Daily Papers

The energy consumption of large-scale ML models is dominated by data movement - shuffling billions of parameters across memory hierarchies and data centers. Effective sparsification to prune redundant parameters is still challenging: existing methods incur significant accuracy degradation, performance overhead, or both. We introduce (Bl)ock (a)nd (S)parse (T)ransformers (BLaST), a general, robust, and reliable sparsification method applicable to linear layers in all settings. Our method iteratively sparsifies weight matrices into a block sparsity pattern suitable for efficient sparse matrix-matrix (SpMM) multiplication. BLaST achieves up to 95% sparsity in MLP weights with negligible accuracy loss. Our fused, highly optimized Sparse MLP kernel delivers up to 16.7x speedup over dense MLPs across 9 architectures and 8 datasets, resulting in up to 1.6x inference speedup, 1.11x pretraining speedup and up to 3.12x inference memory usage reduction. BLaST enables the next generation of large-scale AI systems by reducing energy use, memory footprint, and latency.

Large language models (LLMs) achieved remarkable performance across various tasks. However, they face challenges in managing long documents and extended conversations, due to significantly increased computational requirements, both in memory and inference time, and potential context truncation when the input exceeds the LLM's fixed context length. This paper proposes a method called Selective Context that enhances the inference efficiency of LLMs by identifying and pruning redundancy in the input context to make the input more compact. We test our approach using common data sources requiring long context processing: arXiv papers, news articles, and long conversations, on tasks of summarisation, question answering, and response generation. Experimental results show that Selective Context significantly reduces memory cost and decreases generation latency while maintaining comparable performance compared to that achieved when full context is used. Specifically, we achieve a 50\% reduction in context cost, resulting in a 36\% reduction in inference memory usage and a 32\% reduction in inference time, while observing only a minor drop of .023 in BERTscore and .038 in faithfulness on four downstream applications, indicating that our method strikes a good balance between efficiency and performance.

Large Language Models (LLMs), despite their impressive performance on a wide range of tasks, require significant GPU memory and consume substantial computational resources. In addition to model weights, the memory occupied by KV cache increases linearly with sequence length, becoming a main bottleneck for inference. In this paper, we introduce a novel approach for optimizing the KV cache which significantly reduces its memory footprint. Through a comprehensive investigation, we find that on LLaMA2 series models, (i) the similarity between adjacent tokens' query vectors is remarkably high, and (ii) current query's attention calculation can rely solely on the attention information of a small portion of the preceding queries. Based on these observations, we propose CORM, a KV cache eviction policy that dynamically retains important key-value pairs for inference without finetuning the model. We validate that CORM reduces the inference memory usage of KV cache by up to 70% without noticeable performance degradation across six tasks in LongBench.

Large language models(LLMs) have sparked a new wave of exciting AI applications. Hosting these models at scale requires significant memory resources. One crucial memory bottleneck for the deployment stems from the context window. It is commonly recognized that model weights are memory hungry; however, the size of key-value embedding stored during the generation process (KV cache) can easily surpass the model size. The enormous size of the KV cache puts constraints on the inference batch size, which is crucial for high throughput inference workload. Inspired by an interesting observation of the attention scores, we hypothesize the persistence of importance: only pivotal tokens, which had a substantial influence at one step, will significantly influence future generations. Based on our empirical verification and theoretical analysis around this hypothesis, we propose Scissorhands, a system that maintains the memory usage of the KV cache at a fixed budget without finetuning the model. In essence, Scissorhands manages the KV cache by storing the pivotal tokens with a higher probability. We validate that Scissorhands reduces the inference memory usage of the KV cache by up to 5X without compromising model quality. We further demonstrate that Scissorhands can be combined with 4-bit quantization, traditionally used to compress model weights, to achieve up to 20X compression.

Machine learning models deployed on edge devices have enabled numerous exciting new applications, such as humanoid robots, AR glasses, and autonomous vehicles. However, the computing resources available on these edge devices are not catching up with the ever-growing number of parameters in these models. As the models become bigger and more complicated, the novel yet sophisticated structure challenges the inference runtime optimization. We present FluidML, a generic runtime memory management and optimization framework that can flexibly transform the model execution blueprint to achieve faster and more memory-efficient inference. Evaluations across different platforms show that FluidML can consistently reduce the end-to-end inference latency by up to 25.38% for popular language models and reduce peak memory usage by up to 41.47%, compared to state-of-the-art approaches. FluidML is of ~30K line of codes, built for general-purpose usage, and will be released as an open-source inference runtime optimization framework to the community.

Long-context language models exhibit impressive performance but remain challenging to deploy due to high GPU memory demands during inference. We propose Memory-efficient Offloaded Mini-sequence Inference (MOM), a method that partitions critical layers into smaller "mini-sequences" and integrates seamlessly with KV cache offloading. Experiments on various Llama, Qwen, and Mistral models demonstrate that MOM reduces peak memory usage by over 50\% on average. On Meta-Llama-3.2-8B, MOM extends the maximum context length from 155k to 455k tokens on a single A100 80GB GPU, while keeping outputs identical and not compromising accuracy. MOM also maintains highly competitive throughput due to minimal computational overhead and efficient last-layer processing. Compared to traditional chunked prefill methods, MOM achieves a 35\% greater context length extension. More importantly, our method drastically reduces prefill memory consumption, eliminating it as the longstanding dominant memory bottleneck during inference. This breakthrough fundamentally changes research priorities, redirecting future efforts from prefill-stage optimizations to improving decode-stage residual KV cache efficiency.

In recent years, Mixture-of-Experts (MoE) has emerged as an effective approach for enhancing the capacity of deep neural network (DNN) with sub-linear computational costs. However, storing all experts on GPUs incurs significant memory overhead, increasing the monetary cost of MoE-based inference. To address this, we propose eMoE, a memory efficient inference system for MoE-based large language models (LLMs) by leveraging our observations from experiment measurements. eMoE reduces memory usage by predicting and loading only the required experts based on recurrent patterns in expert routing. To reduce loading latency while maintaining accuracy, as we found using the same experts for subsequent prompts has minimal impact on perplexity, eMoE invokes the expert predictor every few prompts rather than for each prompt. In addition, it skips predictions for tasks less sensitive to routing accuracy. Finally, it has task-aware scheduling to minimize inference latency by considering Service Level Objectives (SLOs), task-specific output lengths, and expert loading latencies. Experimental results show that compared to existing systems, eMoE reduces memory consumption by up to 80% while maintaining accuracy and reduces inference latency by up to 17%. It also enables processing prompts 40x longer, batches 4.5x larger, and achieves 1.5x higher throughput.

The self-attention mechanism in the transformer architecture is capable of capturing long-range dependencies and it is the main reason behind its effectiveness in processing sequential data. Nevertheless, despite their success, transformers have two significant drawbacks that still limit their broader applicability: (1) In order to remember past information, the self-attention mechanism requires access to the whole history to be provided as context. (2) The inference cost in transformers is expensive. In this paper we introduce recurrent alternatives to the transformer self-attention mechanism that offer a context-independent inference cost, leverage long-range dependencies effectively, and perform well in practice. We evaluate our approaches in reinforcement learning problems where the aforementioned computational limitations make the application of transformers nearly infeasible. We quantify the impact of the different components of our architecture in a diagnostic environment and assess performance gains in 2D and 3D pixel-based partially-observable environments. When compared to a state-of-the-art architecture, GTrXL, inference in our approach is at least 40% cheaper while reducing memory use in more than 50%. Our approach either performs similarly or better than GTrXL, improving more than 37% upon GTrXL performance on harder tasks.

Super-resolution (SR) and image generation are important tasks in computer vision and are widely adopted in real-world applications. Most existing methods, however, generate images only at fixed-scale magnification and suffer from over-smoothing and artifacts. Additionally, they do not offer enough diversity of output images nor image consistency at different scales. Most relevant work applied Implicit Neural Representation (INR) to the denoising diffusion model to obtain continuous-resolution yet diverse and high-quality SR results. Since this model operates in the image space, the larger the resolution of image is produced, the more memory and inference time is required, and it also does not maintain scale-specific consistency. We propose a novel pipeline that can super-resolve an input image or generate from a random noise a novel image at arbitrary scales. The method consists of a pretrained auto-encoder, a latent diffusion model, and an implicit neural decoder, and their learning strategies. The proposed method adopts diffusion processes in a latent space, thus efficient, yet aligned with output image space decoded by MLPs at arbitrary scales. More specifically, our arbitrary-scale decoder is designed by the symmetric decoder w/o up-scaling from the pretrained auto-encoder, and Local Implicit Image Function (LIIF) in series. The latent diffusion process is learnt by the denoising and the alignment losses jointly. Errors in output images are backpropagated via the fixed decoder, improving the quality of output images. In the extensive experiments using multiple public benchmarks on the two tasks i.e. image super-resolution and novel image generation at arbitrary scales, the proposed method outperforms relevant methods in metrics of image quality, diversity and scale consistency. It is significantly better than the relevant prior-art in the inference speed and memory usage.

Recent work on tokenizer-free multilingual pretrained models show promising results in improving cross-lingual transfer and reducing engineering overhead (Clark et al., 2022; Xue et al., 2022). However, these works mainly focus on reporting accuracy on a limited set of tasks and data settings, placing less emphasis on other important factors when tuning and deploying the models in practice, such as memory usage, inference speed, and fine-tuning data robustness. We attempt to fill this gap by performing a comprehensive empirical comparison of multilingual tokenizer-free and subword-based models considering these various dimensions. Surprisingly, we find that subword-based models might still be the most practical choice in many settings, achieving better performance for lower inference latency and memory usage. Based on these results, we encourage future work in tokenizer-free methods to consider these factors when designing and evaluating new models.

Transformer-based large language models (LLMs) use the key-value (KV) cache to significantly accelerate inference by storing the key and value embeddings of past tokens. However, this cache consumes significant GPU memory. In this work, we introduce HashEvict, an algorithm that uses locality-sensitive hashing (LSH) to compress the KV cache. HashEvict quickly locates tokens in the cache that are cosine dissimilar to the current query token. This is achieved by computing the Hamming distance between binarized Gaussian projections of the current token query and cached token keys, with a projection length much smaller than the embedding dimension. We maintain a lightweight binary structure in GPU memory to facilitate these calculations. Unlike existing compression strategies that compute attention to determine token retention, HashEvict makes these decisions pre-attention, thereby reducing computational costs. Additionally, HashEvict is dynamic - at every decoding step, the key and value of the current token replace the embeddings of a token expected to produce the lowest attention score. We demonstrate that HashEvict can compress the KV cache by 30%-70% while maintaining high performance across reasoning, multiple-choice, long-context retrieval and summarization tasks.

Visual token reduction lowers inference costs caused by extensive image features in large vision-language models (LVLMs). Unlike relevant studies that prune tokens in self-attention-only LVLMs, our work uniquely addresses cross-attention-based models, which achieve superior performance. We identify that the key-value (KV) cache size for image tokens in cross-attention layers significantly exceeds that of text tokens in self-attention layers, posing a major compute bottleneck. To mitigate this issue, we exploit the sparse nature in cross-attention maps to selectively prune redundant visual features. Our Trimmed Llama effectively reduces KV cache demands without requiring additional training. By benefiting from 50%-reduced visual features, our model can reduce inference latency and memory usage while achieving benchmark parity.

Transformer-based text embedding models have improved their performance on benchmarks like MIRACL and BEIR by increasing their parameter counts. However, this scaling approach introduces significant deployment challenges, including increased inference latency and memory usage. These challenges are particularly severe in retrieval-augmented generation (RAG) applications, where large models' increased memory requirements constrain dataset ingestion capacity, and their higher latency directly impacts query-time performance. While causal language models have addressed similar efficiency challenges using Mixture of Experts (MoE) architectures, this approach hasn't been successfully adapted to the general text embedding setting. In this paper, we introduce Nomic Embed v2, the first general purpose MoE text embedding model. Our model outperforms models in the same parameter class on both monolingual and multilingual benchmarks while also maintaining competitive performance with models twice its size. We open-source all code, models, and evaluation data to ensure full reproducibility of our training pipeline.

Solid results from Transformers have made them prevailing architectures in various natural language and vision tasks. As a default component in Transformers, Layer Normalization (LN) normalizes activations within each token to boost the robustness. However, LN requires on-the-fly statistics calculation in inference as well as division and square root operations, leading to inefficiency on hardware. What is more, replacing LN with other hardware-efficient normalization schemes (e.g., Batch Normalization) results in inferior performance, even collapse in training. We find that this dilemma is caused by abnormal behaviors of activation statistics, including large fluctuations over iterations and extreme outliers across layers. To tackle these issues, we propose Unified Normalization (UN), which can speed up the inference by being fused with other linear operations and achieve comparable performance on par with LN. UN strives to boost performance by calibrating the activation and gradient statistics with a tailored fluctuation smoothing strategy. Meanwhile, an adaptive outlier filtration strategy is applied to avoid collapse in training whose effectiveness is theoretically proved and experimentally verified in this paper. We demonstrate that UN can be an efficient drop-in alternative to LN by conducting extensive experiments on language and vision tasks. Besides, we evaluate the efficiency of our method on GPU. Transformers equipped with UN enjoy about 31% inference speedup and nearly 18% memory reduction. Code will be released at https://github.com/hikvision-research/Unified-Normalization.

Prompt tuning (PT), where a small amount of trainable soft (continuous) prompt vectors is affixed to the input of language models (LM), has shown promising results across various tasks and models for parameter-efficient fine-tuning (PEFT). PT stands out from other PEFT approaches because it maintains competitive performance with fewer trainable parameters and does not drastically scale up its parameters as the model size expands. However, PT introduces additional soft prompt tokens, leading to longer input sequences, which significantly impacts training and inference time and memory usage due to the Transformer's quadratic complexity. Particularly concerning for Large Language Models (LLMs) that face heavy daily querying. To address this issue, we propose Decomposed Prompt Tuning (DePT), which decomposes the soft prompt into a shorter soft prompt and a pair of low-rank matrices that are then optimised with two different learning rates. This allows DePT to achieve better performance while saving over 20% memory and time costs compared to vanilla PT and its variants, without changing trainable parameter sizes. Through extensive experiments on 23 natural language processing (NLP) and vision-language (VL) tasks, we demonstrate that DePT outperforms state-of-the-art PEFT approaches, including the full fine-tuning baseline in some scenarios. Additionally, we empirically show that DEPT grows more efficient as the model size increases. Our further study reveals that DePT integrates seamlessly with parameter-efficient transfer learning in the few-shot learning setting and highlights its adaptability to various model architectures and sizes.

In this technical report, we present Falcon Mamba 7B, a new base large language model based on the novel Mamba architecture. Falcon Mamba 7B is trained on 5.8 trillion tokens with carefully selected data mixtures. As a pure Mamba-based model, Falcon Mamba 7B surpasses leading open-weight models based on Transformers, such as Mistral 7B, Llama3.1 8B, and Falcon2 11B. It is on par with Gemma 7B and outperforms models with different architecture designs, such as RecurrentGemma 9B and RWKV-v6 Finch 7B/14B. Currently, Falcon Mamba 7B is the best-performing Mamba model in the literature at this scale, surpassing both existing Mamba and hybrid Mamba-Transformer models, according to the Open LLM Leaderboard. Due to its architecture, Falcon Mamba 7B is significantly faster at inference and requires substantially less memory for long sequence generation. Despite recent studies suggesting that hybrid Mamba-Transformer models outperform pure architecture designs, we demonstrate that even the pure Mamba design can achieve similar, or even superior results compared to the Transformer and hybrid designs. We make the weights of our implementation of Falcon Mamba 7B publicly available on https://huggingface.co/tiiuae/falcon-mamba-7b, under a permissive license.

