Efficient Large Language Models through Compact Kernelization [30.24187657746638]
  We present a novel Frequency Domain Kernelization approach that enables the transformation of a pre-trained vanilla Transformer into a linear complexity model with little training costs. Experiments demonstrate that the proposed method achieves comparable performance to the original Transformer, but with significantly reduced training costs and much faster inference speeds.
  arXiv  Detail & Related papers  (2024-03-29T02:32:15Z)
• Accelerating Deep Neural Networks via Semi-Structured Activation Sparsity [0.0]
  Exploiting sparsity in the network's feature maps is one of the ways to reduce its inference latency. We propose a solution to induce semi-structured activation sparsity exploitable through minor runtime modifications. Our approach yields a speed improvement of $1.25 times$ with a minimal accuracy drop of $1.1%$ for the ResNet18 model on the ImageNet dataset.
  arXiv  Detail & Related papers  (2023-09-12T22:28:53Z)
• End-to-End Meta-Bayesian Optimisation with Transformer Neural Processes [52.818579746354665]
  This paper proposes the first end-to-end differentiable meta-BO framework that generalises neural processes to learn acquisition functions via transformer architectures. We enable this end-to-end framework with reinforcement learning (RL) to tackle the lack of labelled acquisition data.
  arXiv  Detail & Related papers  (2023-05-25T10:58:46Z)
• Full Stack Optimization of Transformer Inference: a Survey [58.55475772110702]
  Transformer models achieve superior accuracy across a wide range of applications. The amount of compute and bandwidth required for inference of recent Transformer models is growing at a significant rate. There has been an increased focus on making Transformer models more efficient.
  arXiv  Detail & Related papers  (2023-02-27T18:18:13Z)
• Online Convolutional Re-parameterization [51.97831675242173]
  We present online convolutional re- parameterization (OREPA), a two-stage pipeline, aiming to reduce the huge training overhead by squeezing the complex training-time block into a single convolution. Compared with the state-of-the-art re-param models, OREPA is able to save the training-time memory cost by about 70% and accelerate the training speed by around 2x. We also conduct experiments on object detection and semantic segmentation and show consistent improvements on the downstream tasks.
  arXiv  Detail & Related papers  (2022-04-02T09:50:19Z)
• Latency Adjustable Transformer Encoder for Language Understanding [0.8287206589886879]
  This paper proposes an efficient Transformer architecture that adjusts the inference computational cost adaptively with a desired inference latency speedup. The proposed method detects less important hidden sequence elements (word-vectors) and eliminates them in each encoder layer using a proposed Attention Context Contribution (ACC) metric. The proposed method mathematically and experimentally improves the inference latency of BERT_base and GPT-2 by up to 4.8 and 3.72 times with less than 0.75% accuracy drop and passable perplexity on average.
  arXiv  Detail & Related papers  (2022-01-10T13:04:39Z)
• Extrapolation for Large-batch Training in Deep Learning [72.61259487233214]
  We show that a host of variations can be covered in a unified framework that we propose. We prove the convergence of this novel scheme and rigorously evaluate its empirical performance on ResNet, LSTM, and Transformer.
  arXiv  Detail & Related papers  (2020-06-10T08:22:41Z)

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.