Related papers: GenDistiller: Distilling Pre-trained Language Mode…

Recent developments in natural language representations have been accompanied by large and expensive models that leverage vast amounts of general-domain text through self-supervised pre-training. Due to the cost of applying such models to…

In this work we explore deep generative models of text in which the latent representation of a document is itself drawn from a discrete language model distribution. We formulate a variational auto-encoder for inference in this model and…

Dataset distillation aims to distill the knowledge of a large-scale real dataset into small yet informative synthetic data such that a model trained on it performs as well as a model trained on the full dataset. Despite recent progress,…

Motivated by the success of masked language modeling~(MLM) in pre-training natural language processing models, we propose w2v-BERT that explores MLM for self-supervised speech representation learning. w2v-BERT is a framework that combines…

The pre-trained text-image discriminative models, such as CLIP, has been explored for open-vocabulary semantic segmentation with unsatisfactory results due to the loss of crucial localization information and awareness of object shapes.…

Converting a pretrained Transformer into a more efficient hybrid model through distillation offers a promising approach to reducing inference costs. However, achieving high-quality generation in distilled models requires careful joint…

As the size of pre-trained speech recognition models increases, running these large models in low-latency or resource-constrained environments becomes challenging. In this work, we leverage pseudo-labelling to assemble a large-scale…

Running large-scale pre-trained language models in computationally constrained environments remains a challenging problem yet to be addressed, while transfer learning from these models has become prevalent in Natural Language Processing…

This paper studies the potential of distilling knowledge from pre-trained models, especially Masked Autoencoders. Our approach is simple: in addition to optimizing the pixel reconstruction loss on masked inputs, we minimize the distance…

Small CNN-based models usually require transferring knowledge from a large model before they are deployed in computationally resource-limited edge devices. Masked image modeling (MIM) methods achieve great success in various visual tasks…

Knowledge distillation is an effective approach to transferring knowledge from a teacher neural network to a student target network for satisfying the low-memory and fast running requirements in practice use. Whilst being able to create…

Large-scale self-supervised pre-trained speech encoders outperform conventional approaches in speech recognition and translation tasks. Due to the high cost of developing these large models, building new encoders for new tasks and deploying…

Knowledge distillation (KD), a technique widely employed in computer vision, has emerged as a de facto standard for improving the performance of small neural networks. However, prevailing KD-based approaches in video tasks primarily focus…

Recent years have witnessed great strides in self-supervised learning (SSL) on the speech processing. The SSL model is normally pre-trained on a great variety of unlabelled data and a large model size is preferred to increase the modeling…

The push to compress and impart the proficiency of Large Language Models (LLMs) into more deployable and efficient Small Language Models (SLMs) has benefited from improvements in knowledge distillation (KD) techniques. These techniques…

One of the most efficient methods for model compression is hint distillation, where the student model is injected with information (hints) from several different layers of the teacher model. Although the selection of hint points can…

Knowledge distillation addresses the problem of transferring knowledge from a teacher model to a student model. In this process, we typically have multiple types of knowledge extracted from the teacher model. The problem is to make full use…

Diffusion language models intrinsically fail to capture correlations between decoded tokens, which leads to a harsh trade-off between sampling quality and throughput. To solve this issue, we propose DiLaDiff, a variant of masked diffusion…

Neural networks are prone to catastrophic forgetting when trained incrementally on different tasks. Popular incremental learning methods mitigate such forgetting by retaining a subset of previously seen samples and replaying them during the…

Speech super-resolution (SR) is the task that restores high-resolution speech from low-resolution input. Existing models employ simulated data and constrained experimental settings, which limit generalization to real-world SR. Predictive…