Video large language models (VideoLLMs) have demonstrated the capability to process longer video inputs and enable complex reasoning and analysis. However, due to the thousands of visual tokens from the video frames, key-value (KV) cache can significantly increase memory requirements, becoming a bottleneck for inference speed and memory usage. KV cache quantization is a widely used approach to address this problem. In this paper, we find that 2-bit KV quantization of VideoLLMs can hardly hurt the model performance, while the limit of KV cache quantization in even lower bits has not been investigated. To bridge this gap, we introduce VidKV, a plug-and-play KV cache quantization method to compress the KV cache to lower than 2 bits. Specifically, (1) for key, we propose a mixed-precision quantization strategy in the channel dimension, where we perform 2-bit quantization for anomalous channels and 1-bit quantization combined with FFT for normal channels; (2) for value, we implement 1.58-bit quantization while selectively filtering semantically salient visual tokens for targeted preservation, for a better trade-off between precision and model performance. Importantly, our findings suggest that the value cache of VideoLLMs should be quantized in a per-channel fashion instead of the per-token fashion proposed by prior KV cache quantization works for LLMs. Empirically, extensive results with LLaVA-OV-7B and Qwen2.5-VL-7B on six benchmarks show that VidKV effectively compresses the KV cache to 1.5-bit and 1.58-bit precision with almost no performance drop compared to the FP16 counterparts.

Recent advancements in deep generative models present new opportunities for music production but also pose challenges, such as high computational demands and limited audio quality. Moreover, current systems frequently rely solely on text input and typically focus on producing complete musical pieces, which is incompatible with existing workflows in music production. To address these issues, we introduce "Diff-A-Riff," a Latent Diffusion Model designed to generate high-quality instrumental accompaniments adaptable to any musical context. This model offers control through either audio references, text prompts, or both, and produces 48kHz pseudo-stereo audio while significantly reducing inference time and memory usage. We demonstrate the model's capabilities through objective metrics and subjective listening tests, with extensive examples available on the accompanying website: sonycslparis.github.io/diffariff-companion/

In recent years, multilingual machine translation models have achieved promising performance on low-resource language pairs by sharing information between similar languages, thus enabling zero-shot translation. To overcome the "curse of multilinguality", these models often opt for scaling up the number of parameters, which makes their use in resource-constrained environments challenging. We introduce SMaLL-100, a distilled version of the M2M-100 (12B) model, a massively multilingual machine translation model covering 100 languages. We train SMaLL-100 with uniform sampling across all language pairs and therefore focus on preserving the performance of low-resource languages. We evaluate SMaLL-100 on different low-resource benchmarks: FLORES-101, Tatoeba, and TICO-19 and demonstrate that it outperforms previous massively multilingual models of comparable sizes (200-600M) while improving inference latency and memory usage. Additionally, our model achieves comparable results to M2M-100 (1.2B), while being 3.6x smaller and 4.3x faster at inference. Code and pre-trained models: https://github.com/alirezamshi/small100

The application of Transformer-based large models has achieved numerous success in recent years. However, the exponential growth in the parameters of large models introduces formidable memory challenge for edge deployment. Prior works to address this challenge mainly focus on optimizing the model structure and adopting memory swapping methods. However, the former reduces the inference accuracy, and the latter raises the inference latency. This paper introduces PIPELOAD, a novel memory-efficient pipeline execution mechanism. It reduces memory usage by incorporating dynamic memory management and minimizes inference latency by employing parallel model loading. Based on PIPELOAD mechanism, we present Hermes, a framework optimized for large model inference on edge devices. We evaluate Hermes on Transformer-based models of different sizes. Our experiments illustrate that Hermes achieves up to 4.24 X increase in inference speed and 86.7% lower memory consumption than the state-of-the-art pipeline mechanism for BERT and ViT models, 2.58 X increase in inference speed and 90.3% lower memory consumption for GPT-style models.

Tiny deep learning on microcontroller units (MCUs) is challenging due to the limited memory size. We find that the memory bottleneck is due to the imbalanced memory distribution in convolutional neural network (CNN) designs: the first several blocks have an order of magnitude larger memory usage than the rest of the network. To alleviate this issue, we propose a generic patch-by-patch inference scheduling, which operates only on a small spatial region of the feature map and significantly cuts down the peak memory. However, naive implementation brings overlapping patches and computation overhead. We further propose network redistribution to shift the receptive field and FLOPs to the later stage and reduce the computation overhead. Manually redistributing the receptive field is difficult. We automate the process with neural architecture search to jointly optimize the neural architecture and inference scheduling, leading to MCUNetV2. Patch-based inference effectively reduces the peak memory usage of existing networks by 4-8x. Co-designed with neural networks, MCUNetV2 sets a record ImageNet accuracy on MCU (71.8%), and achieves >90% accuracy on the visual wake words dataset under only 32kB SRAM. MCUNetV2 also unblocks object detection on tiny devices, achieving 16.9% higher mAP on Pascal VOC compared to the state-of-the-art result. Our study largely addressed the memory bottleneck in tinyML and paved the way for various vision applications beyond image classification.

The performance of neural networks improves when more parameters are used. However, the model sizes are constrained by the available on-device memory during training and inference. Although applying techniques like quantization can alleviate the constraint, they suffer from performance degradation. In this work, we introduce NeuZip, a new weight compression scheme based on the entropy of floating-point numbers in neural networks. With NeuZip, we are able to achieve memory-efficient training and inference without sacrificing performance. Notably, we significantly reduce the memory footprint of training a Llama-3 8B model from 31GB to less than 16GB, while keeping the training dynamics fully unchanged. In inference, our method can reduce memory usage by more than half while maintaining near-lossless performance. Our code is publicly available.

Large Language Models (LLMs) face challenges for on-device inference due to high memory demands. Traditional methods to reduce memory usage often compromise performance and lack adaptability. We propose FlexInfer, an optimized offloading framework for on-device inference, addressing these issues with techniques like asynchronous prefetching, balanced memory locking, and flexible tensor preservation. These strategies enhance memory efficiency and mitigate I/O bottlenecks, ensuring high performance within user-specified resource constraints. Experiments demonstrate that FlexInfer significantly improves throughput under limited resources, achieving up to 12.5 times better performance than existing methods and facilitating the deployment of large models on resource-constrained devices.

Transformer-based large language models (LLMs) demonstrate impressive performance in long context generation. Extending the context length has disproportionately shifted the memory footprint of LLMs during inference to the key-value cache (KV cache). In this paper, we propose HEADINFER, which offloads the KV cache to CPU RAM while avoiding the need to fully store the KV cache for any transformer layer on the GPU. HEADINFER employs a fine-grained, head-wise offloading strategy, maintaining only selective attention heads KV cache on the GPU while computing attention output dynamically. Through roofline analysis, we demonstrate that HEADINFER maintains computational efficiency while significantly reducing memory footprint. We evaluate HEADINFER on the Llama-3-8B model with a 1-million-token sequence, reducing the GPU memory footprint of the KV cache from 128 GB to 1 GB and the total GPU memory usage from 207 GB to 17 GB, achieving a 92% reduction compared to BF16 baseline inference. Notably, HEADINFER enables 4-million-token inference with an 8B model on a single consumer GPU with 24GB memory (e.g., NVIDIA RTX 4090) without approximation methods.

The rapid growth in machine learning models, especially in natural language processing and computer vision, has led to challenges when running these models on hardware with limited resources. This paper introduces Superpipeline, a new framework designed to optimize the execution of large AI models on constrained hardware during both training and inference. Our approach involves dynamically managing model execution by dividing models into individual layers and efficiently transferring these layers between GPU and CPU memory. Superpipeline reduces GPU memory usage by up to 60% in our experiments while maintaining model accuracy and acceptable processing speeds. This allows models that would otherwise exceed available GPU memory to run effectively. Unlike existing solutions that focus mainly on inference or specific model types, Superpipeline can be applied to large language models (LLMs), vision-language models (VLMs), and vision-based models. We tested Superpipeline's performance across various models and hardware setups. The method includes two key parameters that allow fine-tuning the balance between GPU memory use and processing speed. Importantly, Superpipeline does not require retraining or changing model parameters, ensuring that the original model's output remains unchanged. Superpipeline's simplicity and flexibility make it useful for researchers and professionals working with advanced AI models on limited hardware. It enables the use of larger models or bigger batch sizes on existing hardware, potentially speeding up innovation across many machine learning applications. This work marks an important step toward making advanced AI models more accessible and optimizing their deployment in resource-limited environments. The code for Superpipeline is available at https://github.com/abbasiReza/super-pipeline.

The rapid progress in artificial intelligence-generated content (AIGC), especially with diffusion models, has significantly advanced development of high-quality video generation. However, current video diffusion models exhibit demanding computational requirements and high peak memory usage, especially for generating longer and higher-resolution videos. These limitations greatly hinder the practical application of video diffusion models on standard hardware platforms. To tackle this issue, we present a novel, training-free framework named Streamlined Inference, which leverages the temporal and spatial properties of video diffusion models. Our approach integrates three core components: Feature Slicer, Operator Grouping, and Step Rehash. Specifically, Feature Slicer effectively partitions input features into sub-features and Operator Grouping processes each sub-feature with a group of consecutive operators, resulting in significant memory reduction without sacrificing the quality or speed. Step Rehash further exploits the similarity between adjacent steps in diffusion, and accelerates inference through skipping unnecessary steps. Extensive experiments demonstrate that our approach significantly reduces peak memory and computational overhead, making it feasible to generate high-quality videos on a single consumer GPU (e.g., reducing peak memory of AnimateDiff from 42GB to 11GB, featuring faster inference on 2080Ti).

The size of the key-value (KV) cache plays a critical role in determining both the maximum context length and the number of concurrent requests supported during inference in modern language models. The KV cache size grows proportionally with the number of attention heads and the tokens processed, leading to increased memory consumption and slower inference for long inputs. In this work, we explore the use of MixAttention, a model architecture modification closely related to a blog published by Character.AI. MixAttention combines sliding window attention, where only a small subset of recent tokens is stored in the KV cache, with KV cache sharing across layers. Our experiments demonstrate that MixAttention significantly reduces memory usage and improves inference speed without sacrificing model performance in both short and long-context tasks. We also explore various configurations of this architecture, identifying those that maintain quality across evaluation metrics while optimizing resource efficiency.

The growing computational demands posed by increasingly number of neural network's parameters necessitate low-memory-consumption training approaches. Previous memory reduction techniques, such as Low-Rank Adaptation (LoRA) and ReLoRA, suffer from the limitation of low rank and saddle point issues, particularly during intensive tasks like pre-training. In this paper, we propose Sparse Spectral Training (SST), an advanced training methodology that updates all singular values and selectively updates singular vectors of network weights, thereby optimizing resource usage while closely approximating full-rank training. SST refines the training process by employing a targeted updating strategy for singular vectors, which is determined by a multinomial sampling method weighted by the significance of the singular values, ensuring both high performance and memory reduction. Through comprehensive testing on both Euclidean and hyperbolic neural networks across various tasks, including natural language generation, machine translation, node classification and link prediction, SST demonstrates its capability to outperform existing memory reduction training methods and is comparable with full-rank training in some cases. On OPT-125M, with rank equating to 8.3% of embedding dimension, SST reduces the perplexity gap to full-rank training by 67.6%, demonstrating a significant reduction of the performance loss with prevalent low-rank methods. This approach offers a strong alternative to traditional training techniques, paving the way for more efficient and scalable neural network training solutions.

Test-time compute scaling has emerged as a new axis along which to improve model accuracy, where additional computation is used at inference time to allow the model to think longer for more challenging problems. One promising approach for test-time compute scaling is search against a process reward model, where a model generates multiple potential candidates at each step of the search, and these partial trajectories are then scored by a separate reward model in order to guide the search process. The diversity of trajectories in the tree search process affects the accuracy of the search, since increasing diversity promotes more exploration. However, this diversity comes at a cost, as divergent trajectories have less KV sharing, which means they consume more memory and slow down the search process. Previous search methods either do not perform sufficient exploration, or else explore diverse trajectories but have high latency. We address this challenge by proposing Efficient Tree Search (ETS), which promotes KV sharing by pruning redundant trajectories while maintaining necessary diverse trajectories. ETS incorporates a linear programming cost model to promote KV cache sharing by penalizing the number of nodes retained, while incorporating a semantic coverage term into the cost model to ensure that we retain trajectories which are semantically different. We demonstrate how ETS can achieve 1.8times reduction in average KV cache size during the search process, leading to 1.4times increased throughput relative to prior state-of-the-art methods, with minimal accuracy degradation and without requiring any custom kernel implementation. Code is available at: https://github.com/SqueezeAILab/ETS.

Retrieval-augmented generation (RAG) extends large language models (LLMs) with external data sources to enhance factual correctness and domain coverage. Modern RAG pipelines rely on large datastores, leading to system challenges in latency-sensitive deployments, especially when limited GPU memory is available. To address these challenges, we propose TeleRAG, an efficient inference system that reduces RAG latency with minimal GPU memory requirements. The core innovation of TeleRAG is lookahead retrieval, a prefetching mechanism that anticipates required data and transfers it from CPU to GPU in parallel with LLM generation. By leveraging the modularity of RAG pipelines, the inverted file index (IVF) search algorithm and similarities between queries, TeleRAG optimally overlaps data movement and computation. Experimental results show that TeleRAG reduces end-to-end RAG inference latency by up to 1.72x on average compared to state-of-the-art systems, enabling faster, more memory-efficient deployments of advanced RAG applications.

Mixture-of-Experts (MoE) models have gained popularity in achieving state-of-the-art performance in a wide range of tasks in computer vision and natural language processing. They effectively expand the model capacity while incurring a minimal increase in computation cost during training. However, deploying such models for inference is difficult due to their large size and complex communication pattern. In this work, we provide a characterization of two MoE workloads, namely Language Modeling (LM) and Machine Translation (MT) and identify their sources of inefficiencies at deployment. We propose three optimization techniques to mitigate sources of inefficiencies, namely (1) Dynamic gating, (2) Expert Buffering, and (3) Expert load balancing. We show that dynamic gating improves maximum throughput by 6.21-11.23times for LM, 5.75-10.98times for MT Encoder and 2.58-5.71times for MT Decoder. It also reduces memory usage by up to 1.36times for LM and up to 1.1times for MT. We further propose Expert Buffering, a new caching mechanism that only keeps hot, active experts in GPU memory while buffering the rest in CPU memory. This reduces static memory allocation by up to 1.47times. We finally propose a load balancing methodology that provides additional scalability to the workload.

