Non-autoregressive Sequence-to-Sequence Vision-Language Models
- URL: http://arxiv.org/abs/2403.02249v1
- Date: Mon, 4 Mar 2024 17:34:59 GMT
- Title: Non-autoregressive Sequence-to-Sequence Vision-Language Models
- Authors: Kunyu Shi, Qi Dong, Luis Goncalves, Zhuowen Tu, Stefano Soatto
- Abstract summary: We propose a parallel decoding sequence-to-sequence vision-language model that marginalizes over multiple inference paths in the decoder.
The model achieves performance on-par with its state-of-the-art autoregressive counterpart, but is faster at inference time.
- Score: 63.77614880533488
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: Sequence-to-sequence vision-language models are showing promise, but their
applicability is limited by their inference latency due to their autoregressive
way of generating predictions. We propose a parallel decoding
sequence-to-sequence vision-language model, trained with a Query-CTC loss, that
marginalizes over multiple inference paths in the decoder. This allows us to
model the joint distribution of tokens, rather than restricting to conditional
distribution as in an autoregressive model. The resulting model, NARVL,
achieves performance on-par with its state-of-the-art autoregressive
counterpart, but is faster at inference time, reducing from the linear
complexity associated with the sequential generation of tokens to a paradigm of
constant time joint inference.
Related papers
- COrAL: Order-Agnostic Language Modeling for Efficient Iterative Refinement [80.18490952057125]
Iterative refinement has emerged as an effective paradigm for enhancing the capabilities of large language models (LLMs) on complex tasks.
We propose Context-Wise Order-Agnostic Language Modeling (COrAL) to overcome these challenges.
Our approach models multiple token dependencies within manageable context windows, enabling the model to perform iterative refinement internally.
arXiv Detail & Related papers (2024-10-12T23:56:19Z) - Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusion [61.03681839276652]
Diffusion Forcing is a new training paradigm where a diffusion model is trained to denoise a set of tokens with independent per-token noise levels.
We apply Diffusion Forcing to sequence generative modeling by training a causal next-token prediction model to generate one or several future tokens.
arXiv Detail & Related papers (2024-07-01T15:43:25Z) - Fast and Robust Early-Exiting Framework for Autoregressive Language
Models with Synchronized Parallel Decoding [43.659680579686544]
We propose a Fast and Robust Early-Exiting framework, which incorporates a shallow-deep module and a synchronized parallel decoding.
Our framework enables faster inference by synchronizing the decoding process of the current token with previously stacked early-exited tokens.
As parallel decoding allows us to observe predictions from both shallow and deep models, we present a novel adaptive threshold estimator.
arXiv Detail & Related papers (2023-10-09T05:53:05Z) - RecycleGPT: An Autoregressive Language Model with Recyclable Module [13.243551482623623]
We present RecycleGPT, a generative language model with fast decoding speed.
Our approach relies on the observation that adjacent tokens in a sequence usually have strong correlations.
Experiments and analysis demonstrate the effectiveness of our approach in lowering inference latency, achieving up to 1.4x speedup.
arXiv Detail & Related papers (2023-08-07T09:14:33Z) - SequenceMatch: Imitation Learning for Autoregressive Sequence Modelling with Backtracking [60.109453252858806]
A maximum-likelihood (MLE) objective does not match a downstream use-case of autoregressively generating high-quality sequences.
We formulate sequence generation as an imitation learning (IL) problem.
This allows us to minimize a variety of divergences between the distribution of sequences generated by an autoregressive model and sequences from a dataset.
Our resulting method, SequenceMatch, can be implemented without adversarial training or architectural changes.
arXiv Detail & Related papers (2023-06-08T17:59:58Z) - Conditional Denoising Diffusion for Sequential Recommendation [62.127862728308045]
Two prominent generative models, Generative Adversarial Networks (GANs) and Variational AutoEncoders (VAEs)
GANs suffer from unstable optimization, while VAEs are prone to posterior collapse and over-smoothed generations.
We present a conditional denoising diffusion model, which includes a sequence encoder, a cross-attentive denoising decoder, and a step-wise diffuser.
arXiv Detail & Related papers (2023-04-22T15:32:59Z) - Symbolic Music Generation with Diffusion Models [4.817429789586127]
We present a technique for training diffusion models on sequential data by parameterizing the discrete domain in the continuous latent space of a pre-trained variational autoencoder.
We show strong unconditional generation and post-hoc conditional infilling results compared to autoregressive language models operating over the same continuous embeddings.
arXiv Detail & Related papers (2021-03-30T05:48:05Z) - Synergetic Learning of Heterogeneous Temporal Sequences for
Multi-Horizon Probabilistic Forecasting [48.8617204809538]
We propose Variational Synergetic Multi-Horizon Network (VSMHN), a novel deep conditional generative model.
To learn complex correlations across heterogeneous sequences, a tailored encoder is devised to combine the advances in deep point processes models and variational recurrent neural networks.
Our model can be trained effectively using variational inference and generates predictions with Monte-Carlo simulation.
arXiv Detail & Related papers (2021-01-31T11:00:55Z) - Fast Sequence Generation with Multi-Agent Reinforcement Learning [40.75211414663022]
Non-autoregressive decoding has been proposed in machine translation to speed up the inference time by generating all words in parallel.
We propose a simple and efficient model for Non-Autoregressive sequence Generation (NAG) with a novel training paradigm: Counterfactuals-critical Multi-Agent Learning (CMAL)
On MSCOCO image captioning benchmark, our NAG method achieves a performance comparable to state-of-the-art autoregressive models, while brings 13.9x decoding speedup.
arXiv Detail & Related papers (2021-01-24T12:16:45Z)
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.