MoE-Mamba: Efficient Selective State Space Models with Mixture of
Experts
- URL: http://arxiv.org/abs/2401.04081v2
- Date: Mon, 26 Feb 2024 17:04:41 GMT
- Title: MoE-Mamba: Efficient Selective State Space Models with Mixture of
Experts
- Authors: Maciej Pi\'oro, Kamil Ciebiera, Krystian Kr\'ol, Jan Ludziejewski,
Micha{\l} Krutul, Jakub Krajewski, Szymon Antoniak, Piotr Mi{\l}o\'s, Marek
Cygan, Sebastian Jaszczur
- Abstract summary: State Space Models (SSMs) have become serious contenders in the field of sequential modeling.
MoE has significantly improved Transformer-based Large Language Models, including recent state-of-the-art open models.
We propose that to unlock the potential of SSMs for scaling, they should be combined with MoE.
- Score: 4.293771840782942
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: State Space Models (SSMs) have become serious contenders in the field of
sequential modeling, challenging the dominance of Transformers. At the same
time, Mixture of Experts (MoE) has significantly improved Transformer-based
Large Language Models, including recent state-of-the-art open models. We
propose that to unlock the potential of SSMs for scaling, they should be
combined with MoE. We showcase this on Mamba, a recent SSM-based model that
achieves remarkable performance. Our model, MoE-Mamba, outperforms both Mamba
and baseline Transformer-MoE. In particular, MoE-Mamba reaches the same
performance as Mamba in $2.35\times$ fewer training steps while preserving the
inference performance gains of Mamba against Transformer.
Related papers
- MambaVision: A Hybrid Mamba-Transformer Vision Backbone [54.965143338206644]
We propose a novel hybrid Mamba-Transformer backbone, denoted as MambaVision, which is specifically tailored for vision applications.
Our core contribution includes redesigning the Mamba formulation to enhance its capability for efficient modeling of visual features.
We conduct a comprehensive ablation study on the feasibility of integrating Vision Transformers (ViT) with Mamba.
arXiv Detail & Related papers (2024-07-10T23:02:45Z) - Venturing into Uncharted Waters: The Navigation Compass from Transformer to Mamba [77.21394300708172]
Transformer, a deep neural network architecture, has long dominated the field of natural language processing and beyond.
The recent introduction of Mamba challenges its supremacy, sparks considerable interest among researchers, and gives rise to a series of Mamba-based models that have exhibited notable potential.
This survey paper orchestrates a comprehensive discussion, diving into essential research dimensions, covering: (i) the functioning of the Mamba mechanism and its foundation on the principles of structured state space models; (ii) the proposed improvements and the integration of Mamba with various networks, exploring its potential as a substitute for Transformers; (iii) the combination of
arXiv Detail & Related papers (2024-06-24T15:27:21Z) - An Empirical Study of Mamba-based Language Models [69.74383762508805]
Selective state-space models (SSMs) like Mamba overcome some shortcomings of Transformers.
We present a direct comparison between 8B-context Mamba, Mamba-2, and Transformer models trained on the same datasets.
We find that the 8B Mamba-2-Hybrid exceeds the 8B Transformer on all 12 standard tasks.
arXiv Detail & Related papers (2024-06-12T05:25:15Z) - Mamba State-Space Models Can Be Strong Downstream Learners [1.6385815610837167]
Mamba state-space models (SSMs) have recently outperformed state-of-the-art (SLLMs) in various tasks.
Mixed-precision (MPFT) and fine-tuning (PEFT) are under-evaluated.
We show that combining MPFT and PEFT enables up to 215 times more tokens-per-second and 65.5% reduced per-memory.
arXiv Detail & Related papers (2024-05-31T21:46:23Z) - Demystify Mamba in Vision: A Linear Attention Perspective [72.93213667713493]
Mamba is an effective state space model with linear computation complexity.
We show that Mamba shares surprising similarities with linear attention Transformer.
We propose a Mamba-Like Linear Attention (MLLA) model by incorporating the merits of these two key designs into linear attention.
arXiv Detail & Related papers (2024-05-26T15:31:09Z) - MambaOut: Do We Really Need Mamba for Vision? [70.60495392198686]
Mamba, an architecture with RNN-like token mixer of state space model (SSM), was recently introduced to address the quadratic complexity of the attention mechanism.
This paper conceptually concludes that Mamba is ideally suited for tasks with long-sequence and autoregressive characteristics.
We construct a series of models named MambaOut through stacking Mamba blocks while removing their core token mixer, SSM.
arXiv Detail & Related papers (2024-05-13T17:59:56Z) - CLIP-Mamba: CLIP Pretrained Mamba Models with OOD and Hessian Evaluation [18.383760896304604]
This report introduces the first attempt to train a Mamba model utilizing contrastive technical-image pretraining (CLIP)
A Mamba model 67 million parameters is on par with a 307 million- parameters Vision Transformer (ViT) model in zero-shot classification tasks.
arXiv Detail & Related papers (2024-04-30T09:40:07Z) - The Hidden Attention of Mamba Models [54.50526986788175]
The Mamba layer offers an efficient selective state space model (SSM) that is highly effective in modeling multiple domains.
We show that such models can be viewed as attention-driven models.
This new perspective enables us to empirically and theoretically compare the underlying mechanisms to that of the self-attention layers in transformers.
arXiv Detail & Related papers (2024-03-03T18:58:21Z) - BlackMamba: Mixture of Experts for State-Space Models [10.209192169793772]
State-space models (SSMs) have recently demonstrated competitive performance to transformers at large-scale language modeling benchmarks.
MoE models have shown remarkable performance while significantly reducing the compute and latency costs of inference.
We present BlackMamba, a novel architecture that combines the Mamba SSM with MoE to obtain the benefits of both.
arXiv Detail & Related papers (2024-02-01T07:15:58Z)
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.