LLM inference for popular enterprise use cases, such as summarization, RAG, and code-generation, typically observes orders of magnitude longer prompt lengths than generation lengths. This characteristic leads to high cost of prefill and increased response latency. In this paper, we present SwiftKV, a novel model transformation and distillation procedure specifically designed to reduce the time and cost of processing prompt tokens while preserving high quality of generated tokens. SwiftKV combines three key mechanisms: i) SingleInputKV, which prefills later layers' KV cache using a much earlier layer's output, allowing prompt tokens to skip much of the model computation, ii) AcrossKV, which merges the KV caches of neighboring layers to reduce the memory footprint and support larger batch size for higher throughput, and iii) a knowledge-preserving distillation procedure that can adapt existing LLMs for SwiftKV with minimal accuracy impact and low compute and data requirement. For Llama-3.1-8B and 70B, SwiftKV reduces the compute requirement of prefill by 50% and the memory requirement of the KV cache by 62.5% while incurring minimum quality degradation across a wide range of tasks. In the end-to-end inference serving using an optimized vLLM implementation, SwiftKV realizes up to 2x higher aggregate throughput and 60% lower time per output token. It can achieve a staggering 560 TFlops/GPU of normalized inference throughput, which translates to 16K tokens/s for Llama-3.1-70B in 16-bit precision on 4x H100 GPUs.

The increasing context window size in large language models (LLMs) has improved their ability to handle complex, long-text tasks. However, as the conversation rounds continue, it is required to store a large amount of KV cache in GPU memory, which significantly affects the efficiency and even availability of the model serving systems. This paper analyzes dialogue data from real users and discovers that the LLM inference manifests a watershed layer, after which the distribution of round-level attention shows notable similarity. We propose Round Attention, a novel round-level attention mechanism that only recalls and computes the KV cache of the most relevant rounds. The experiments show that our method saves 55\% memory usage without compromising model performance.

Optimizing the Key-Value (KV) cache of the Large Language Model (LLM) has been considered critical to saving the cost of inference. Most of the existing KV-cache compression algorithms attempted to sparsify the sequence of tokens by taking advantage of the different importance of tokens. In this work, we found that by identifying the importance of attention layers, we could optimize the KV-cache jointly from two dimensions. Based on our observations regarding layer-wise importance in inference, we propose SqueezeAttention to precisely optimize the allocation of KV-cache budget among layers on-the-fly and then incorporate three representative token sparsification algorithms to compress the KV-cache for each layer with its very own budget. By optimizing the KV-cache from both sequence's and layer's dimensions, SqueezeAttention achieves around 30% to 70% of the memory reductions and up to 2.2 times of throughput improvements in a wide range of LLMs and benchmarks. The code is available at https://github.com/hetailang/SqueezeAttention.

MLLMs have demonstrated remarkable comprehension and reasoning capabilities with complex language and visual data. These advances have spurred the vision of establishing a generalist robotic MLLM proficient in understanding complex human instructions and accomplishing various embodied tasks. However, developing MLLMs for real-world robots is challenging due to the typically limited computation and memory capacities available on robotic platforms. In contrast, the inference of MLLMs involves storing billions of parameters and performing tremendous computation, imposing significant hardware demands. In our paper, we propose a Dynamic Early-Exit Framework for Robotic Vision-Language-Action Model (DeeR-VLA, or simply DeeR) that automatically adjusts the size of the activated MLLM based on each situation at hand. The approach leverages a multi-exit architecture in MLLMs, which allows the model to terminate processing once a proper size of the model has been activated for a specific situation, thus avoiding further redundant computation. Additionally, we develop novel algorithms that establish early-termination criteria for DeeR, conditioned on predefined demands such as average computational cost (i.e., power consumption), as well as peak computational consumption (i.e., latency) and GPU memory usage. These enhancements ensure that DeeR operates efficiently under varying resource constraints while maintaining competitive performance. On the CALVIN robot manipulation benchmark, DeeR demonstrates significant reductions in computational costs of LLM by 5.2-6.5x and GPU memory of LLM by 2-6x without compromising performance. Code and checkpoints are available at https://github.com/yueyang130/DeeR-VLA.

Auto-regressive inference of transformers benefit greatly from Key-Value (KV) caching, but can lead to major memory bottlenecks as model size, batch size, and sequence length grow at scale. We introduce Multi-Layer Key-Value (MLKV) sharing, a novel approach extending KV sharing across transformer layers to reduce memory usage beyond what was possible with Multi-Query Attention (MQA) and Grouped-Query Attention (GQA). Evaluations on various NLP benchmarks and inference metrics using uptrained Pythia-160M variants demonstrate that MLKV significantly reduces memory usage with minimal performance loss, reducing KV cache size down to a factor of 6x compared to MQA. These results highlight MLKV's potential for efficient deployment of transformer models at scale. We provide code at https://github.com/zaydzuhri/pythia-mlkv

Model compression has emerged as a mainstream solution to reduce memory usage and computational overhead. This paper presents Group Quantization and Sparse Acceleration (GQSA), a novel compression technique tailored for LLMs. Traditional methods typically focus exclusively on either quantization or sparsification, but relying on a single strategy often results in significant performance loss at high compression rates. In contrast, GQSA integrates quantization and sparsification in a tightly coupled manner, leveraging GPU-friendly structured group sparsity and quantization for efficient acceleration. Building upon system-algorithm co-design principles, we propose a two-stage sparse optimization strategy that ensures the performance superiority of the compressed model. On the engine side, we introduce a "task-centric" parallel strategy, which, to the best of our knowledge, is the first application in the domain of sparse computing. Compared to the traditional 2:4 sparse method, the GQSA offers a more flexible and adjustable sparsity rate, as well as a higher weight compression rate, and is efficiently compatible with weight-only quantization methods. Experimental results demonstrate that, under the GQSA W4S50% compression setting, the model's accuracy surpasses that of both 2:4 pruning and W2 quantization. Furthermore, at the inference level, GQSA outperforms W2 by 1.26times and 2:4 pruning by 2.35times in terms of speed.

This paper introduces PowerInfer-2, a framework designed for high-speed inference of Large Language Models (LLMs) on smartphones, particularly effective for models whose sizes exceed the device's memory capacity. The key insight of PowerInfer-2 is to utilize the heterogeneous computation, memory, and I/O resources in smartphones by decomposing traditional matrix computations into fine-grained neuron cluster computations. Specifically, PowerInfer-2 features a polymorphic neuron engine that adapts computational strategies for various stages of LLM inference. Additionally, it introduces segmented neuron caching and fine-grained neuron-cluster-level pipelining, which effectively minimize and conceal the overhead caused by I/O operations. The implementation and evaluation of PowerInfer-2 demonstrate its capability to support a wide array of LLM models on two smartphones, achieving up to a 29.2x speed increase compared with state-of-the-art frameworks. Notably, PowerInfer-2 is the first system to serve the TurboSparse-Mixtral-47B model with a generation rate of 11.68 tokens per second on a smartphone. For models that fit entirely within the memory, PowerInfer-2 can achieve approximately a 40% reduction in memory usage while maintaining inference speeds comparable to llama.cpp and MLC-LLM. For more details, including a demonstration video, please visit the project site at www.powerinfer.ai/v2.

Large language models (LLMs) based on transformers have made significant strides in recent years, the success of which is driven by scaling up their model size. Despite their high algorithmic performance, the computational and memory requirements of LLMs present unprecedented challenges. To tackle the high compute requirements of LLMs, the Mixture-of-Experts (MoE) architecture was introduced which is able to scale its model size without proportionally scaling up its computational requirements. Unfortunately, MoE's high memory demands and dynamic activation of sparse experts restrict its applicability to real-world problems. Previous solutions that offload MoE's memory-hungry expert parameters to CPU memory fall short because the latency to migrate activated experts from CPU to GPU incurs high performance overhead. Our proposed Pre-gated MoE system effectively tackles the compute and memory challenges of conventional MoE architectures using our algorithm-system co-design. Pre-gated MoE employs our novel pre-gating function which alleviates the dynamic nature of sparse expert activation, allowing our proposed system to address the large memory footprint of MoEs while also achieving high performance. We demonstrate that Pre-gated MoE is able to improve performance, reduce GPU memory consumption, while also maintaining the same level of model quality. These features allow our Pre-gated MoE system to cost-effectively deploy large-scale LLMs using just a single GPU with high performance.

Long-context Multimodal Large Language Models (MLLMs) demand substantial computational resources for inference as the growth of their multimodal Key-Value (KV) cache, in response to increasing input lengths, challenges memory and time efficiency. Unlike single-modality LLMs that manage only textual contexts, the KV cache of long-context MLLMs includes representations from multiple images with temporal and spatial relationships and related textual contexts. The predominance of image tokens means traditional optimizations for LLMs' KV caches are unsuitable for multimodal long-context settings, and no prior works have addressed this challenge. In this work, we introduce LOOK-M, a pioneering, fine-tuning-free approach that efficiently reduces the multimodal KV cache size while maintaining performance comparable to a full cache. We observe that during prompt prefill, the model prioritizes more textual attention over image features, and based on the multimodal interaction observation, a new proposed text-prior method is explored to compress the KV cache. Furthermore, to mitigate the degradation of image contextual information, we propose several compensatory strategies using KV pairs merging. LOOK-M demonstrates that with a significant reduction in KV Cache memory usage, such as reducing it by 80% in some cases, it not only achieves up to 1.5x faster decoding but also maintains or even enhances performance across a variety of long context multimodal tasks.

Large language models face significant computational and memory challenges when processing long contexts. During inference, efficient management of the key-value (KV) cache, which stores intermediate activations for autoregressive generation, is critical to reducing memory overhead and improving computational efficiency. Traditional token-level efficient KV caching methods overlook semantic information, treating tokens independently without considering their semantic relationships. Meanwhile, existing semantic-preserving KV cache management approaches often suffer from substantial memory usage and high time-to-first-token. To address these limitations, we propose SentenceKV, a novel sentence-level semantic KV caching approach designed to enhance inference efficiency while preserving semantic coherence. During prefilling, SentenceKV groups tokens based on sentence-level semantic similarity, compressing sentence representations into concise semantic vectors stored directly on the GPU, while individual KV pairs are offloaded to CPU. During decoding, SentenceKV generates tokens by selectively retrieving semantically relevant sentence-level KV entries, leveraging the semantic similarity between the prefilling-stage semantic vectors and decoding-stage queries. This ensures efficient and contextually accurate predictions, minimizing the loading of redundant or irrelevant data into GPU memory and significantly reducing memory overhead while maintaining stable inference latency, even for extremely long contexts. Extensive evaluations on benchmarks including PG-19, LongBench, and Needle-In-A-Haystack demonstrate that SentenceKV significantly outperforms state-of-the-art methods in both efficiency and memory usage, without compromising model accuracy.

Transformers, driven by attention mechanisms, form the foundation of large language models (LLMs). As these models scale up, efficient GPU attention kernels become essential for high-throughput and low-latency inference. Diverse LLM applications demand flexible and high-performance attention solutions. We present FlashInfer: a customizable and efficient attention engine for LLM serving. FlashInfer tackles KV-cache storage heterogeneity using block-sparse format and composable formats to optimize memory access and reduce redundancy. It also offers a customizable attention template, enabling adaptation to various settings through Just-In-Time (JIT) compilation. Additionally, FlashInfer's load-balanced scheduling algorithm adjusts to dynamism of user requests while maintaining compatibility with CUDAGraph which requires static configuration. FlashInfer have been integrated into leading LLM serving frameworks like SGLang, vLLM and MLC-Engine. Comprehensive kernel-level and end-to-end evaluations demonstrate FlashInfer's ability to significantly boost kernel performance across diverse inference scenarios: compared to state-of-the-art LLM serving solutions, FlashInfer achieve 29-69% inter-token-latency reduction compared to compiler backends for LLM serving benchmark, 28-30% latency reduction for long-context inference, and 13-17% speedup for LLM serving with parallel generation.

Large Language Models (LLMs) have shown remarkable comprehension abilities but face challenges in GPU memory usage during inference, hindering their scalability for real-time applications like chatbots. To accelerate inference, we store computed keys and values (KV cache) in the GPU memory. Existing methods study the KV cache compression to reduce memory by pruning the pre-computed KV cache. However, they neglect the inter-layer dependency between layers and huge memory consumption in pre-computation. To explore these deficiencies, we find that the number of crucial keys and values that influence future generations decreases layer by layer and we can extract them by the consistency in attention weights. Based on the findings, we propose PyramidInfer, a method that compresses the KV cache by layer-wise retaining crucial context. PyramidInfer saves significant memory by computing fewer keys and values without sacrificing performance. Experimental results show PyramidInfer improves 2.2x throughput compared to Accelerate with over 54% GPU memory reduction in KV cache.

Large language model (LLM) inference is computation and memory intensive, so we adapt lexical shortlisting to it hoping to improve both. While lexical shortlisting is well-explored in tasks like machine translation, it requires modifications before being suitable for LLMs as the intended applications vary significantly. Our work studies two heuristics to shortlist sub-vocabulary at LLM inference time: Unicode-based script filtering and corpus-based selection. We explore different LLM families and sizes, and we find that lexical shortlisting can reduce the memory usage of some models by nearly 50\% and has an upper bound of 25\% improvement in generation speed. In this pilot study, we also identify the drawbacks of such vocabulary selection methods and propose avenues for future research.

Large language models (LLMs) have shown remarkable advances in supporting long-context comprehension and processing tasks. However, scaling the generation inference of LLMs to such long contexts incurs significant additional computation load, and demands a substantial GPU memory footprint to maintain the key-value (KV) cache of transformer-based LLMs. Existing KV cache compression methods, such as quantization, face memory bottlenecks as context length increases, while static-sized caches, such as eviction, suffer from inefficient policies. These limitations restrict deployment on consumer-grade devices like a single Nvidia 4090 GPU. To overcome this, we propose Locret, a framework for long-context LLM inference that introduces retaining heads to evaluate the causal importance of KV cache units, allowing for more accurate eviction within a fixed cache size. Locret is fine-tuned on top of the frozen backbone LLM using a minimal amount of data from standard long-context SFT datasets. During inference, we evict low-importance cache units along with a chunked prefill pattern, significantly reducing peak GPU memory usage. We conduct an extensive empirical study to evaluate Locret, where the experimental results show that Locret outperforms the recent competitive approaches, including InfLLM, Quantization, SirLLM, and MInference, in terms of memory efficiency and the quality of generated contents -- Locret achieves over a 20x and 8x KV cache compression ratio compared to the full KV cache for Phi-3-mini-128K and Llama-3.1-8B-instruct. Additionally, Locret can be combined with other methods, such as quantization and token merging. To our knowledge, Locret is the first framework capable of deploying Llama-3.1-8B or similar models on a single Nvidia 4090 GPU, enabling 128K long-context inference without compromising generation quality, and requiring little additional system optimizations.

Large-scale text encoders in text-to-image (T2I) diffusion models have demonstrated exceptional performance in generating high-quality images from textual prompts. Unlike denoising modules that rely on multiple iterative steps, text encoders require only a single forward pass to produce text embeddings. However, despite their minimal contribution to total inference time and floating-point operations (FLOPs), text encoders demand significantly higher memory usage, up to eight times more than denoising modules. To address this inefficiency, we propose Skip and Re-use layers (Skrr), a simple yet effective pruning strategy specifically designed for text encoders in T2I diffusion models. Skrr exploits the inherent redundancy in transformer blocks by selectively skipping or reusing certain layers in a manner tailored for T2I tasks, thereby reducing memory consumption without compromising performance. Extensive experiments demonstrate that Skrr maintains image quality comparable to the original model even under high sparsity levels, outperforming existing blockwise pruning methods. Furthermore, Skrr achieves state-of-the-art memory efficiency while preserving performance across multiple evaluation metrics, including the FID, CLIP, DreamSim, and GenEval scores.

