Neural Information Processing Systems http://nips.cc/

Modern neural networks (NN) contain an ever-growing number of parameters, substantially increasing the memory and computational cost of inference. Researchers have explored various ways to reduce the inference cost of NNs by reducing the model size before deployment and dynamically pruning the inference computation at runtime. In this work, we present ExPrune, a general, dynamic pruning optimization that enables multi-granularity partial computation on a per-input basis. ExPrune requires no change to the model architecture or the training algorithm. ExPrune is based on our theoretical results that the relationship between certain model parameters and intermediate values can be described by a statistical property called exchangeability. By identifying exchangeable parameters and values in the model, we are able to first partially evaluate the network, analyze the statistics of the partial results, and make pruning decisions on the fly. Because ExPrune is theory grounded, it generalizes across model architectures in different problem domains. We evaluate ExPrune on one computer vision models, one graph model and one language model. ExPrune provides 10.98--17.33% reduction in FLOPs with negligible accuracy drop and 21.61--27.16% reduction in FLOPs with at most 1% accuracy drop. We also demonstrate that ExPrune composes with static magnitude pruning. On models that have been aggressively statically pruned, ExPrune still provides additional 10.24--11.11% reduction in FLOPs with negligible accuracy drop and 13.91--14.39% reduction in FLOPs with at most 1% accuracy drop.

These benchmarks contain only short audio clips and thus do not reflect the complexity of achieving long-context Speech Large Language Models (Speech LLMs) represent a understanding and extracting precise information from lengthy significant advancement in speech language understanding and audio sequences. To systematically assess the unique challenges processing, as they leverage contextual reasoning capabilities of posed by SIR, we present SPIRAL (Speech Informational large language models to process audio inputs [1]. Unlike traditional Retrieval and Lookup), a 1,012-sample benchmark specifically cascaded pipelines, where automatic speech recognition crafted to evaluate Speech LLM performance on long-form (ASR) and language modeling are handled by separate modules, audio sequences (around 90 seconds in duration). On a high Speech LLMs unify audio processing, cross-modal fusion, and level, SPIRAL constructs SIR questions by embedding a critical language modeling in a single architecture [2]. These unified piece of information within lengthy and potentially distracting models can perform multiple tasks like speech recognition, dialogues, thereby assessing the model ability to pinpoint and speech translation, speaker identification and emotion recognition, retrieve essential content from long-form inputs.

With the proliferation of domain-specific models, model merging has emerged as a set of techniques that combine the capabilities of multiple models into one that can multitask without the cost of additional training. In this paper, we propose a new model merging technique, Drop and rEscaLe via sampLing with mAgnitude (DELLA-Merging), that employs a novel pruning technique, MAGPRUNE, which shows significant advantages over DARE and TIES. MAGPRUNE first ranks the parameters in order of their magnitude and assigns higher dropout probabilities (p) to parameters with lower ranks corresponding to lower magnitudes. To approximate the original embeddings, MAGPRUNE employs a rescaling operation on the parameters that survive the random dropping by 1/(1 - p). On three different expert models considered for merging (LM, Math, Code) and corresponding benchmark datasets (AlpacaEval, GSM8K, MBPP), DELLA shows an average improvement of 2.4 points over baseline methods employing delta parameter pruning (an improvement of 3.6 points over TIES, 1.2 points over DARE), and 11.1 points over the no-pruning baseline (TA). We release the source code at: https://github.com/declare-lab/della.

Transformer-based Language Models have become ubiquitous in Natural Language Processing (NLP) due to their impressive performance on various tasks. However, expensive training as well as inference remains a significant impediment to their widespread applicability. While enforcing sparsity at various levels of the model architecture has found promise in addressing scaling and efficiency issues, there remains a disconnect between how sparsity affects network topology. Inspired by brain neuronal networks, we explore sparsity approaches through the lens of network topology. Specifically, we exploit mechanisms seen in biological networks, such as preferential attachment and redundant synapse pruning, and show that principled, model-agnostic sparsity approaches are performant and efficient across diverse NLP tasks, spanning both classification (such as natural language inference) and generation (summarization, machine translation), despite our sole objective not being optimizing performance. NeuroPrune is competitive with (or sometimes superior to) baselines on performance and can be up to $10$x faster in terms of training time for a given level of sparsity, simultaneously exhibiting measurable improvements in inference time in many cases.

