Tagged articles
pruning
25 articles · Page 1 of 1
This article presents PRIME, a unified integer linear programming approach that jointly decides per-expert retention, merging, or pruning based on activation frequency and output similarity, reducing routed-expert count by 25% with only ~1% accuracy drop on code/math benchmarks, and demonstrates up to 2.23× throughput improvement in high-concurrency long-output serving via working-set reduction and KV-capacity expansion.
MoEexpert compressioninteger linear programming
0 likes · 21 min read

TinySR introduces a lightweight diffusion transformer for real-world image super-resolution, achieving 5.68x inference speedup and 83% parameter reduction over its teacher model TSD-SR through learnable pruning, dynamic block activation, VAE compression, and removal of time/prompt modules, while maintaining competitive perceptual quality on benchmarks.
TinySRVAE lightweightingdiffusion model
0 likes · 17 min read

This article explains how JD's team built an attribution‑analysis Agent that maps analysts' investigative steps into a plan‑and‑action loop, uses parallel search with pruning, script constraints, and dynamic structured memory to make data‑driven root‑cause analysis faster, more reliable, and interactive.
AI AgentAttribution AnalysisDynamic Memory
0 likes · 13 min read

The article explains SINDI, a sparse‑vector retrieval index designed for VSAG, detailing its window‑based weighted inverted lists, multi‑layer pruning, quantization and reranking techniques that achieve dramatically faster index construction, smaller index size, and high query throughput compared with traditional inverted or graph indexes.
Maximum Inner Product SearchSparse VectorVSAG
0 likes · 21 min read

The article explains how to choose suitable large AI models for product use by focusing on three key aspects—model inference, post‑training optimization techniques such as quantization, pruning and distillation, and the role of GPUs, including their kernels, software stack, and computing platforms.
GPUPost-Traininginference
0 likes · 8 min read


AIWalkerMar 15, 2025 · Artificial Intelligence
SANA 1.5 introduces an efficient model‑growth pipeline, depth‑pruning, and inference‑time scaling that reuse a 1.6 B‑parameter foundation to train a 4.8 B model with 8× lower memory, 60 % less training time, and GenEval scores that rival or surpass much larger diffusion models.
Efficient TrainingInference ScalingModel Scaling
0 likes · 17 min read

KwaiCoder-23BA4-v1 is a 23B wide MoE code‑completion model that achieves state‑of‑the‑art performance on HumanEval, BigCodeBench and Fill‑in‑Middle benchmarks by using high‑quality data, a cost‑effective training pipeline that combines model pruning, knowledge distillation and fine‑grained merging, and extensive ablation studies.
AIModel Trainingbenchmark
0 likes · 10 min read

The article explains LeetCode 814, where a binary tree of 0s and 1s is pruned by recursively removing subtrees lacking a 1, using a post‑order traversal that returns null for nodes with value 0 and no retained children, achieving O(n) time and O(h) space.
AlgorithmC++Java
0 likes · 6 min read

The article first shares a frustrated JD interview experience regarding a 'personality' assessment, then introduces LeetCode problem 814 – Binary Tree Pruning – detailing the problem statement, examples, and providing concise post‑order traversal solutions in Java, C++, and Python.
C++JavaLeetCode
0 likes · 6 min read

The article chronicles Baidu Search’s Model Architecture Group’s evolution of deep‑learning‑driven search, detailing the shift from inverted‑index to semantic vector indexing, the use of transformer‑based models for text and image queries, large‑scale offline/online pipelines, and extensive GPU‑centric optimizations such as pruning, quantization and distillation, all aimed at delivering precise, cost‑effective results to hundreds of millions of users.
ERNIEGPU InferenceSearch Architecture
0 likes · 14 min read

This article presents a comprehensive overview of inference optimization methods—including model pruning, quantization, knowledge distillation, caching, instruction‑set acceleration, and operator fusion—and details a GPU‑centric parallel acceleration methodology with CUDA basics, performance‑analysis tools, theoretical limits, and practical case studies, all illustrated with real‑world examples from Tencent's intelligent dialogue products.
CachingGPU accelerationOperator Fusion
0 likes · 18 min read

This article provides a comprehensive overview of recent advances in deep learning model compression, focusing on pruning methods—including unstructured, structured, filter-wise, channel-wise, shape-wise, and stripe-wise approaches—and quantization techniques such as linear, non‑linear, clustering, power‑of‑two, binary, and 8‑bit quantization, while discussing evaluation criteria, sparsity ratios, fine‑tuning, and training‑aware quantization.
deep learningmodel compressionneural networks
0 likes · 23 min read