In-context learning (ICL) approaches typically leverage prompting to condition decoder-only language model generation on reference information. Just-in-time processing of a context is inefficient due to the quadratic cost of self-attention operations, and caching is desirable. However, caching transformer states can easily require almost as much space as the model parameters. When the right context isn't known in advance, caching ICL can be challenging. This work addresses these limitations by introducing models that, inspired by the encoder-decoder architecture, use cross-attention to condition generation on reference text without the prompt. More precisely, we leverage pre-trained decoder-only models and only train a small number of added layers. We use Question-Answering (QA) as a testbed to evaluate the ability of our models to perform conditional generation and observe that they outperform ICL, are comparable to fine-tuned prompted LLMs, and drastically reduce the space footprint relative to standard KV caching by two orders of magnitude.

Optimizing inference for long-context Large Language Models (LLMs) is increasingly important due to the quadratic compute and linear memory complexity of Transformers. Existing approximation methods, such as key-value (KV) cache dropping, sparse attention, and prompt compression, typically rely on rough predictions of token or KV pair importance. We propose a novel framework for approximate LLM inference that leverages small draft models to more accurately predict the importance of tokens and KV pairs. Specifically, we introduce two instantiations of our proposed framework: (i) SpecKV, which leverages a draft output to accurately assess the importance of each KV pair for more effective KV cache dropping, and (ii) SpecPC, which uses the draft model's attention activations to identify and discard unimportant prompt tokens. To the best of our knowledge, this is the first work to use draft models for approximate LLM inference acceleration, extending their utility beyond traditional lossless speculative decoding. We motivate our methods with theoretical and empirical analyses, and show a strong correlation between the attention patterns of draft and target models. Extensive experiments on long-context benchmarks show that our methods consistently achieve higher accuracy than existing baselines, while preserving the same improvements in memory usage, latency, and throughput. Our code is available at https://github.com/furiosa-ai/draft-based-approx-llm.

Transformers have emerged as the underpinning architecture for Large Language Models (LLMs). In generative language models, the inference process involves two primary phases: prompt processing and token generation. Token generation, which constitutes the majority of the computational workload, primarily entails vector-matrix multiplications and interactions with the Key-Value (KV) Cache. This phase is constrained by memory bandwidth due to the overhead of transferring weights and KV cache values from the memory system to the computing units. This memory bottleneck becomes particularly pronounced in applications that require long-context and extensive text generation, both of which are increasingly crucial for LLMs. This paper introduces "Keyformer", an innovative inference-time approach, to mitigate the challenges associated with KV cache size and memory bandwidth utilization. Keyformer leverages the observation that approximately 90% of the attention weight in generative inference focuses on a specific subset of tokens, referred to as "key" tokens. Keyformer retains only the key tokens in the KV cache by identifying these crucial tokens using a novel score function. This approach effectively reduces both the KV cache size and memory bandwidth usage without compromising model accuracy. We evaluate Keyformer's performance across three foundational models: GPT-J, Cerebras-GPT, and MPT, which employ various positional embedding algorithms. Our assessment encompasses a variety of tasks, with a particular emphasis on summarization and conversation tasks involving extended contexts. Keyformer's reduction of KV cache reduces inference latency by 2.1x and improves token generation throughput by 2.4x, while preserving the model's accuracy.

While Chain-of-Thought (CoT) prompting improves reasoning in large language models (LLMs), the excessive length of reasoning tokens increases latency and KV cache memory usage, and may even truncate final answers under context limits. We propose ThinkLess, an inference-efficient framework that terminates reasoning generation early and maintains output quality without modifying the model. Atttention analysis reveals that answer tokens focus minimally on earlier reasoning steps and primarily attend to the reasoning terminator token, due to information migration under causal masking. Building on this insight, ThinkLess inserts the terminator token at earlier positions to skip redundant reasoning while preserving the underlying knowledge transfer. To prevent format discruption casued by early termination, ThinkLess employs a lightweight post-regulation mechanism, relying on the model's natural instruction-following ability to produce well-structured answers. Without fine-tuning or auxiliary data, ThinkLess achieves comparable accuracy to full-length CoT decoding while greatly reducing decoding time and memory consumption.

Long text generation, such as novel writing and discourse-level translation with extremely long contexts, presents significant challenges to current language models. Existing methods mainly focus on extending the model's context window through strategies like length extrapolation. However, these approaches demand substantial hardware resources during the training and/or inference phases. Our proposed method, Temp-Lora, introduces an alternative concept. Instead of relying on the KV cache to store all context information, we embeds this information directly into a temporary Lora module. In the process of long text generation, this module is progressively trained with text generated previously. This approach not only efficiently preserves contextual knowledge but also prevents any permanent alteration to the model's parameters given that the module is discarded post-generation. Extensive experiments on the PG19 language modeling benchmark and the GuoFeng discourse-level translation benchmark validate the effectiveness of Temp-Lora. Our results show that: 1) Temp-Lora substantially enhances generation quality for long text, as indicated by a 13.2% decrease in perplexity (PPL) on a subset of PG19, and a 29.3% decrease in PPL along with a 113.2% increase in BLEU score on a subset of GuoFeng, 2) Temp-Lora is compatible with and enhances most existing long text generation methods, and 3) Temp-Lora can greatly reduce computational costs by shortening the context window. For example, we can ensure a moderate improvement in generation quality (a decrease of 3.8% in PPL) while enabling a 51.5% memory usage reduction and a 60.0% decrease in latency for inference.

With the advancements in long-context inference capabilities of large language models (LLMs), the KV cache has become one of the foundational components. However, its substantial GPU memory consumption makes KV cache compression a key technique for enabling efficient LLM inference in industrial scenarios. While recent studies have focused on optimizing the memory occupied by the KV cache, they overlook two critical factors: preserving semantic coherence and considering task-specific characteristic during compression. To address these limitations, we propose a novel task-adaptive KV cache window selection method, WindowKV. WindowKV dynamically selects local semantic windows consisting of consecutive tokens, according to task-specific characteristics, ensuring the retained KV cache captures continuous, essential context. Additionally, we introduce an intra-group layer KV cache indices sharing strategy to reduce computational overhead, achieving a balance between performance and efficiency. We rigorously evaluate WindowKV on the LongBench benchmark, and the results demonstrate that it maintains a performance comparable to full KV cache retention while using only 12% of the original KV cache, significantly reducing memory requirements. Furthermore, our method also achieves state-of-the-art results in the Needle-in-a-Haystack evaluation, highlighting its effectiveness and robustness.

As large language model (LLM) inference demands ever-greater resources, there is a rapid growing trend of using low-bit weights to shrink memory usage and boost inference efficiency. However, these low-bit LLMs introduce the need for mixed-precision matrix multiplication (mpGEMM), which is a crucial yet under-explored operation that involves multiplying lower-precision weights with higher-precision activations. Unfortunately, current hardware does not natively support mpGEMM, resulting in indirect and inefficient dequantization-based implementations. To address the mpGEMM requirements in low-bit LLMs, we explored the lookup table (LUT)-based approach for mpGEMM. However, a conventional LUT implementation falls short of its potential. To fully harness the power of LUT-based mpGEMM, we introduce LUT Tensor Core, a software-hardware co-design optimized for low-bit LLM inference. Specifically, we introduce software-based operator fusion and table symmetrization techniques to optimize table precompute and table storage, respectively. Then, LUT Tensor Core proposes the hardware design featuring an elongated tiling shape design to enhance table reuse and a bit-serial design to support various precision combinations in mpGEMM. Moreover, we design an end-to-end compilation stack with new instructions for LUT-based mpGEMM, enabling efficient LLM compilation and optimizations. The evaluation on low-bit LLMs (e.g., BitNet, LLAMA) shows that LUT Tensor Core achieves more than a magnitude of improvements on both compute density and energy efficiency.

Transformer-based large Language Models (LLMs) become increasingly important in various domains. However, the quadratic time complexity of attention operation poses a significant challenge for scaling to longer contexts due to the extremely high inference latency and GPU memory consumption for caching key-value (KV) vectors. This paper proposes RetrievalAttention, a training-free approach to accelerate attention computation. To leverage the dynamic sparse property of attention, RetrievalAttention builds approximate nearest neighbor search (ANNS) indexes upon KV vectors in CPU memory and retrieves the most relevant ones via vector search during generation. Due to the out-of-distribution (OOD) between query vectors and key vectors, off-the-shelf ANNS indexes still need to scan O(N) (usually 30% of all keys) data for accurate retrieval, which fails to exploit the high sparsity. RetrievalAttention first identifies the OOD challenge of ANNS-based attention, and addresses it via an attention-aware vector search algorithm that can adapt to queries and only access 1--3% of data, thus achieving a sub-linear time complexity. RetrievalAttention greatly reduces the inference cost of long-context LLM with much lower GPU memory requirements while maintaining the model accuracy. Especially, RetrievalAttention only needs 16GB GPU memory for serving 128K tokens in LLMs with 8B parameters, which is capable of generating one token in 0.188 seconds on a single NVIDIA RTX4090 (24GB).

With the fast growth of parameter size, it becomes increasingly challenging to deploy large generative models as they typically require large GPU memory consumption and massive computation. Unstructured model pruning has been a common approach to reduce both GPU memory footprint and the overall computation while retaining good model accuracy. However, the existing solutions do not provide a highly-efficient support for handling unstructured sparsity on modern GPUs, especially on the highly-structured Tensor Core hardware. Therefore, we propose Flash-LLM for enabling low-cost and highly-efficient large generative model inference with the sophisticated support of unstructured sparsity on high-performance but highly restrictive Tensor Cores. Based on our key observation that the main bottleneck of generative model inference is the several skinny matrix multiplications for which Tensor Cores would be significantly under-utilized due to low computational intensity, we propose a general Load-as-Sparse and Compute-as-Dense methodology for unstructured sparse matrix multiplication. The basic insight is to address the significant memory bandwidth bottleneck while tolerating redundant computations that are not critical for end-to-end performance on Tensor Cores. Based on this, we design an effective software framework for Tensor Core based unstructured SpMM, leveraging on-chip resources for efficient sparse data extraction and computation/memory-access overlapping. At SpMM kernel level, Flash-LLM significantly outperforms the state-of-the-art library, i.e., Sputnik and SparTA by an average of 2.9x and 1.5x, respectively. At end-to-end framework level on OPT-30B/66B/175B models, for tokens per GPU-second, Flash-LLM achieves up to 3.8x and 3.6x improvement over DeepSpeed and FasterTransformer, respectively, with significantly lower inference cost.

Many LLM tasks are performed in large batches or even offline, and the performance indictor for which is throughput. These tasks usually show the characteristic of prefix sharing, where different prompt input can partially show the common prefix. However, the existing LLM inference engines tend to optimize the streaming requests and show limitations of supporting the large batched tasks with the prefix sharing characteristic. The existing solutions use the LRU-based cache to reuse the KV context of common prefix. The KV context that is about to be reused may prematurely be evicted with the implicit cache management. Even if not evicted, the lifetime of the shared KV context is extended since requests sharing the same context are not scheduled together, resulting in larger memory usage. These streaming oriented systems schedule the requests in the first-come-first-serve or similar order. As a result, the requests with larger ratio of decoding steps may be scheduled too late to be able to mix with the prefill chunks to increase the hardware utilization. Besides, the token and request number based batching can limit the size of token-batch, which keeps the GPU from saturating for the iterations dominated by decoding tokens. We propose BatchLLM to address the above problems. BatchLLM explicitly identifies the common prefixes globally. The requests sharing the same prefix will be scheduled together to reuse the KV context the best, which also shrinks the lifetime of common KV memory. BatchLLM reorders the requests and schedules the requests with larger ratio of decoding first to better mix the decoding tokens with the latter prefill chunks and applies memory-centric token batching to enlarge the token-batch sizes, which helps to increase the GPU utilization. Extensive evaluation shows that BatchLLM outperforms vLLM by 1.1x to 2x on a set of microbenchmarks and two typical industry workloads.

Large Language Models (LLMs), despite their recent impressive accomplishments, are notably cost-prohibitive to deploy, particularly for applications involving long-content generation, such as dialogue systems and story writing. Often, a large amount of transient state information, referred to as the KV cache, is stored in GPU memory in addition to model parameters, scaling linearly with the sequence length and batch size. In this paper, we introduce a novel approach for implementing the KV cache which significantly reduces its memory footprint. Our approach is based on the noteworthy observation that a small portion of tokens contributes most of the value when computing attention scores. We call these tokens Heavy Hitters (H_2). Through a comprehensive investigation, we find that (i) the emergence of H_2 is natural and strongly correlates with the frequent co-occurrence of tokens in the text, and (ii) removing them results in significant performance degradation. Based on these insights, we propose Heavy Hitter Oracle (H_2O), a KV cache eviction policy that dynamically retains a balance of recent and H_2 tokens. We formulate the KV cache eviction as a dynamic submodular problem and prove (under mild assumptions) a theoretical guarantee for our novel eviction algorithm which could help guide future work. We validate the accuracy of our algorithm with OPT, LLaMA, and GPT-NeoX across a wide range of tasks. Our implementation of H_2O with 20% heavy hitters improves the throughput over three leading inference systems DeepSpeed Zero-Inference, Hugging Face Accelerate, and FlexGen by up to 29times, 29times, and 3times on OPT-6.7B and OPT-30B. With the same batch size, H2O can reduce the latency by up to 1.9times. The code is available at https://github.com/FMInference/H2O.

Recent breakthroughs in solving reasoning, math and coding problems with Large Language Models (LLMs) have been enabled by investing substantial computation budgets at inference time. Therefore, inference speed is one of the most critical properties of LLM architectures, and there is a growing need for LLMs that are efficient and fast at inference. Recently, LLMs built on the xLSTM architecture have emerged as a powerful alternative to Transformers, offering linear compute scaling with sequence length and constant memory usage, both highly desirable properties for efficient inference. However, such xLSTM-based LLMs have yet to be scaled to larger models and assessed and compared with respect to inference speed and efficiency. In this work, we introduce xLSTM 7B, a 7-billion-parameter LLM that combines xLSTM's architectural benefits with targeted optimizations for fast and efficient inference. Our experiments demonstrate that xLSTM 7B achieves performance on downstream tasks comparable to other similar-sized LLMs, while providing significantly faster inference speeds and greater efficiency compared to Llama- and Mamba-based LLMs. These results establish xLSTM 7B as the fastest and most efficient 7B LLM, offering a solution for tasks that require large amounts of test-time computation. Our work highlights xLSTM's potential as a foundational architecture for methods building on heavy use of LLM inference. Our model weights, model code and training code are open-source.

Efficient deployment of Large Language Models (LLMs) requires batching multiple requests together to improve throughput. As the batch size, context length, or model size increases, the size of the key and value (KV) cache can quickly become the main contributor to GPU memory usage and the bottleneck of inference latency. Quantization has emerged as an effective technique for KV cache compression, but existing methods still fail at very low bit widths. We observe that distinct channels of a key/value activation embedding are highly inter-dependent, and the joint entropy of multiple channels grows at a slower rate than the sum of their marginal entropies. Based on this insight, we propose Coupled Quantization (CQ), which couples multiple key/value channels together to exploit their inter-dependency and encode the activations in a more information-efficient manner. Extensive experiments reveal that CQ outperforms or is competitive with existing baselines in preserving model quality. Furthermore, we demonstrate that CQ can preserve model quality with KV cache quantized down to 1-bit.