We propose to prune a random forest (RF) for resource-constrained prediction. We first construct a RF and then prune it to optimize expected feature cost & accuracy. We pose pruning RFs as a novel 0-1 integer program with linear constraints that encourages feature re-use. We establish total unimodularity of the constraint set to prove that the corresponding LP relaxation solves the original integer program. We then exploit connections to combinatorial optimization and develop an efficient primal-dual algorithm, scalable to large datasets. In contrast to our bottom-up approach, which benefits from good RF initialization, conventional methods are top-down acquiring features based on their utility value and is generally intractable, requiring heuristics. Empirically, our pruning algorithm outperforms existing state-of-the-art resource-constrained algorithms.

Tree ensembles are powerful models that achieve excellent predictive performances, but can grow to unwieldy sizes. These ensembles are often post-processed (pruned) to reduce memory footprint and improve interpretability. We present ForestPrune, a novel optimization framework to post-process tree ensembles by pruning depth layers from individual trees. Since the number of nodes in a decision tree increases exponentially with tree depth, pruning deep trees drastically compactifies ensembles. We develop a specialized optimization algorithm to efficiently obtain high-quality solutions to problems under ForestPrune. Our algorithm typically reaches good solutions in seconds for medium-size datasets and ensembles, with 10000s of rows and 100s of trees, resulting in significant speedups over existing approaches. Our experiments demonstrate that ForestPrune produces parsimonious models that outperform models extracted by existing post-processing algorithms.

Ragnarock is a virtual reality (VR) rhythm game in which you play a Viking captain competing in a longship race. With two hammers, the task is to crush the incoming runes in sync with epic Viking music. The runes are defined by a beat map which the player can manually create. The creation of beat maps takes hours. This work aims to automate the process of beat map creation, also known as the task of learning to choreograph. The assignment is broken down into three parts: determining the timing of the beats (action placement), determining where in space the runes connected with the chosen beats should be placed (action selection) and web-application creation. For the first task of action placement, extraction of predominant local pulse (PLP) information from music recordings is used. This approach allows to learn where and how many beats are supposed to be placed. For the second task of action selection, Recurrent Neural Networks (RNN) are used, specifically Gated recurrent unit (GRU) to learn sequences of beats, and their patterns to be able to recreate those rules and receive completely new levels. Then the last task was to build a solution for non-technical players, the task was to combine the results of the first and the second parts into a web application for easy use. For this task the frontend was built using JavaScript and React and the backend - python and FastAPI.

Actually, that is a misconception of Machine Learning. One of the biggest kept secrets in the Machine Learning communities is that it does not require high level of statistics, mathematics, or any computer science degree to master. Why do most believe that? It is true that the invention and paradigms used in Machine Learning was created by people with high level degrees in these fields. But let me ask you a question.

This paper proposes weight regularization for a faster neural vocoder. Pruning time-consuming DNN modules is a promising way to realize a real-time vocoder on a CPU (e.g. WaveRNN, LPCNet). Regularization that encourages sparsity is also effective in avoiding the quality degradation created by pruning. However, the orders of weight matrices must be contiguous in SIMD size for fast vocoding. To ensure this order, we propose explicit SIMD size aware regularization. Our proposed method reshapes a weight matrix into a tensor so that the weights are aligned by group size in advance, and then computes the group Lasso-like regularization loss. Experiments on 70% sparse subband WaveRNN show that pruning in conventional Lasso and column-wise group Lasso degrades the synthetic speech's naturalness. The vocoder with proposed regularization 1) achieves comparable naturalness to that without pruning and 2) performs meaningfully faster than other conventional vocoders using regularization.