BlueLM-V-3B: Algorithm and System Co-Design for Multimodal Large Language Models on Mobile Devices

The emergence and growing popularity of multimodal large language models (MLLMs) have significant potential to enhance various aspects of daily life, from improving communication to facilitating learning and problem-solving. Mobile phones, as essential daily companions, represent the most effective and accessible deployment platform for MLLMs, enabling seamless integration into everyday tasks. However, deploying MLLMs on mobile phones presents challenges due to limitations in memory size and computational capability, making it difficult to achieve smooth and real-time processing without extensive optimization. In this paper, we present BlueLM-V-3B, an algorithm and system co-design approach specifically tailored for the efficient deployment of MLLMs on mobile platforms. To be specific, we redesign the dynamic resolution scheme adopted by mainstream MLLMs and implement system optimization for hardware-aware deployment to optimize model inference on mobile phones. BlueLM-V-3B boasts the following key highlights: (1) Small Size: BlueLM-V-3B features a language model with 2.7B parameters and a vision encoder with 400M parameters. (2) Fast Speed: BlueLM-V-3B achieves a generation speed of 24.4 token/s on the MediaTek Dimensity 9300 processor with 4-bit LLM weight quantization. (3) Strong Performance: BlueLM-V-3B has attained the highest average score of 66.1 on the OpenCompass benchmark among models with $\leq$ 4B parameters and surpassed a series of models with much larger parameter sizes (e.g., MiniCPM-V-2.6, InternVL2-8B).

ThinK: Thinner Key Cache by Query-Driven Pruning

Large Language Models (LLMs) have revolutionized the field of natural language processing, achieving unprecedented performance across a variety of applications by leveraging increased model sizes and sequence lengths. However, the associated rise in computational and memory costs poses significant challenges, particularly in managing long sequences due to the quadratic complexity of the transformer attention mechanism. This paper focuses on the long-context scenario, addressing the inefficiencies in KV cache memory consumption during inference. Unlike existing approaches that optimize the memory based on the sequence lengths, we uncover that the channel dimension of the KV cache exhibits significant redundancy, characterized by unbalanced magnitude distribution and low-rank structure in attention weights. Based on these observations, we propose ThinK, a novel query-dependent KV cache pruning method designed to minimize attention weight loss while selectively pruning the least significant channels. Our approach not only maintains or enhances model accuracy but also achieves a reduction in memory costs by over 20% compared with vanilla KV cache eviction methods. Extensive evaluations on the LLaMA3 and Mistral models across various long-sequence datasets confirm the efficacy of ThinK, setting a new precedent for efficient LLM deployment without compromising performance. We also outline the potential of extending our method to value cache pruning, demonstrating ThinK's versatility and broad applicability in reducing both memory and computational overheads.

FabGPT: An Efficient Large Multimodal Model for Complex Wafer Defect Knowledge Queries

Intelligence is key to advancing integrated circuit (IC) fabrication. Recent breakthroughs in Large Multimodal Models (LMMs) have unlocked unparalleled abilities in understanding images and text, fostering intelligent fabrication. Leveraging the power of LMMs, we introduce FabGPT, a customized IC fabrication large multimodal model for wafer defect knowledge query. FabGPT manifests expertise in conducting defect detection in Scanning Electron Microscope (SEM) images, performing root cause analysis, and providing expert question-answering (Q&A) on fabrication processes. FabGPT matches enhanced multimodal features to automatically detect minute defects under complex wafer backgrounds and reduce the subjectivity of manual threshold settings. Besides, the proposed modulation module and interactive corpus training strategy embed wafer defect knowledge into the pre-trained model, effectively balancing Q&A queries related to defect knowledge and original knowledge and mitigating the modality bias issues. Experiments on in-house fab data (SEM-WaD) show that our FabGPT achieves significant performance improvement in wafer defect detection and knowledge querying.

SPP: Sparsity-Preserved Parameter-Efficient Fine-Tuning for Large Language Models

Large Language Models (LLMs) have become pivotal in advancing the field of artificial intelligence, yet their immense sizes pose significant challenges for both fine-tuning and deployment. Current post-training pruning methods, while reducing the sizes of LLMs, often fail to maintain their original performance. To address these challenges, this paper introduces SPP, a Sparsity-Preserved Parameter-efficient fine-tuning method. Different from existing post-training pruning approaches that struggle with performance retention, SPP proposes to employ lightweight learnable column and row matrices to optimize sparse LLM weights, keeping the structure and sparsity of pruned pre-trained models intact. By element-wise multiplication and residual addition, SPP ensures the consistency of model sparsity pattern and ratio during both training and weight-merging processes. We demonstrate the effectiveness of SPP by applying it to the LLaMA and LLaMA-2 model families with recent post-training pruning methods. Our results show that SPP significantly enhances the performance of models with different sparsity patterns (i.e. unstructured and N:M sparsity), especially for those with high sparsity ratios (e.g. 75%), making it a promising solution for the efficient fine-tuning of sparse LLMs. Code will be made available at https://github.com/Lucky-Lance/SPP.

TerDiT: Ternary Diffusion Models with Transformers

Recent developments in large-scale pre-trained text-to-image diffusion models have significantly improved the generation of high-fidelity images, particularly with the emergence of diffusion models based on transformer architecture (DiTs). Among these diffusion models, diffusion transformers have demonstrated superior image generation capabilities, boosting lower FID scores and higher scalability. However, deploying large-scale DiT models can be expensive due to their extensive parameter numbers. Although existing research has explored efficient deployment techniques for diffusion models such as model quantization, there is still little work concerning DiT-based models. To tackle this research gap, in this paper, we propose TerDiT, a quantization-aware training (QAT) and efficient deployment scheme for ternary diffusion models with transformers. We focus on the ternarization of DiT networks and scale model sizes from 600M to 4.2B. Our work contributes to the exploration of efficient deployment strategies for large-scale DiT models, demonstrating the feasibility of training extremely low-bit diffusion transformer models from scratch while maintaining competitive image generation capacities compared to full-precision models. Code will be available at https://github.com/Lucky-Lance/TerDiT.