Key-value (KV) caching has become the de-facto to accelerate generation speed for large language models (LLMs) inference. However, the growing cache demand with increasing sequence length has transformed LLM inference to be a memory bound problem, significantly constraining the system throughput. Existing methods rely on dropping unimportant tokens or quantizing all entries uniformly. Such methods, however, often incur high approximation errors to represent the compressed matrices. The autoregressive decoding process further compounds the error of each step, resulting in critical deviation in model generation and deterioration of performance. To tackle this challenge, we propose GEAR, an efficient KV cache compression framework that achieves near-lossless high-ratio compression. GEAR first applies quantization to majority of entries of similar magnitudes to ultra-low precision. It then employs a low rank matrix to approximate the quantization error, and a sparse matrix to remedy individual errors from outlier entries. By adeptly integrating three techniques, GEAR is able to fully exploit their synergistic potentials. Our experiments demonstrate that compared to alternatives, GEAR achieves near-lossless 4-bit KV cache compression with up to 2.38x throughput improvement, while reducing peak-memory size up to 2.29x. Our code is publicly available at https://github.com/HaoKang-Timmy/GEAR.

Efficient long-context inference is critical as large language models (LLMs) adopt context windows of ranging from 128K to 1M tokens. However, the growing key-value (KV) cache and the high computational complexity of attention create significant bottlenecks in memory usage and latency. In this paper, we find that attention in diverse long-context tasks exhibits sparsity, and LLMs implicitly "know" which tokens can be dropped or evicted at the head level after the pre-filling stage. Based on this insight, we propose Self-Attention Guided Eviction~(SAGE-KV), a simple and effective KV eviction cache method for long-context inference. After prefilling, our method performs a one-time top-k selection at both the token and head levels to compress the KV cache, enabling efficient inference with the reduced cache. Evaluations on LongBench and three long-context LLMs (Llama3.1-8B-Instruct-128k, Llama3-8B-Prolong-512k-Instruct, and Qwen2.5-7B-Instruct-128k) show that SAGE-KV maintains accuracy comparable to full attention while significantly improving efficiency. Specifically, SAGE-KV achieves 4x higher memory efficiency with improved accuracy over the static KV cache selection method StreamLLM, and 2x higher memory efficiency with better accuracy than the dynamic KV cache selection method Quest.

Large Language Models (LLMs) are increasingly used in applications requiring long context lengths, but the key-value (KV) cache often becomes a memory bottleneck on GPUs as context grows. To address this, we propose Commutative Vector Quantization (CommVQ) to significantly reduce memory usage for long-context LLM inference. We first introduce additive quantization with a lightweight encoder and codebook to compress the KV cache, which can be decoded via simple matrix multiplication. To further reduce computational costs during decoding, we design the codebook to be commutative with Rotary Position Embedding (RoPE) and train it using an Expectation-Maximization (EM) algorithm. This enables efficient integration of decoding into the self-attention mechanism. Our approach achieves high accuracy with additive quantization and low overhead via the RoPE-commutative codebook. Experiments on long-context benchmarks and GSM8K show that our method reduces FP16 KV cache size by 87.5% with 2-bit quantization, while outperforming state-of-the-art KV cache quantization methods. Notably, it enables 1-bit KV cache quantization with minimal accuracy loss, allowing a LLaMA-3.1 8B model to run with a 128K context length on a single RTX 4090 GPU. The source code is available at: https://github.com/UMass-Embodied-AGI/CommVQ.

In this study, we introduce adaptive KV cache compression, a plug-and-play method that reduces the memory footprint of generative inference for Large Language Models (LLMs). Different from the conventional KV cache that retains key and value vectors for all context tokens, we conduct targeted profiling to discern the intrinsic structure of attention modules. Based on the recognized structure, we then construct the KV cache in an adaptive manner: evicting long-range contexts on attention heads emphasizing local contexts, discarding non-special tokens on attention heads centered on special tokens, and only employing the standard KV cache for attention heads that broadly attend to all tokens. Moreover, with the lightweight attention profiling used to guide the construction of the adaptive KV cache, FastGen can be deployed without resource-intensive fine-tuning or re-training. In our experiments across various asks, FastGen demonstrates substantial reduction on GPU memory consumption with negligible generation quality loss. We will release our code and the compatible CUDA kernel for reproducibility.

Diffusion models have shown remarkable performance in image synthesis, but they demand extensive computational and memory resources for training, fine-tuning and inference. Although advanced quantization techniques have successfully minimized memory usage for inference, training and fine-tuning these quantized models still require large memory possibly due to dequantization for accurate computation of gradients and/or backpropagation for gradient-based algorithms. However, memory-efficient fine-tuning is particularly desirable for applications such as personalization that often must be run on edge devices like mobile phones with private data. In this work, we address this challenge by quantizing a diffusion model with personalization via Textual Inversion and by leveraging a zeroth-order optimization on personalization tokens without dequantization so that it does not require gradient and activation storage for backpropagation that consumes considerable memory. Since a gradient estimation using zeroth-order optimization is quite noisy for a single or a few images in personalization, we propose to denoise the estimated gradient by projecting it onto a subspace that is constructed with the past history of the tokens, dubbed Subspace Gradient. In addition, we investigated the influence of text embedding in image generation, leading to our proposed time steps sampling, dubbed Partial Uniform Timestep Sampling for sampling with effective diffusion timesteps. Our method achieves comparable performance to prior methods in image and text alignment scores for personalizing Stable Diffusion with only forward passes while reducing training memory demand up to 8.2times.

As large language models increasingly gain popularity in real-world applications, processing extremely long contexts, often exceeding the model's pre-trained context limits, has emerged as a critical challenge. While existing approaches to efficient long-context processing show promise, recurrent compression-based methods struggle with information preservation, whereas random access approaches require substantial memory resources. We introduce REFORM, a novel inference framework that efficiently handles long contexts through a two-phase approach. First, it incrementally processes input chunks while maintaining a compressed KV cache, constructs cross-layer context embeddings, and utilizes early exit strategy for improved efficiency. Second, it identifies and gathers essential tokens via similarity matching and selectively recomputes the KV cache. Compared to baselines, REFORM achieves over 50% and 27% performance gains on RULER and BABILong respectively at 1M context length. It also outperforms baselines on Infinite-Bench and MM-NIAH, demonstrating flexibility across diverse tasks and domains. Additionally, REFORM reduces inference time by 30% and peak memory usage by 5%, achieving both efficiency and superior performance.

Large Language Models (LLMs) face significant deployment challenges due to their substantial resource requirements. While low-bit quantized weights can reduce memory usage and improve inference efficiency, current hardware lacks native support for mixed-precision General Matrix Multiplication (mpGEMM), resulting in inefficient dequantization-based implementations. Moreover, uniform quantization methods often fail to capture weight distributions adequately, leading to performance degradation. We propose GANQ (GPU-Adaptive Non-Uniform Quantization), a layer-wise post-training non-uniform quantization framework optimized for hardware-efficient lookup table-based mpGEMM. GANQ achieves superior quantization performance by utilizing a training-free, GPU-adaptive optimization algorithm to efficiently reduce layer-wise quantization errors. Extensive experiments demonstrate GANQ's ability to reduce the perplexity gap from the FP16 baseline compared to state-of-the-art methods for both 3-bit and 4-bit quantization. Furthermore, when deployed on a single NVIDIA RTX 4090 GPU, GANQ's quantized models achieve up to 2.57times speedup over the baseline, advancing memory and inference efficiency in LLM deployment.

Large Language Models (LLMs) have demonstrated remarkable capabilities in handling long context inputs, but this comes at the cost of increased computational resources and latency. Our research introduces a novel approach for the long context bottleneck to accelerate LLM inference and reduce GPU memory consumption. Our research demonstrates that LLMs can identify relevant tokens in the early layers before generating answers to a query. Leveraging this insight, we propose an algorithm that uses early layers of an LLM as filters to select and compress input tokens, significantly reducing the context length for subsequent processing. Our method, GemFilter, demonstrates substantial improvements in both speed and memory efficiency compared to existing techniques, such as standard attention and SnapKV/H2O. Notably, it achieves a 2.4times speedup and 30\% reduction in GPU memory usage compared to SOTA methods. Evaluation on the Needle in a Haystack task shows that GemFilter significantly outperforms standard attention, SnapKV and demonstrates comparable performance on the LongBench challenge. GemFilter is simple, training-free, and broadly applicable across different LLMs. Crucially, it provides interpretability by allowing humans to inspect the selected input sequence. These findings not only offer practical benefits for LLM deployment, but also enhance our understanding of LLM internal mechanisms, paving the way for further optimizations in LLM design and inference. Our code is available at https://github.com/SalesforceAIResearch/GemFilter.

Efficiently serving large language models (LLMs) requires batching many requests together to reduce the cost per request. Yet, the key-value (KV) cache, which stores attention keys and values to avoid re-computations, significantly increases memory demands and becomes the new bottleneck in speed and memory usage. This memory demand increases with larger batch sizes and longer context lengths. Additionally, the inference speed is limited by the size of KV cache, as the GPU's SRAM must load the entire KV cache from the main GPU memory for each token generated, causing the computational core to be idle during this process. A straightforward and effective solution to reduce KV cache size is quantization, which decreases the total bytes taken by KV cache. However, there is a lack of in-depth studies that explore the element distribution of KV cache to understand the hardness and limitation of KV cache quantization. To fill the gap, we conducted a comprehensive study on the element distribution in KV cache of popular LLMs. Our findings indicate that the key cache should be quantized per-channel, i.e., group elements along the channel dimension and quantize them together. In contrast, the value cache should be quantized per-token. From this analysis, we developed a tuning-free 2bit KV cache quantization algorithm, named KIVI. With the hardware-friendly implementation, KIVI can enable Llama (Llama-2), Falcon, and Mistral models to maintain almost the same quality while using 2.6times less peak memory usage (including the model weight). This reduction in memory usage enables up to 4times larger batch size, bringing 2.35times sim 3.47times throughput on real LLM inference workload. The source code is available at https://github.com/jy-yuan/KIVI.

Large Language models (LLMs) have become a research hotspot. To accelerate the inference of LLMs, storing computed caches in memory has become the standard technique. However, as the inference length increases, growing KV caches might lead to out-of-memory issues. Many existing methods address this issue through KV cache compression, primarily by preserving key tokens throughout all layers to reduce information loss. Most of them allocate a uniform budget size for each layer to retain. However, we observe that the minimum budget sizes needed to retain essential information vary across layers and models based on the perspectives of attention and hidden state output. Building on this observation, this paper proposes a simple yet effective KV cache compression method that leverages layer uncertainty to allocate budget size for each layer. Experimental results show that the proposed method can reduce memory usage of the KV caches to only sim20\% when compared to Full KV inference while achieving nearly lossless performance.

Large model inference is shifting from cloud to edge due to concerns about the privacy of user interaction data. However, edge devices often struggle with limited computing power, memory, and bandwidth, requiring collaboration across multiple devices to run and speed up LLM inference. Pipeline parallelism, the mainstream solution, is inefficient for single-user scenarios, while tensor parallelism struggles with frequent communications. In this paper, we argue that tensor parallelism can be more effective than pipeline on low-resource devices, and present a compute- and memory-efficient tensor parallel inference system, named TPI-LLM, to serve 70B-scale models. TPI-LLM keeps sensitive raw data local in the users' devices and introduces a sliding window memory scheduler to dynamically manage layer weights during inference, with disk I/O latency overlapped with the computation and communication. This allows larger models to run smoothly on memory-limited devices. We analyze the communication bottleneck and find that link latency, not bandwidth, emerges as the main issue, so a star-based allreduce algorithm is implemented. Through extensive experiments on both emulated and real testbeds, TPI-LLM demonstrated over 80% less time-to-first-token and token latency compared to Accelerate, and over 90% compared to Transformers and Galaxy, while cutting the peak memory footprint of Llama 2-70B by 90%, requiring only 3.1 GB of memory for 70B-scale models.

Large pre-trained models (LPMs), such as LLaMA and GLM, have shown exceptional performance across various tasks through fine-tuning. Although low-rank adaption (LoRA) has emerged to cheaply fine-tune these LPMs on downstream tasks, their deployment is still hindered by the vast model scale and computational costs. Neural network pruning offers a way to compress LPMs. However, the current pruning methods designed for LPMs are not compatible with LoRA. This is due to their utilization of unstructured pruning on LPMs, impeding the merging of LoRA weights, or their dependence on the gradients of pre-trained weights to guide pruning, which can impose significant memory overhead. To this end, we propose LoRAPrune, a new framework that delivers an accurate, compact model for efficient inference in a highly memory-effective manner. Specifically, we first design a LoRA-guided pruning criterion, which uses the weights and gradients of LoRA, rather than the gradients of pre-trained weights for importance estimation. We then propose a structured iterative pruning procedure, to remove redundant channels and heads. Extensive experimental results demonstrate the superior performance of our LoRAPrune over existing approaches on the LLaMA series models. For instance, at a 50\% compression rate, LoRAPrune outperforms LLM-Pruner by a perplexity reduction of 8.0 on WikiText2 and 16.05 on PTB datasets, while concurrently reducing memory usage by 52.6\%. The code will be released after review

Current approaches to question answering rely on pre-trained language models (PLMs) like RoBERTa. This work challenges the existing question-answer encoding convention and explores finer representations. We begin with testing various pooling methods compared to using the begin-of-sentence token as a question representation for better quality. Next, we explore opportunities to simultaneously embed all answer candidates with the question. This enables cross-reference between answer choices and improves inference throughput via reduced memory usage. Despite their simplicity and effectiveness, these methods have yet to be widely studied in current frameworks. We experiment with different PLMs, and with and without the integration of knowledge graphs. Results prove that the memory efficacy of the proposed techniques with little sacrifice in performance. Practically, our work enhances 38-100% throughput with 26-65% speedups on consumer-grade GPUs by allowing for considerably larger batch sizes. Our work sends a message to the community with promising directions in both representation quality and efficiency for the question-answering task in natural language processing.

Large language models (LLMs) have demonstrated remarkable performance across various downstream tasks. However, the high computational and memory requirements of LLMs are a major bottleneck. To address this, parameter-efficient fine-tuning (PEFT) methods such as low-rank adaptation (LoRA) have been proposed to reduce computational costs while ensuring minimal loss in performance. Additionally, knowledge distillation (KD) has been a popular choice for obtaining compact student models from teacher models. In this work, we present KD-LoRA, a novel fine-tuning method that combines LoRA with KD. Our results demonstrate that KD-LoRA achieves performance comparable to full fine-tuning (FFT) and LoRA while significantly reducing resource requirements. Specifically, KD-LoRA retains 98% of LoRA's performance on the GLUE benchmark, while being 40% more compact. Additionally, KD-LoRA reduces GPU memory usage by 30% compared to LoRA, while decreasing inference time by 30% compared to both FFT and LoRA. We evaluate KD-LoRA across three encoder-only models: BERT, RoBERTa, and DeBERTaV3. Code is available at https://github.com/rambodazimi/KD-LoRA.