This article provides a comprehensive overview of deep learning model compression and acceleration methods, detailing pruning strategies, various pruning types, evaluation criteria, sparsity ratios, fine‑tuning procedures, as well as linear and non‑linear quantization approaches, their implementations, and practical considerations.
Efficiencydeep learningmodel compression
0 likes · 26 min read


Code DAODec 29, 2021 · Artificial Intelligence
This article dissects Optuna’s internal design by building three miniature versions (Minituna v1‑v3) that illustrate its main components, storage layer, sampling APIs, pruning mechanisms, and joint‑sampling concepts, while comparing them with Optuna’s full implementation.
MinitunaOptunaPython
0 likes · 24 min read

This article reviews BERT’s architecture, analyzes the storage and compute costs of each layer, and systematically presents compression methods—including quantization, pruning, knowledge distillation (Distilled BiLSTM and MobileBERT), and structure‑preserving techniques—aimed at enabling efficient deployment on resource‑constrained mobile devices.
BERTMobile Deploymentknowledge distillation
0 likes · 15 min read

This article examines the internal structure of BERT and systematically presents various model‑compression strategies—including quantization, pruning, knowledge distillation, and structure‑preserving techniques—highlighting their impact on storage, computational cost, and inference speed for deployment on resource‑constrained mobile devices.
BERTknowledge distillationmobile AI
0 likes · 16 min read

Hammer is an integrated hardware-aware model compression tool developed by Kuaishou in collaboration with universities, combining pruning, quantization, search, and distillation to achieve efficient and accurate neural network models tailored to specific hardware.
AI frameworkKuaishouNAS
0 likes · 9 min read

This article reviews four mainstream model compression and acceleration methods—structural optimization, pruning, quantization, and knowledge distillation—explaining their principles, implementations, and performance, and presents practical examples such as DistillBERT, TinyBERT, and FastBERT with comparative results.
AIdeep learningknowledge distillation
0 likes · 14 min read


Didi TechOct 21, 2020 · Artificial Intelligence
The article reviews deep‑model compression methods—ADMM‑based structured pruning, low‑bit quantization, and teacher‑student knowledge distillation—and their automated AutoCompress workflow, demonstrating how these techniques shrink neural networks enough to run real‑time driver‑monitoring and other intelligent cockpit functions on resource‑limited automotive hardware while preserving accuracy.
ADMMdeep learningedge AI
0 likes · 16 min read


AntTechJun 9, 2020 · Artificial Intelligence
This article reviews the motivations, challenges, and a comprehensive set of algorithmic, framework, and hardware methods—including structural optimization, quantization, pruning, and knowledge distillation—to compress and accelerate deep learning models for deployment on mobile devices, highlighting benefits such as reduced server load, lower latency, improved reliability, and enhanced privacy.
knowledge distillationmobile AImodel compression
0 likes · 17 min read

Deep learning models often suffer from slow training and deployment due to their size, but a range of advanced acceleration methods—including model architecture optimization, pruning, quantization, knowledge distillation, and distributed training techniques—can dramatically improve speed and efficiency while maintaining performance.
Model Accelerationdeep learningdistributed training
0 likes · 14 min read

This article reviews neural‑network model quantization, explaining why quantization is needed, detailing forward‑ and backward‑propagation issues, presenting three main mitigation strategies, discussing subsequent pruning, performance‑recovery techniques, and outlining future research avenues in efficient machine learning.
efficient machine learningmodel quantizationneural networks
0 likes · 27 min read

This article reviews why deep neural networks are over‑parameterized, outlines the challenges of deploying them on mobile and embedded devices, and presents six major strategies—pruning, low‑rank approximation, filter selection, quantization, knowledge distillation, and novel architecture design—to accelerate and compress models while preserving performance.
Model Accelerationdeep learningknowledge distillation
0 likes · 11 min read

Decision trees are tree-structured classifiers that split data using attributes chosen for maximal purity measured by Gini impurity or entropy, with algorithms like ID3 selecting splits by information gain, while overfitting is mitigated through constraints and pruning techniques such as REP, PEP, and CCP.
AlgorithmGini ImpurityID3
0 likes · 13 min read

This article explains the backtracking algorithm as an enumeration‑like depth‑first search technique, outlines its fundamental concepts, basic ideas, typical problem‑solving steps, and provides both non‑recursive and recursive pseudo‑code frameworks for implementation.
Algorithmbacktrackingdepth-first-search
0 likes · 4 min read