Not All Experts are Equal: Efficient Expert Pruning and Skipping for Mixture-of-Experts Large Language Models

A pivotal advancement in the progress of large language models (LLMs) is the emergence of the Mixture-of-Experts (MoE) LLMs. Compared to traditional LLMs, MoE LLMs can achieve higher performance with fewer parameters, but it is still hard to deploy them due to their immense parameter sizes. Different from previous weight pruning methods that rely on specifically designed hardware, this paper mainly aims to enhance the deployment efficiency of MoE LLMs by introducing plug-and-play expert-level sparsification techniques. Specifically, we propose, for the first time to our best knowledge, post-training approaches for task-agnostic and task-specific expert pruning and skipping of MoE LLMs, tailored to improve deployment efficiency while maintaining model performance across a wide range of tasks. Extensive experiments show that our proposed methods can simultaneously reduce model sizes and increase the inference speed, while maintaining satisfactory performance. Data and code will be available at https://github.com/Lucky-Lance/Expert_Sparsity.

ImageBind-LLM: Multi-modality Instruction Tuning

We present ImageBind-LLM, a multi-modality instruction tuning method of large language models (LLMs) via ImageBind. Existing works mainly focus on language and image instruction tuning, different from which, our ImageBind-LLM can respond to multi-modality conditions, including audio, 3D point clouds, video, and their embedding-space arithmetic by only image-text alignment training. During training, we adopt a learnable bind network to align the embedding space between LLaMA and ImageBind's image encoder. Then, the image features transformed by the bind network are added to word tokens of all layers in LLaMA, which progressively injects visual instructions via an attention-free and zero-initialized gating mechanism. Aided by the joint embedding of ImageBind, the simple image-text training enables our model to exhibit superior multi-modality instruction-following capabilities. During inference, the multi-modality inputs are fed into the corresponding ImageBind encoders, and processed by a proposed visual cache model for further cross-modal embedding enhancement. The training-free cache model retrieves from three million image features extracted by ImageBind, which effectively mitigates the training-inference modality discrepancy. Notably, with our approach, ImageBind-LLM can respond to instructions of diverse modalities and demonstrate significant language generation quality. Code is released at https://github.com/OpenGVLab/LLaMA-Adapter.

Zero-shot information extraction from radiological reports using ChatGPT

Electronic health records contain an enormous amount of valuable information, but many are recorded in free text. Information extraction is the strategy to transform the sequence of characters into structured data, which can be employed for secondary analysis. However, the traditional information extraction components, such as named entity recognition and relation extraction, require annotated data to optimize the model parameters, which has become one of the major bottlenecks in building information extraction systems. With the large language models achieving good performances on various downstream NLP tasks without parameter tuning, it becomes possible to use large language models for zero-shot information extraction. In this study, we aim to explore whether the most popular large language model, ChatGPT, can extract useful information from the radiological reports. We first design the prompt template for the interested information in the CT reports. Then, we generate the prompts by combining the prompt template with the CT reports as the inputs of ChatGPT to obtain the responses. A post-processing module is developed to transform the responses into structured extraction results. We conducted the experiments with 847 CT reports collected from Peking University Cancer Hospital. The experimental results indicate that ChatGPT can achieve competitive performances for some extraction tasks compared with the baseline information extraction system, but some limitations need to be further improved.

From Isolated Islands to Pangea: Unifying Semantic Space for Human Action Understanding

Action understanding matters and attracts attention. It can be formed as the mapping from the action physical space to the semantic space. Typically, researchers built action datasets according to idiosyncratic choices to define classes and push the envelope of benchmarks respectively. Thus, datasets are incompatible with each other like "Isolated Islands" due to semantic gaps and various class granularities, e.g., do housework in dataset A and wash plate in dataset B. We argue that a more principled semantic space is an urgent need to concentrate the community efforts and enable us to use all datasets together to pursue generalizable action learning. To this end, we design a Poincare action semantic space given verb taxonomy hierarchy and covering massive actions. By aligning the classes of previous datasets to our semantic space, we gather (image/video/skeleton/MoCap) datasets into a unified database in a unified label system, i.e., bridging "isolated islands" into a "Pangea". Accordingly, we propose a bidirectional mapping model between physical and semantic space to fully use Pangea. In extensive experiments, our system shows significant superiority, especially in transfer learning. Code and data will be made publicly available.

Towards Multi-perspective conformance checking with fuzzy sets

Conformance checking techniques are widely adopted to pinpoint possible discrepancies between process models and the execution of the process in reality. However, state of the art approaches adopt a crisp evaluation of deviations, with the result that small violations are considered at the same level of significant ones. This affects the quality of the provided diagnostics, especially when there exists some tolerance with respect to reasonably small violations, and hampers the flexibility of the process. In this work, we propose a novel approach which allows to represent actors' tolerance with respect to violations and to account for severity of deviations when assessing executions compliance. We argue that besides improving the quality of the provided diagnostics, allowing some tolerance in deviations assessment also enhances the flexibility of conformance checking techniques and, indirectly, paves the way for improving the resilience of the overall process management system.