Large Language Models (LLMs) have shown remarkable performance across various tasks, but the escalating demands on computational resources pose significant challenges, particularly in the extensive utilization of full fine-tuning for downstream tasks. To address this, parameter-efficient fine-tuning (PEFT) methods have been developed, but they often underperform compared to full fine-tuning and struggle with memory efficiency. In this work, we introduce Gradient Weight-Normalized Low-Rank Projection (GradNormLoRP), a novel approach that enhances both parameter and memory efficiency while maintaining comparable performance to full fine-tuning. GradNormLoRP normalizes the weight matrix to improve gradient conditioning, facilitating better convergence during optimization. Additionally, it applies low-rank approximations to the weight and gradient matrices, significantly reducing memory usage during training. Extensive experiments demonstrate that our 8-bit GradNormLoRP reduces optimizer memory usage by up to 89.5% and enables the pre-training of large LLMs, such as LLaMA 7B, on consumer-level GPUs like the NVIDIA RTX 4090, without additional inference costs. Moreover, GradNormLoRP outperforms existing low-rank methods in fine-tuning tasks. For instance, when fine-tuning the RoBERTa model on all GLUE tasks with a rank of 8, GradNormLoRP achieves an average score of 80.65, surpassing LoRA's score of 79.23. These results underscore GradNormLoRP as a promising alternative for efficient LLM pre-training and fine-tuning. Source code: https://github.com/Jhhuangkay/Gradient-Weight-normalized-Low-rank-Projection-for-Efficient-LLM-Training

We present RWKV-7 "Goose", a new sequence modeling architecture, along with pre-trained language models that establish a new state-of-the-art in downstream performance at the 3 billion parameter scale on multilingual tasks, and match current SoTA English language performance despite being trained on dramatically fewer tokens than other top 3B models. Nevertheless, RWKV-7 models require only constant memory usage and constant inference time per token. RWKV-7 introduces a newly generalized formulation of the delta rule with vector-valued gating and in-context learning rates, as well as a relaxed value replacement rule. We show that RWKV-7 can perform state tracking and recognize all regular languages, while retaining parallelizability of training. This exceeds the capabilities of Transformers under standard complexity conjectures, which are limited to TC^0. To demonstrate RWKV-7's language modeling capability, we also present an extended open source 3.1 trillion token multilingual corpus, and train four RWKV-7 models ranging from 0.19 billion to 2.9 billion parameters on this dataset. To foster openness, reproduction, and adoption, we release our models and dataset component listing at https://huggingface.co/RWKV, and our training and inference code at https://github.com/RWKV/RWKV-LM all under the Apache 2.0 License.

Large language models (LLMs) have shown remarkable performance in complex reasoning tasks, but their efficiency is hindered by the substantial memory and computational costs associated with generating lengthy tokens. In this paper, we propose LightThinker, a novel method that enables LLMs to dynamically compress intermediate thoughts during reasoning. Inspired by human cognitive processes, LightThinker compresses verbose thought steps into compact representations and discards the original reasoning chains, thereby significantly reducing the number of tokens stored in the context window. This is achieved by training the model on when and how to perform compression through data construction, mapping hidden states to condensed gist tokens, and creating specialized attention masks. Additionally, we introduce the Dependency (Dep) metric to quantify the degree of compression by measuring the reliance on historical tokens during generation. Extensive experiments on four datasets and two models show that LightThinker reduces peak memory usage and inference time, while maintaining competitive accuracy. Our work provides a new direction for improving the efficiency of LLMs in complex reasoning tasks without sacrificing performance. Code will be released at https://github.com/zjunlp/LightThinker.

We propose SLoPe, a Double-Pruned Sparse Plus Lazy Low-rank Adapter Pretraining method for LLMs that improves the accuracy of sparse LLMs while accelerating their pretraining and inference and reducing their memory footprint. Sparse pretraining of LLMs reduces the accuracy of the model, to overcome this, prior work uses dense models during fine-tuning. SLoPe improves the accuracy of sparsely pretrained models by adding low-rank adapters in the final 1% iterations of pretraining without adding significant overheads to the model pretraining and inference. In addition, SLoPe uses a double-pruned backward pass formulation that prunes the transposed weight matrix using N:M sparsity structures to enable an accelerated sparse backward pass. SLoPe accelerates the training and inference of models with billions of parameters up to 1.14times and 1.34times respectively (OPT-33B and OPT-66B) while reducing their memory usage by up to 0.77times and 0.51times for training and inference respectively.

Rapid advancements in large language models (LLMs) have increased interest in deploying them on mobile devices for on-device AI applications. Mobile users interact differently with LLMs compared to desktop users, creating unique expectations and data biases. Current benchmark datasets primarily target at server and desktop environments, and there is a notable lack of extensive datasets specifically designed for mobile contexts. Additionally, mobile devices face strict limitations in storage and computing resources, constraining model size and capabilities, thus requiring optimized efficiency and prioritized knowledge. To address these challenges, we introduce Mobile-MMLU, a large-scale benchmark dataset tailored for mobile intelligence. It consists of 16,186 questions across 80 mobile-related fields, designed to evaluate LLM performance in realistic mobile scenarios. A challenging subset, Mobile-MMLU-Pro, provides advanced evaluation similar in size to MMLU-Pro but significantly more difficult than our standard full set. Both benchmarks use multiple-choice, order-invariant questions focused on practical mobile interactions, such as recipe suggestions, travel planning, and essential daily tasks. The dataset emphasizes critical mobile-specific metrics like inference latency, energy consumption, memory usage, and response quality, offering comprehensive insights into model performance under mobile constraints. Moreover, it prioritizes privacy and adaptability, assessing models' ability to perform on-device processing, maintain user privacy, and adapt to personalized usage patterns. Mobile-MMLU family offers a standardized framework for developing and comparing mobile-optimized LLMs, enabling advancements in productivity and decision-making within mobile computing environments. Our code and data are available at: https://github.com/VILA-Lab/Mobile-MMLU.

Transformers have revolutionized deep learning across various tasks, including audio representation learning, due to their powerful modeling capabilities. However, they often suffer from quadratic complexity in both GPU memory usage and computational inference time, affecting their efficiency. Recently, state space models (SSMs) like Mamba have emerged as a promising alternative, offering a more efficient approach by avoiding these complexities. Given these advantages, we explore the potential of SSM-based models in audio tasks. In this paper, we introduce Self-Supervised Audio Mamba (SSAMBA), the first self-supervised, attention-free, and SSM-based model for audio representation learning. SSAMBA leverages the bidirectional Mamba to capture complex audio patterns effectively. We incorporate a self-supervised pretraining framework that optimizes both discriminative and generative objectives, enabling the model to learn robust audio representations from large-scale, unlabeled datasets. We evaluated SSAMBA on various tasks such as audio classification, keyword spotting, and speaker identification. Our results demonstrate that SSAMBA outperforms the Self-Supervised Audio Spectrogram Transformer (SSAST) in most tasks. Notably, SSAMBA is approximately 92.7% faster in batch inference speed and 95.4% more memory-efficient than SSAST for the tiny model size with an input token size of 22k. These efficiency gains, combined with superior performance, underscore the effectiveness of SSAMBA's architectural innovation, making it a compelling choice for a wide range of audio processing applications.

The mixture of experts (MoE) model is a sparse variant of large language models (LLMs), designed to hold a better balance between intelligent capability and computational overhead. Despite its benefits, MoE is still too expensive to deploy on resource-constrained edge devices, especially with the demands of on-device inference services. Recent research efforts often apply model compression techniques, such as quantization, pruning and merging, to restrict MoE complexity. Unfortunately, due to their predefined static model optimization strategies, they cannot always achieve the desired quality-overhead trade-off when handling multiple requests, finally degrading the on-device quality of service. These limitations motivate us to propose the D^2MoE, an algorithm-system co-design framework that matches diverse task requirements by dynamically allocating the most proper bit-width to each expert. Specifically, inspired by the nested structure of matryoshka dolls, we propose the matryoshka weight quantization (MWQ) to progressively compress expert weights in a bit-nested manner and reduce the required runtime memory. On top of it, we further optimize the I/O-computation pipeline and design a heuristic scheduling algorithm following our hottest-expert-bit-first (HEBF) principle, which maximizes the expert parallelism between I/O and computation queue under constrained memory budgets, thus significantly reducing the idle temporal bubbles waiting for the experts to load. Evaluations on real edge devices show that D^2MoE improves the overall inference throughput by up to 1.39times and reduces the peak memory footprint by up to 53% over the latest on-device inference frameworks, while still preserving comparable serving accuracy as its INT8 counterparts.

Text to speech (TTS) has been broadly used to synthesize natural and intelligible speech in different scenarios. Deploying TTS in various end devices such as mobile phones or embedded devices requires extremely small memory usage and inference latency. While non-autoregressive TTS models such as FastSpeech have achieved significantly faster inference speed than autoregressive models, their model size and inference latency are still large for the deployment in resource constrained devices. In this paper, we propose LightSpeech, which leverages neural architecture search~(NAS) to automatically design more lightweight and efficient models based on FastSpeech. We first profile the components of current FastSpeech model and carefully design a novel search space containing various lightweight and potentially effective architectures. Then NAS is utilized to automatically discover well performing architectures within the search space. Experiments show that the model discovered by our method achieves 15x model compression ratio and 6.5x inference speedup on CPU with on par voice quality. Audio demos are provided at https://speechresearch.github.io/lightspeech.

Recent advancements in speculative decoding have demonstrated considerable speedup across a wide array of large language model (LLM) tasks. Speculative decoding inherently relies on sacrificing extra memory allocations to generate several candidate tokens, of which acceptance rate drives the speedup. However, deploying speculative decoding on memory-constrained devices, such as mobile GPUs, remains as a significant challenge in real-world scenarios. In this work, we present a device-aware inference engine named SpecMemo that can smartly control memory allocations at finer levels to enable multi-turn chatbots with speculative decoding on such limited memory devices. Our methodology stems from theoretically modeling memory footprint of speculative decoding to determine a lower bound on the required memory budget while retaining speedup. SpecMemo empirically acquires a careful balance between minimizing redundant memory allocations for rejected candidate tokens and maintaining competitive performance gains from speculation. Notably, with SpecMemo's memory management, we maintain 96% of overall throughput from speculative decoding on MT-Bench, with reduced generation-memory by 65% on single Nvidia Titan RTX. Given multiple constrained GPUs, we build on top of previous speculative decoding architectures to facilitate big-model inference by distributing Llama-2-70B-Chat model, on which we provide novel batched speculative decoding to increase usability of multiple small server GPUs. This novel framework demonstrates 2x speedup over distributed and batched vanilla decoding with the base model on eight AMD MI250 GPUs. Moreover, inference throughput increases remarkably 8x with batch size 10. Our work contributes to democratized LLM applications in resource-constrained environments, providing a pathway for faster and cheaper deployment of real-world LLM applications with robust performance.

State-of-the-art transformer-based large multimodal models (LMMs) struggle to handle hour-long video inputs due to the quadratic complexity of the causal self-attention operations, leading to high computational costs during training and inference. Existing token compression-based methods reduce the number of video tokens but often incur information loss and remain inefficient for extremely long sequences. In this paper, we explore an orthogonal direction to build a hybrid Mamba-Transformer model (VAMBA) that employs Mamba-2 blocks to encode video tokens with linear complexity. Without any token reduction, VAMBA can encode more than 1024 frames (640times360) on a single GPU, while transformer-based models can only encode 256 frames. On long video input, VAMBA achieves at least 50% reduction in GPU memory usage during training and inference, and nearly doubles the speed per training step compared to transformer-based LMMs. Our experimental results demonstrate that VAMBA improves accuracy by 4.3% on the challenging hour-long video understanding benchmark LVBench over prior efficient video LMMs, and maintains strong performance on a broad spectrum of long and short video understanding tasks.

The field of neural machine translation (NMT) has changed with the advent of large language models (LLMs). Much of the recent emphasis in natural language processing (NLP) has been on modeling machine translation and many other problems using a single pre-trained Transformer decoder, while encoder-decoder architectures, which were the standard in earlier NMT models, have received relatively less attention. In this paper, we explore translation models that are universal, efficient, and easy to optimize, by marrying the world of LLMs with the world of NMT. We apply LLMs to NMT encoding and leave the NMT decoder unchanged. We also develop methods for adapting LLMs to work better with the NMT decoder. Furthermore, we construct a new dataset involving multiple tasks to assess how well the machine translation system generalizes across various tasks. Evaluations on the WMT and our datasets show that results using our method match or surpass a range of baselines in terms of translation quality, but achieve 2.4 sim 6.5 times inference speedups and a 75% reduction in the memory footprint of the KV cache. It also demonstrates strong generalization across a variety of translation-related tasks.

The Large Vision-Language Model (LVLM) integrates computer vision and natural language processing techniques, offering substantial application potential. However, these models demand extensive resources during inference. Adaptive attention techniques can dynamically reduce computational redundancy and thus improve efficiency. Although current adaptive attention methods significantly reduce the memory requirements of Transformer-based language models, they are not tailored for LVLMs. We observe that LVLMs generate responses from both remote image tokens and local text tokens, and different modalities have different attention patterns. This observation inspires us to manage the attention for each modality separately. Specifically, for visual input, we store the cache of potentially useful information but only compute the most critical parts. For language input, we care more about local information. Based on our observation and analysis of vision-language attention patterns, we develop A-VL, a plug-and-play adaptive attention tailored for LVLM inference. Extensive evaluations on three vision-language tasks and five datasets show the effectiveness of our designs. Our approach A-VL outperforms existing adaptive attention methods in reducing memory usage and computational load without compromising performance.

Recent works on neural network pruning advocate that reducing the depth of the network is more effective in reducing run-time memory usage and accelerating inference latency than reducing the width of the network through channel pruning. In this regard, some recent works propose depth compression algorithms that merge convolution layers. However, the existing algorithms have a constricted search space and rely on human-engineered heuristics. In this paper, we propose a novel depth compression algorithm which targets general convolution operations. We propose a subset selection problem that replaces inefficient activation layers with identity functions and optimally merges consecutive convolution operations into shallow equivalent convolution operations for efficient end-to-end inference latency. Since the proposed subset selection problem is NP-hard, we formulate a surrogate optimization problem that can be solved exactly via two-stage dynamic programming within a few seconds. We evaluate our methods and baselines by TensorRT for a fair inference latency comparison. Our method outperforms the baseline method with higher accuracy and faster inference speed in MobileNetV2 on the ImageNet dataset. Specifically, we achieve 1.41times speed-up with 0.11\%p accuracy gain in MobileNetV2-1.0 on the ImageNet.

