MatFormer: Nested Transformer for Elastic Inference
- URL: http://arxiv.org/abs/2310.07707v2
- Date: Sun, 15 Dec 2024 03:45:36 GMT
- Title: MatFormer: Nested Transformer for Elastic Inference
- Authors: Devvrit, Sneha Kudugunta, Aditya Kusupati, Tim Dettmers, Kaifeng Chen, Inderjit Dhillon, Yulia Tsvetkov, Hannaneh Hajishirzi, Sham Kakade, Ali Farhadi, Prateek Jain,
- Abstract summary: MatFormer is a novel Transformer architecture designed to provide elastic inference across diverse deployment constraints.
MatFormer achieves this by incorporating a nested Feed Forward Network (FFN) block structure within a standard Transformer model.
We show that a 850M decoder-only MatFormer language model (MatLM) allows us to extract multiple smaller models spanning from 582M to 850M parameters.
- Score: 91.45687988953435
- License:
- Abstract: Foundation models are applied in a broad spectrum of settings with different inference constraints, from massive multi-accelerator clusters to resource-constrained standalone mobile devices. However, the substantial costs associated with training these models often limit the number of unique model sizes that can be offered. Consequently, practitioners are compelled to select a model that may not be optimally aligned with their specific latency and cost requirements. We present MatFormer, a novel Transformer architecture designed to provide elastic inference across diverse deployment constraints. MatFormer achieves this by incorporating a nested Feed Forward Network (FFN) block structure within a standard Transformer model. During training, we optimize the parameters of multiple nested FFN blocks with varying sizes, enabling the extraction of hundreds of accurate smaller models without incurring additional computational costs. We empirically validate the efficacy of MatFormer across different model classes (decoders and encoders) and modalities (language and vision), demonstrating its potential for real-world deployment. We show that a 850M decoder-only MatFormer language model (MatLM) allows us to extract multiple smaller models spanning from 582M to 850M parameters, each exhibiting better validation loss and one-shot downstream evaluations than independently trained counterparts. Furthermore, we observe that smaller encoders extracted from a universal MatFormer-based ViT (MatViT) encoder preserve the metric-space structure for adaptive large-scale retrieval. Finally, we showcase that speculative decoding with the accurate and consistent submodels extracted from MatFormer can lead to significant reduction in inference latency. Project website: https://devvrit.github.io/matformer/
Related papers
- TokenFormer: Rethinking Transformer Scaling with Tokenized Model Parameters [102.1116808722299]
We introduce TokenFormer, a scalable architecture for scaling Transformers.
By treating model parameters as tokens, we replace all the linear projections in Transformers.
Our model scales from 124M to 1.4B parameters by incrementally adding new key-value parameter pairs.
arXiv Detail & Related papers (2024-10-30T16:19:00Z) - MatMamba: A Matryoshka State Space Model [24.85566171753877]
MatMamba is a state space model which combines Matryoshka-style learning with Mamba2.
MatMamba allows for efficient and adaptive deployment across various model sizes.
We train language and image models at a variety of parameter sizes from 35M to 1.4B.
arXiv Detail & Related papers (2024-10-09T09:41:34Z) - XMoE: Sparse Models with Fine-grained and Adaptive Expert Selection [30.687511115573038]
tool is a novel MoE designed to enhance both the efficacy and efficiency of sparse MoE models.
tool can enhance model performance while decreasing the computation load at MoE layers by over 50% without sacrificing performance.
arXiv Detail & Related papers (2024-02-27T08:18:02Z) - Quantized Transformer Language Model Implementations on Edge Devices [1.2979415757860164]
Large-scale transformer-based models like the Bidirectional Representations from Transformers (BERT) are widely used for Natural Language Processing (NLP) applications.
These models are initially pre-trained with a large corpus with millions of parameters and then fine-tuned for a downstream NLP task.
One of the major limitations of these large-scale models is that they cannot be deployed on resource-constrained devices due to their large model size and increased inference latency.
arXiv Detail & Related papers (2023-10-06T01:59:19Z) - Fourier Transformer: Fast Long Range Modeling by Removing Sequence
Redundancy with FFT Operator [24.690247474891958]
Fourier Transformer is able to significantly reduce computational costs while retain the ability to inherit from various large pretrained models.
Our model achieves state-of-the-art performances among all transformer-based models on the long-range modeling benchmark LRA.
For generative seq-to-seq tasks including CNN/DailyMail and ELI5, by inheriting the BART weights our model outperforms the standard BART.
arXiv Detail & Related papers (2023-05-24T12:33:06Z) - Bilaterally Slimmable Transformer for Elastic and Efficient Visual
Question Answering [75.86788916930377]
bilaterally slimmable Transformer (BST) integrated into arbitrary Transformer-based VQA models.
One slimmed MCAN-BST submodel achieves comparable accuracy on VQA-v2.
Smallest MCAN-BST submodel has 9M parameters and 0.16G FLOPs during inference.
arXiv Detail & Related papers (2022-03-24T02:26:04Z) - MoEfication: Conditional Computation of Transformer Models for Efficient
Inference [66.56994436947441]
Transformer-based pre-trained language models can achieve superior performance on most NLP tasks due to large parameter capacity, but also lead to huge computation cost.
We explore to accelerate large-model inference by conditional computation based on the sparse activation phenomenon.
We propose to transform a large model into its mixture-of-experts (MoE) version with equal model size, namely MoEfication.
arXiv Detail & Related papers (2021-10-05T02:14:38Z) - Ensemble Distillation for Robust Model Fusion in Federated Learning [72.61259487233214]
Federated Learning (FL) is a machine learning setting where many devices collaboratively train a machine learning model.
In most of the current training schemes the central model is refined by averaging the parameters of the server model and the updated parameters from the client side.
We propose ensemble distillation for model fusion, i.e. training the central classifier through unlabeled data on the outputs of the models from the clients.
arXiv Detail & Related papers (2020-06-12T14:49:47Z) - Model Fusion via Optimal Transport [64.13185244219353]
We present a layer-wise model fusion algorithm for neural networks.
We show that this can successfully yield "one-shot" knowledge transfer between neural networks trained on heterogeneous non-i.i.d. data.
arXiv Detail & Related papers (2019-10-12T22:07:15Z)
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.