相关论文: ZipIt! Merging Models from Different Tasks without…

Model merging seeks to integrate task-specific expert models into a unified architecture while preserving multi-task generalization capabilities, yet parameter interference between constituent models frequently induces performance…

Generalist robot policies, trained on large and diverse datasets, have demonstrated the ability to generalize across a wide spectrum of behaviors, enabling a single policy to act in varied real-world environments. However, they still fall…

Model merging aims to integrate task-specific abilities from individually fine-tuned models into a single model without extra training. In recent model merging methods, task vector has become a fundamental building block, as it can…

Model merging aims to integrate multiple task-specific models into a unified model that inherits the capabilities of the task-specific models, without additional training. Existing model merging methods often lack consideration of the…

Recent advances in deep learning have led to a surge of open-source models across diverse domains. While model merging offers a promising way to combine their strengths, existing approaches often suffer from parameter conflicts that degrade…

Model merging has emerged as a lightweight alternative to joint multi-task learning (MTL), yet the generalization properties of merged models remain largely unexplored. Establishing such theoretical guarantees is non-trivial, as the merging…

In a setting where segmentation models have to be built for multiple datasets, each with its own corresponding label set, a straightforward way is to learn one model for every dataset and its labels. Alternatively, multi-task architectures…

Recent work on permutation-based model merging has shown impressive low- or zero-barrier mode connectivity between models from completely different initializations. However, this line of work has not yet extended to the Transformer…

Current network training paradigms primarily focus on either centralized or decentralized data regimes. However, in practice, data availability often exhibits a hybrid nature, where both regimes coexist. This hybrid setting presents new…

At Pinterest, we utilize image embeddings throughout our search and recommendation systems to help our users navigate through visual content by powering experiences like browsing of related content and searching for exact products for…

Large language models, such as ChatGPT, Claude, or LLaMA, are gigantic, monolithic, and possess the superpower to simultaneously support thousands of tasks. However, high-throughput applications often prefer smaller task-specific models…

Multi-task learning (MTL) is a subfield of machine learning in which multiple tasks are simultaneously learned by a shared model. Such approaches offer advantages like improved data efficiency, reduced overfitting through shared…

Large deep networks have demonstrated competitive performance in single image super-resolution (SISR), with a huge volume of data involved. However, in real-world scenarios, due to the limited accessible training pairs, large models exhibit…

Model merging, typically on Instruct and Thinking models, has shown remarkable performance for efficient reasoning. In this paper, we systematically revisit the simplest merging method that interpolates two weights directly. Particularly,…

Deep learning models for vision tasks are trained on large datasets under the assumption that there exists a universal representation that can be used to make predictions for all samples. Whereas high complexity models are proven to be…

We introduce a new method for training deep Boltzmann machines jointly. Prior methods of training DBMs require an initial learning pass that trains the model greedily, one layer at a time, or do not perform well on classification tasks. In…

Deep networks have gained immense popularity in Computer Vision and other fields in the past few years due to their remarkable performance on recognition/classification tasks surpassing the state-of-the art. One of the keys to their success…

The development of language models have moved from encoder-decoder to decoder-only designs. In addition, we observe that the two most popular multimodal tasks, the generative and contrastive tasks, are nontrivial to accommodate in one…

In this paper, we present a novel data-free method for merging neural networks in weight space. Differently from most existing works, our method optimizes for the permutations of network neurons globally across all layers. This allows us to…

Deep Neural Network (DNN) models are usually trained sequentially from one layer to another, which causes forward, backward and update locking's problems, leading to poor performance in terms of training time. The existing parallel…