Diffusion models have been proven highly effective at generating high-quality images. However, as these models grow larger, they require significantly more memory and suffer from higher latency, posing substantial challenges for deployment. In this work, we aim to accelerate diffusion models by quantizing their weights and activations to 4 bits. At such an aggressive level, both weights and activations are highly sensitive, where conventional post-training quantization methods for large language models like smoothing become insufficient. To overcome this limitation, we propose SVDQuant, a new 4-bit quantization paradigm. Different from smoothing which redistributes outliers between weights and activations, our approach absorbs these outliers using a low-rank branch. We first consolidate the outliers by shifting them from activations to weights, then employ a high-precision low-rank branch to take in the weight outliers with Singular Value Decomposition (SVD). This process eases the quantization on both sides. However, na\"{\i}vely running the low-rank branch independently incurs significant overhead due to extra data movement of activations, negating the quantization speedup. To address this, we co-design an inference engine Nunchaku that fuses the kernels of the low-rank branch into those of the low-bit branch to cut off redundant memory access. It can also seamlessly support off-the-shelf low-rank adapters (LoRAs) without the need for re-quantization. Extensive experiments on SDXL, PixArt-Sigma, and FLUX.1 validate the effectiveness of SVDQuant in preserving image quality. We reduce the memory usage for the 12B FLUX.1 models by 3.5times, achieving 3.0times speedup over the 4-bit weight-only quantized baseline on the 16GB laptop 4090 GPU, paving the way for more interactive applications on PCs. Our quantization library and inference engine are open-sourced.

Latest instruction-tuned large language models (LLM) show great results on various tasks, however, they often face performance degradation for non-English input. There is evidence that the reason lies in inefficient tokenization caused by low language representation in pre-training data which hinders the comprehension of non-English instructions, limiting the potential of target language instruction-tuning. In this work we investigate the possibility of addressing the issue with vocabulary substitution in the context of LLaMa Russian language adaptation. We explore three variants of vocabulary adaptation and test their performance on Saiga instruction-tuning and fine-tuning on Russian Super Glue benchmark. The results of automatic evaluation show that vocabulary substitution not only improves the model's quality in Russian but also accelerates fine-tuning (35%) and inference (up to 60%) while reducing memory consumption. Additional human evaluation of the instruction-tuned models demonstrates that models with Russian-adapted vocabulary generate answers with higher user preference than the original Saiga-LLaMa model.

Large language models (LLMs) predominantly employ decoder-only transformer architectures, necessitating the retention of keys/values information for historical tokens to provide contextual information and avoid redundant computation. However, the substantial size and parameter volume of these LLMs require massive GPU memory. This memory demand increases with the length of the input text, leading to an urgent need for more efficient methods of information storage and processing. This study introduces the Anchor-based LLM (AnLLM), which utilizes an innovative anchor-based self-attention network (AnSAN) and also an anchor-based inference strategy. This approach enables LLMs to compress sequence information into an anchor token, reducing the keys/values cache and enhancing inference efficiency. Experiments show that the AnLLM maintains comparable accuracy with up to 99% keys/values cache reduction and up to 3.5 times faster inference. Despite a minor compromise in accuracy, the AnLLM significantly improves computational efficiency and resource utilization, demonstrating the potential of the anchor-based attention approach in the context of LLMs for real-time inference in practical applications.

Transformers have become the cornerstone of modern large-scale language models; however, their dependence on softmax attention poses a major computational bottleneck, particularly in long-context settings. In this work, rather than following prevalent approaches such as linear attention (or SSMs) and local attention, we introduce an intermediate design called \rat between recurrence and attention mechanisms. It partitions the input into chunks, applies a simple linear recurrence within each chunk to capture local dependencies, and then performs softmax attention across chunks to model long-range interactions. By adjusting the size of the chunk, \rat enables flexible trade-offs, combining the strengths of RNN and attention. Empirically, with a chunk size of 16, the \rat layer achieves a \(7\times\) improvement in training speed with 100K token sequences and \(9\times\) in generation at 4K sequence length, while maintaining similar or sometimes even better accuracy compared to standard attention. We demonstrate this by training 1.3B parameter models from scratch and performing large-scale evaluations, including short- and long-context benchmarks, as well as supervised fine-tuning~(SFT). We further propose a hybrid architecture that interleaves \rat with local attention. By combining efficient long-range modeling with strong local interactions, this hybrid design not only improves inference speed and reduces cache memory usage compared to attention, but also consistently enhances performance, for example, achieving an average 1 point gain in commonsense reasoning tasks, up to 4 points on code tasks, and a 1 point Rouge-L increase in a summarization SFT task. Code is available at https://github.com/CLAIRE-Labo/RAT

Parameter-efficient fine-tuning (PEFT) has emerged as the predominant technique for fine-tuning in the era of large language models. However, existing PEFT methods still have inadequate training efficiency. Firstly, the utilization of large-scale foundation models during the training process is excessively redundant for certain fine-tuning tasks. Secondly, as the model size increases, the growth in trainable parameters of empirically added PEFT modules becomes non-negligible and redundant, leading to inefficiency. To achieve task-specific efficient fine-tuning, we propose the Light-PEFT framework, which includes two methods: Masked Early Pruning of the Foundation Model and Multi-Granularity Early Pruning of PEFT. The Light-PEFT framework allows for the simultaneous estimation of redundant parameters in both the foundation model and PEFT modules during the early stage of training. These parameters can then be pruned for more efficient fine-tuning. We validate our approach on GLUE, SuperGLUE, QA tasks, and various models. With Light-PEFT, parameters of the foundation model can be pruned by up to over 40%, while still controlling trainable parameters to be only 25% of the original PEFT method. Compared to utilizing the PEFT method directly, Light-PEFT achieves training and inference speedup, reduces memory usage, and maintains comparable performance and the plug-and-play feature of PEFT.

Test-Time Scaling (TTS) methods for enhancing Large Language Model (LLM) reasoning often incur substantial computational costs, primarily due to extensive reliance on external Process Reward Models (PRMs) or sampling methods like Best-of-N (BoN). This paper introduces Guided by Gut (GG), an efficient self-guided TTS framework that achieves PRM-level performance without costly external verifier models. Our method employs a lightweight tree search guided solely by intrinsic LLM signals, token-level confidence and step novelty. One critical innovation is improving the reliability of internal confidence estimates via a targeted reinforcement learning fine-tuning phase. Empirical evaluations on challenging mathematical reasoning benchmarks demonstrate that GG enables smaller models (e.g., 1.5B parameters) to achieve accuracy matching or surpassing significantly larger models (e.g., 32B-70B parameters), while reducing GPU memory usage by up to 10x. Compared to PRM-based methods, GG achieves comparable accuracy with 8x faster inference speeds and 4-5x lower memory usage. Additionally, GG reduces KV cache memory usage by approximately 50% compared to the BoN strategy, facilitating more efficient and practical deployment of TTS techniques.

Network traffic classification is a crucial research area aiming to enhance service quality, streamline network management, and bolster cybersecurity. To address the growing complexity of transmission encryption techniques, various machine learning and deep learning methods have been proposed. However, existing approaches face two main challenges. Firstly, they struggle with model inefficiency due to the quadratic complexity of the widely used Transformer architecture. Secondly, they suffer from inadequate traffic representation because of discarding important byte information while retaining unwanted biases. To address these challenges, we propose NetMamba, an efficient linear-time state space model equipped with a comprehensive traffic representation scheme. We adopt a specially selected and improved unidirectional Mamba architecture for the networking field, instead of the Transformer, to address efficiency issues. In addition, we design a traffic representation scheme to extract valid information from massive traffic data while removing biased information. Evaluation experiments on six public datasets encompassing three main classification tasks showcase NetMamba's superior classification performance compared to state-of-the-art baselines. It achieves an accuracy rate of nearly 99% (some over 99%) in all tasks. Additionally, NetMamba demonstrates excellent efficiency, improving inference speed by up to 60 times while maintaining comparably low memory usage. Furthermore, NetMamba exhibits superior few-shot learning abilities, achieving better classification performance with fewer labeled data. To the best of our knowledge, NetMamba is the first model to tailor the Mamba architecture for networking.

Significant advancements have been achieved in the realm of large-scale pre-trained text-to-video Diffusion Models (VDMs). However, previous methods either rely solely on pixel-based VDMs, which come with high computational costs, or on latent-based VDMs, which often struggle with precise text-video alignment. In this paper, we are the first to propose a hybrid model, dubbed as Show-1, which marries pixel-based and latent-based VDMs for text-to-video generation. Our model first uses pixel-based VDMs to produce a low-resolution video of strong text-video correlation. After that, we propose a novel expert translation method that employs the latent-based VDMs to further upsample the low-resolution video to high resolution. Compared to latent VDMs, Show-1 can produce high-quality videos of precise text-video alignment; Compared to pixel VDMs, Show-1 is much more efficient (GPU memory usage during inference is 15G vs 72G). We also validate our model on standard video generation benchmarks. Our code and model weights are publicly available at https://github.com/showlab/Show-1.

In Transformer architectures, tokens\textemdash discrete units derived from raw data\textemdash are formed by segmenting inputs into fixed-length chunks. Each token is then mapped to an embedding, enabling parallel attention computations while preserving the input's essential information. Due to the quadratic computational complexity of transformer self-attention mechanisms, token reduction has primarily been used as an efficiency strategy. This is especially true in single vision and language domains, where it helps balance computational costs, memory usage, and inference latency. Despite these advances, this paper argues that token reduction should transcend its traditional efficiency-oriented role in the era of large generative models. Instead, we position it as a fundamental principle in generative modeling, critically influencing both model architecture and broader applications. Specifically, we contend that across vision, language, and multimodal systems, token reduction can: (i) facilitate deeper multimodal integration and alignment, (ii) mitigate "overthinking" and hallucinations, (iii) maintain coherence over long inputs, and (iv) enhance training stability, etc. We reframe token reduction as more than an efficiency measure. By doing so, we outline promising future directions, including algorithm design, reinforcement learning-guided token reduction, token optimization for in-context learning, and broader ML and scientific domains. We highlight its potential to drive new model architectures and learning strategies that improve robustness, increase interpretability, and better align with the objectives of generative modeling.

The deployment of large language models (LLMs) is often constrained by memory bandwidth, where the primary bottleneck is the cost of transferring model parameters from the GPU's global memory to its registers. When coupled with custom kernels that fuse the dequantization and matmul operations, weight-only quantization can thus enable faster inference by reducing the amount of memory movement. However, developing high-performance kernels for weight-quantized LLMs presents substantial challenges, especially when the weights are compressed to non-evenly-divisible bit widths (e.g., 3 bits) with non-uniform, lookup table (LUT) quantization. This paper describes FLUTE, a flexible lookup table engine for LUT-quantized LLMs, which uses offline restructuring of the quantized weight matrix to minimize bit manipulations associated with unpacking, and vectorization and duplication of the lookup table to mitigate shared memory bandwidth constraints. At batch sizes < 32 and quantization group size of 128 (typical in LLM inference), the FLUTE kernel can be 2-4x faster than existing GEMM kernels. As an application of FLUTE, we explore a simple extension to lookup table-based NormalFloat quantization and apply it to quantize LLaMA3 to various configurations, obtaining competitive quantization performance against strong baselines while obtaining an end-to-end throughput increase of 1.5 to 2 times.

Large Multimodal Models (LMMs) have demonstrated impressive performance in short video understanding tasks but face great challenges when applied to long video understanding. In contrast, Large Language Models (LLMs) exhibit outstanding capabilities in modeling long texts. Existing work attempts to address this issue by introducing long video-text pairs during training. However, these approaches require substantial computational and data resources. In this paper, we tackle the challenge of long video understanding from the perspective of context windows, aiming to apply LMMs to long video tasks without retraining on long video datasets. We first conduct an in-depth analysis of why pretrained LMMs struggle to understand lengthy video content, identifying that discrepancies between visual and language modalities lead to different context windows for visual and language tokens, making it difficult to directly extend the visual tokens to match the language context window. Based on this, we propose to adapt LMMs for long video understanding tasks by extending the visual context window, eliminating the need for retraining on large scalelong video datasets. To further mitigate the significant memory consumption caused by long sequences, we introduce a progressive pooling inference strategy that selectively adjusts the spatial resolution of frame embeddings, reducing the number of visual tokens while retaining important spatial information. Across multiple long video understanding benchmarks, our method consistently improves the performance as the number of video frames increases. On the MLVU benchmark, our method outperforms GPT-4o, even though our model size is only 7B. Additionally, in the 256-frame setting, our method reduces memory usage by approximately 45% compared to the baseline, without introducing any performance loss.

We introduce a decoder-decoder architecture, YOCO, for large language models, which only caches key-value pairs once. It consists of two components, i.e., a cross-decoder stacked upon a self-decoder. The self-decoder efficiently encodes global key-value (KV) caches that are reused by the cross-decoder via cross-attention. The overall model behaves like a decoder-only Transformer, although YOCO only caches once. The design substantially reduces GPU memory demands, yet retains global attention capability. Additionally, the computation flow enables prefilling to early exit without changing the final output, thereby significantly speeding up the prefill stage. Experimental results demonstrate that YOCO achieves favorable performance compared to Transformer in various settings of scaling up model size and number of training tokens. We also extend YOCO to 1M context length with near-perfect needle retrieval accuracy. The profiling results show that YOCO improves inference memory, prefill latency, and throughput by orders of magnitude across context lengths and model sizes. Code is available at https://aka.ms/YOCO.

Video large language models (VLLMs) have significantly advanced recently in processing complex video content, yet their inference efficiency remains constrained because of the high computational cost stemming from the thousands of visual tokens generated from the video inputs. We empirically observe that, unlike single image inputs, VLLMs typically attend visual tokens from different frames at different decoding iterations, making a one-shot pruning strategy prone to removing important tokens by mistake. Motivated by this, we present DyCoke, a training-free token compression method to optimize token representation and accelerate VLLMs. DyCoke incorporates a plug-and-play temporal compression module to minimize temporal redundancy by merging redundant tokens across frames, and applies dynamic KV cache reduction to prune spatially redundant tokens selectively. It ensures high-quality inference by dynamically retaining the critical tokens at each decoding step. Extensive experimental results demonstrate that DyCoke can outperform the prior SoTA counterparts, achieving 1.5X inference speedup, 1.4X memory reduction against the baseline VLLM, while still improving the performance, with no training.

A fundamental problem in learning robust and expressive visual representations lies in efficiently estimating the spatial relationships of visual semantics throughout the entire image. In this study, we propose vHeat, a novel vision backbone model that simultaneously achieves both high computational efficiency and global receptive field. The essential idea, inspired by the physical principle of heat conduction, is to conceptualize image patches as heat sources and model the calculation of their correlations as the diffusion of thermal energy. This mechanism is incorporated into deep models through the newly proposed module, the Heat Conduction Operator (HCO), which is physically plausible and can be efficiently implemented using DCT and IDCT operations with a complexity of O(N^{1.5}). Extensive experiments demonstrate that vHeat surpasses Vision Transformers (ViTs) across various vision tasks, while also providing higher inference speeds, reduced FLOPs, and lower GPU memory usage for high-resolution images. The code will be released at https://github.com/MzeroMiko/vHeat.

Despite the impressive advancements of Large Vision-Language Models (LVLMs), existing approaches suffer from a fundamental bottleneck: inefficient visual-language integration. Current methods either disrupt the model's inherent structure or introduce severe long-context computational burden, severely limiting scalability and efficiency. In this paper, we rethink multimodal integration and present LaVi, a novel LVLM that enables seamless and efficient vision-language fusion through internal feature modulation within the Large Language Models (LLMs). Unlike dominant LVLMs that rely on visual token concatenation, LaVi bypasses long-context expansion by introducing a lightweight and adaptive transformation, which incorporates visual context by injecting token-wise vision-conditioned deltas into the affine parameters of layer normalization. This mechanism directly modulates linguistic hidden states based on visual input, ensuring precise vision-language alignment while preserving the LLM's linguistic priors and drastically reducing computational costs. Extensive evaluations across 15 image and video benchmarks demonstrate that LaVi not only achieves state-of-the-art multimodal performance but also dramatically enhances efficiency. Compared to LLaVA-OV-7B, LaVi reduces FLOPs by 94.0%, improves inference speed by 3.1 times, and cuts memory usage in half - establishing LaVi as a scalable and practical solution for real-time multimodal reasoning. The code and models will be released soon.

We propose E-MD3C (Efficient Masked Diffusion Transformer with Disentangled Conditions and Compact Collector), a highly efficient framework for zero-shot object image customization. Unlike prior works reliant on resource-intensive Unet architectures, our approach employs lightweight masked diffusion transformers operating on latent patches, offering significantly improved computational efficiency. The framework integrates three core components: (1) an efficient masked diffusion transformer for processing autoencoder latents, (2) a disentangled condition design that ensures compactness while preserving background alignment and fine details, and (3) a learnable Conditions Collector that consolidates multiple inputs into a compact representation for efficient denoising and learning. E-MD3C outperforms the existing approach on the VITON-HD dataset across metrics such as PSNR, FID, SSIM, and LPIPS, demonstrating clear advantages in parameters, memory efficiency, and inference speed. With only 1{4} of the parameters, our Transformer-based 468M model delivers 2.5times faster inference and uses 2{3} of the GPU memory compared to an 1720M Unet-based latent diffusion model.

The emergence of long-context text applications utilizing large language models (LLMs) has presented significant scalability challenges, particularly in memory footprint. The linear growth of the Key-Value (KV) cache responsible for storing attention keys and values to minimize redundant computations can lead to substantial increases in memory consumption, potentially causing models to fail to serve with limited memory resources. To address this issue, we propose a novel approach called Cache Sparse Representation (CSR), which converts the KV cache by transforming the dense Key-Value cache tensor into sparse indexes and weights, offering a more memory-efficient representation during LLM inference. Furthermore, we introduce NeuralDict, a novel neural network-based method for automatically generating the dictionary used in our sparse representation. Our extensive experiments demonstrate that CSR achieves performance comparable to state-of-the-art KV cache quantization algorithms while maintaining robust functionality in memory-constrained environments.

Large-scale multi-modal deep learning models have revolutionized domains such as healthcare, highlighting the importance of computational power. However, in resource-constrained regions like Low and Middle-Income Countries (LMICs), limited access to GPUs and data poses significant challenges, often leaving CPUs as the sole resource. To address this, we advocate for leveraging vector embeddings to enable flexible and efficient computational methodologies, democratizing multimodal deep learning across diverse contexts. Our paper investigates the efficiency and effectiveness of using vector embeddings from single-modal foundation models and multi-modal Vision-Language Models (VLMs) for multimodal deep learning in low-resource environments, particularly in healthcare. Additionally, we propose a simple yet effective inference-time method to enhance performance by aligning image-text embeddings. Comparing these approaches with traditional methods, we assess their impact on computational efficiency and model performance using metrics like accuracy, F1-score, inference time, training time, and memory usage across three medical modalities: BRSET (ophthalmology), HAM10000 (dermatology), and SatelliteBench (public health). Our findings show that embeddings reduce computational demands without compromising model performance. Furthermore, our alignment method improves performance in medical tasks. This research promotes sustainable AI practices by optimizing resources in constrained environments, highlighting the potential of embedding-based approaches for efficient multimodal learning. Vector embeddings democratize multimodal deep learning in LMICs, particularly in healthcare, enhancing AI adaptability in varied use cases.

While Transformer self-attention offers strong parallelism, the Key-Value (KV) cache grows linearly with sequence length and becomes a bottleneck for inference efficiency. Multi-head latent attention was recently developed to compress the KV cache into a low-rank latent space. This paper proposes Multi-head Temporal Latent Attention (MTLA), which further reduces the KV cache size along the temporal dimension, greatly lowering the memory footprint of self-attention inference. MTLA employs a hyper-network to dynamically merge temporally adjacent KV cache vectors. To address the mismatch between the compressed KV cache and processed sequence lengths, a stride-aware causal mask is proposed to ensure efficient parallel training and consistency with inference behaviour. Experiments across tasks, including speech translation, speech recognition, speech understanding and text summarisation, demonstrate that MTLA achieves competitive performance compared to standard Multi-Head Attention (MHA), while greatly improving inference speed and GPU memory usage. For example, on a English-German speech translation task, MTLA achieves a 5.3x speedup and a reduction in GPU memory usage by a factor of 8.3 compared to MHA, while maintaining translation quality.

To alleviate the data scarcity problem in End-to-end speech translation (ST), pre-training on data for speech recognition and machine translation is considered as an important technique. However, the modality gap between speech and text prevents the ST model from efficiently inheriting knowledge from the pre-trained models. In this work, we propose AdaTranS for end-to-end ST. It adapts the speech features with a new shrinking mechanism to mitigate the length mismatch between speech and text features by predicting word boundaries. Experiments on the MUST-C dataset demonstrate that AdaTranS achieves better performance than the other shrinking-based methods, with higher inference speed and lower memory usage. Further experiments also show that AdaTranS can be equipped with additional alignment losses to further improve performance.

Large Reasoning Models (LRMs) significantly improve the reasoning ability of Large Language Models (LLMs) by learning to reason, exhibiting promising performance in complex task-solving. However, their deliberative reasoning process leads to inefficiencies in token usage, memory consumption, and inference time. Thus, this survey provides a review of efficient inference methods designed specifically for LRMs, focusing on mitigating token inefficiency while preserving the reasoning quality. First, we introduce a taxonomy to group the recent methods into two main categories: (a) explicit compact Chain-of-Thought (CoT), which reduces tokens while keeping the explicit reasoning structure, and (b) implicit latent CoT, which encodes reasoning steps within hidden representations instead of explicit tokens. Meanwhile, we discuss their strengths and weaknesses. Then, we conduct empirical analyses on existing methods from performance and efficiency aspects. Besides, we present open challenges in this field, including human-centric controllable reasoning, trade-off between interpretability and efficiency of reasoning, ensuring safety of efficient reasoning, and broader applications of efficient reasoning. In addition, we highlight key insights for enhancing LRMs' inference efficiency via techniques such as model merging, new architectures, and agent routers. We hope this work serves as a valuable guide, helping researchers overcome challenges in this vibrant fieldhttps://github.com/yueliu1999/Awesome-Efficient-Inference-for-LRMs.

Large Language Models (LLMs) have long held sway in the realms of artificial intelligence research. Numerous efficient techniques, including weight pruning, quantization, and distillation, have been embraced to compress LLMs, targeting memory reduction and inference acceleration, which underscore the redundancy in LLMs. However, most model compression techniques concentrate on weight optimization, overlooking the exploration of optimal architectures. Besides, traditional architecture search methods, limited by the elevated complexity with extensive parameters, struggle to demonstrate their effectiveness on LLMs. In this paper, we propose a training-free architecture search framework to identify optimal subnets that preserve the fundamental strengths of the original LLMs while achieving inference acceleration. Furthermore, after generating subnets that inherit specific weights from the original LLMs, we introduce a reformation algorithm that utilizes the omitted weights to rectify the inherited weights with a small amount of calibration data. Compared with SOTA training-free structured pruning works that can generate smaller networks, our method demonstrates superior performance across standard benchmarks. Furthermore, our generated subnets can directly reduce the usage of GPU memory and achieve inference acceleration.

Attention mechanisms, particularly softmax attention, have been instrumental in the success of transformer-based models such as GPT. However, the quadratic memory complexity of softmax attention with respect to sequence length poses significant challenges for processing longer sequences. We introduce Cottention, a novel attention mechanism that replaces the softmax operation with cosine similarity. By leveraging the properties of cosine similarity and rearranging the attention equation, Cottention achieves native linear memory complexity with respect to sequence length, making it inherently more memory-efficient than softmax attention. We demonstrate that Cottention can be reformulated as a recurrent neural network (RNN) with a finite hidden state, allowing for constant memory usage during inference. We evaluate Cottention on both the bidirectional BERT and causal GPT tasks, demonstrating comparable performance to softmax attention while significantly reducing memory requirements. To ensure efficient computation, we develop a custom CUDA kernel for Cottention. Our results show that Cottention is a promising alternative to softmax attention, enabling the processing of longer sequences without sacrificing performance, due to its native linear memory complexity and ability to maintain a constant memory footprint during inference.

As recently demonstrated, Deep Neural Networks (DNN), usually trained using single precision IEEE 754 floating point numbers (binary32), can also work using lower precision. Therefore, 16-bit and 8-bit compressed format have attracted considerable attention. In this paper, we focused on two families of formats that have already achieved interesting results in compressing binary32 numbers in machine learning applications, without sensible degradation of the accuracy: bfloat and posit. Even if 16-bit and 8-bit bfloat/posit are routinely used for reducing the storage of the weights/biases of trained DNNs, the inference still often happens on the 32-bit FPU of the CPU (especially if GPUs are not available). In this paper we propose a way to decompress a tensor of bfloat/posits just before computations, i.e., after the compressed operands have been loaded within the vector registers of a vector capable CPU, in order to save bandwidth usage and increase cache efficiency. Finally, we show the architectural parameters and considerations under which this solution is advantageous with respect to the uncompressed one.

As large language models (LLMs) process increasing context windows, the memory usage of KV cache has become a critical bottleneck during inference. The mainstream KV compression methods, including KV pruning and KV quantization, primarily focus on either token or precision dimension and seldom explore the efficiency of their combination. In this paper, we comprehensively investigate the token-precision trade-off in KV cache compression. Experiments demonstrate that storing more tokens in the KV cache with lower precision, i.e., quantized pruning, can significantly enhance the long-context performance of LLMs. Furthermore, in-depth analysis regarding token-precision trade-off from a series of key aspects exhibit that, quantized pruning achieves substantial improvements in retrieval-related tasks and consistently performs well across varying input lengths. Moreover, quantized pruning demonstrates notable stability across different KV pruning methods, quantization strategies, and model scales. These findings provide valuable insights into the token-precision trade-off in KV cache compression. We plan to release our code in the near future.

In this paper, we introduce Positional Skip-wisE (PoSE) training for efficient adaptation of large language models~(LLMs) to extremely long context windows. PoSE decouples train length from target context window size by simulating long inputs using a fixed context window with manipulated position indices during training. Concretely, we select several short chunks from a long input sequence, and introduce distinct skipping bias terms to modify the position indices of each chunk. These bias terms, along with the length of each chunk, are altered for each training example, allowing the model to adapt to all positions within the target context window without training on full length inputs. Experiments show that, compared with fine-tuning on the full length, PoSE greatly reduces memory and time overhead with minimal impact on performance. Leveraging this advantage, we have successfully extended the LLaMA model to 128k tokens. Furthermore, we empirically confirm that PoSE is compatible with all RoPE-based LLMs and various position interpolation strategies. Notably, by decoupling fine-tuning length from target context window, PoSE can theoretically extend the context window infinitely, constrained only by memory usage for inference. With ongoing advancements for efficient inference, we believe PoSE holds great promise for scaling the context window even further.

Large language models (LLMs) represent a groundbreaking advancement in the domain of natural language processing due to their impressive reasoning abilities. Recently, there has been considerable interest in increasing the context lengths for these models to enhance their applicability to complex tasks. However, at long context lengths and large batch sizes, the key-value (KV) cache, which stores the attention keys and values, emerges as the new bottleneck in memory usage during inference. To address this, we propose Eigen Attention, which performs the attention operation in a low-rank space, thereby reducing the KV cache memory overhead. Our proposed approach is orthogonal to existing KV cache compression techniques and can be used synergistically with them. Through extensive experiments over OPT, MPT, and Llama model families, we demonstrate that Eigen Attention results in up to 40% reduction in KV cache sizes and up to 60% reduction in attention operation latency with minimal drop in performance.

Large language models (LLMs) achieve remarkable performance in natural language understanding but require substantial computation and memory resources. Post-training quantization (PTQ) is a powerful compression technique extensively investigated in LLMs. However, existing PTQ methods are still not ideal in terms of accuracy and efficiency, especially with below 4 bit-widths. Standard PTQ methods using group-wise quantization suffer difficulties in quantizing LLMs accurately to such low-bit, but advanced methods remaining high-precision weights element-wisely are hard to realize their theoretical hardware efficiency. This paper presents a Salience-Driven Mixed-Precision Quantization scheme for LLMs, namely SliM-LLM. The scheme exploits the salience distribution of weights to determine optimal bit-width and quantizers for accurate LLM quantization, while aligning bit-width partition to groups for compact memory usage and fast integer inference. Specifically, the proposed SliM-LLM mainly relies on two novel techniques: (1) Salience-Determined Bit Allocation utilizes the clustering characteristics of salience distribution to allocate the bit-widths of each group, increasing the accuracy of quantized LLMs and maintaining the inference efficiency; (2) Salience-Weighted Quantizer Calibration optimizes the parameters of the quantizer by considering the element-wise salience within the group, balancing the maintenance of salient information and minimization of errors. Comprehensive experiments show that SliM-LLM significantly improves the accuracy of LLMs at ultra-low bits, e.g., 2-bit LLaMA-7B achieves a 5.5-times memory-saving than original model on NVIDIA A800 GPUs, and 48% decrease of perplexity compared to the state-of-the-art gradient-free PTQ method. Moreover, SliM-LLM+, which is integrated from the extension of SliM-LLM with gradient-based quantizers, further reduces perplexity by 35.1%.

Lowering the memory requirement in full-parameter training on large models has become a hot research area. MeZO fine-tunes the large language models (LLMs) by just forward passes in a zeroth-order SGD optimizer (ZO-SGD), demonstrating excellent performance with the same GPU memory usage as inference. However, the simulated perturbation stochastic approximation for gradient estimate in MeZO leads to severe oscillations and incurs a substantial time overhead. Moreover, without momentum regularization, MeZO shows severe over-fitting problems. Lastly, the perturbation-irrelevant momentum on ZO-SGD does not improve the convergence rate. This study proposes ZO-AdaMU to resolve the above problems by adapting the simulated perturbation with momentum in its stochastic approximation. Unlike existing adaptive momentum methods, we relocate momentum on simulated perturbation in stochastic gradient approximation. Our convergence analysis and experiments prove this is a better way to improve convergence stability and rate in ZO-SGD. Extensive experiments demonstrate that ZO-AdaMU yields better generalization for LLMs fine-tuning across various NLP tasks than MeZO and its momentum variants.

Existing adaptation techniques typically require architectural modifications or added parameters, leading to high computational costs and complexity. We introduce Attention Projection Layer Adaptation (APLA), a simple approach to adapt vision transformers (ViTs) without altering the architecture or adding parameters. Through a systematic analysis, we find that the layer immediately after the attention mechanism is crucial for adaptation. By updating only this projection layer, or even just a random subset of this layer's weights, APLA achieves state-of-the-art performance while reducing GPU memory usage by up to 52.63% and training time by up to 43.0%, with no extra cost at inference. Across 46 datasets covering a variety of tasks including scene classification, medical imaging, satellite imaging, and fine-grained classification, APLA consistently outperforms 17 other leading adaptation methods, including full fine-tuning, on classification, segmentation, and detection tasks. The code is available at https://github.com/MoeinSorkhei/APLA.