14 dépôts
Capabilities for modifying input dimensions, batch sizes, and memory layouts during model execution.
Distinct from Dynamic Sync Shapes: None of the candidates address neural network tensor shapes; most relate to UI layouts or audio timbre.
Explore 14 awesome GitHub repositories matching artificial intelligence & ml · Dynamic Tensor Shapes. Refine with filters or upvote what's useful.
OpenVINO is an AI inference engine and model serving platform designed to execute optimized deep learning models across CPUs, GPUs, and NPUs through a unified API. It includes a model optimization toolkit for converting, quantizing, and compressing models from various frameworks, alongside a specialized generative AI runtime for large language models. The project distinguishes itself through a plugin-based hardware acceleration layer that maps neural network operations to vendor-specific drivers. It features advanced execution mechanisms such as continuous batching, speculative decoding, and
Allows modifying batch size and input shapes during execution to optimize the balance between throughput and latency.
This project is a structured learning curriculum and technical reference for mastering deep learning with TensorFlow. It provides a comprehensive guide for building, training, and deploying neural networks, combining theoretical fundamentals with practical implementation examples. The repository distinguishes itself by covering the end-to-end machine learning workflow, from low-level tensor mathematics and linear algebra to the creation of complex model architectures. It includes specific guidance on developing data pipelines for diverse data types, such as images, text, and time-series seque
Expands smaller tensor dimensions to match larger ones for compatible element-wise operations.
tensorrtx is a computer vision inference engine and model implementation library designed for graphics processor acceleration. It provides a framework for optimizing deep learning models through a GPU inference optimizer, a deep learning model converter for transforming weights from frameworks like TensorFlow and PyTorch, and a custom plugin library to implement operations not natively supported by the TensorRT API. The project distinguishes itself through a comprehensive collection of pre-defined network implementations, ranging from various YOLO versions and DETR transformers for object det
Provides capabilities for managing variable batch sizes and input dimensions through optimization profiles during model execution.
Flashlight est une bibliothèque de machine learning et de tenseurs autonome en C++ utilisée pour construire et entraîner des réseaux de neurones. Elle fonctionne comme un framework complet de réseaux de neurones et un moteur de différenciation automatique, fournissant les outils pour construire des graphes de calcul et calculer les gradients via la rétropropagation. Le projet sert de framework d'entraînement distribué, utilisant des opérations all-reduce pour synchroniser les gradients et les paramètres sur plusieurs nœuds de calcul et appareils. Il se distingue par une intégration profonde de la manipulation de tenseurs haute performance, l'interopérabilité native de la mémoire des appareils et un système pour synchroniser les poids entre les workers distribués afin d'accélérer l'entraînement de modèles à grande échelle. Le framework couvre un large éventail de capacités de deep learning, incluant la composition modulaire de couches pour concevoir des architectures complexes comme des blocs résiduels et des cellules récurrentes. Il fournit des utilitaires étendus de gestion de données pour l'ingestion et le préchargement, ainsi que des systèmes de sérialisation pour persister les états de modèle. De plus, il inclut une suite d'outils de surveillance et d'observabilité pour suivre les métriques d'entraînement et mesurer les erreurs de séquence. La bibliothèque est implémentée en C++.
Changes dimensions and permutes axes of tensors during model execution.
Pyrefly is a static type checker for Python that operates as a language server, delivering real-time diagnostics, completions, and navigation in any editor supporting the Language Server Protocol. It also performs static tensor shape analysis, using symbolic dimension variables and arithmetic to verify shape consistency in deep learning models without runtime execution. Beyond core type checking, Pyrefly supports gradual adoption workflows: it can generate a baseline of known errors so only new issues are reported, migrate configuration from other type checkers, and automatically suppress exi
Validates tensor shape consistency at compile time using symbolic arithmetic and shape rules for operations.
TileLang is a Python-embedded domain-specific language compiler that JIT-compiles and autotunes GPU kernels. It uses a tile-based DSL, automatic software pipelining, and parallel autotuning to generate optimized GPU kernels at runtime. It supports tensor core operations with Pythonic syntax, automatic memory management, and thread mapping. The compiler searches over tile sizes, thread counts, and scheduling policies, compiling and benchmarking candidates in parallel to find the fastest kernel. It also caches compiled binaries and tuning results to disk for reuse across sessions. TileLang inc
Queries supported tensor instruction shapes from the target architecture for instruction selection.
This project is a comprehensive instructional resource and course for building neural networks using PyTorch. It covers the fundamental building blocks of deep learning, including tensor manipulation, automatic differentiation, and the construction of modular neural network components. The repository serves as a technical guide for several specialized domains. It provides implementation details for computer vision tasks such as image classification, object detection, and semantic segmentation, as well as natural language processing workflows involving transformers, recurrent networks, and gen
Changes tensor structures through operations such as reshaping, squeezing, permuting, and rotating axes.
ExecuTorch is a lightweight C++ runtime for deploying PyTorch models on mobile, embedded, and edge hardware. It provides an ahead-of-time compilation pipeline that exports, quantizes, and lowers model graphs into compact serialized programs, then executes them through a minimal runtime with hardware acceleration and on-device large language model inference capabilities. The project distinguishes itself through a hardware accelerator delegate system that partitions model subgraphs and offloads computation to specialized backends including NPUs, GPUs, and DSPs from Apple, Arm, Intel, MediaTek,
Handles tensor dimension changes between inference runs without requiring model recompilation.
PaddleRec est une bibliothèque de recommandation par apprentissage profond et un framework d'entraînement de modèle distribué basé sur le framework PaddlePaddle. Il fournit une suite d'algorithmes et de modèles à l'échelle industrielle pour le matching d'utilisateurs et le classement de contenu personnalisé. Le projet inclut un moteur d'inférence de recommandation pour exporter et servir des modèles entraînés vers des environnements de production pour des requêtes en ligne en temps réel. Il permet l'implémentation d'algorithmes de recommandation par apprentissage profond pour traiter des jeux de données comportementaux massifs. Le framework couvre l'entraînement de modèles à grande échelle sur des clusters de calcul distribués et le développement de systèmes pour classer les éléments en fonction des préférences personnelles.
Supports dynamic tensor shapes to handle variable-length input sequences in recommendation ranking.
pytorch-summary est une collection d'utilitaires pour les réseaux de neurones PyTorch conçus pour générer des résumés de modèles, calculer les besoins en mémoire et visualiser les formes des tenseurs couche par couche. Il fonctionne comme un outil de reporting qui fournit des ventilations détaillées des couches du réseau et des formes de sortie pour aider au débogage et à l'inspection des modèles. Le projet fournit des capacités spécialisées pour estimer l'utilisation totale de la mémoire des passes avant et arrière en fonction des dimensions d'entrée et du nombre de paramètres. Il génère des visualisations lisibles par l'homme des structures de modèles pour vérifier les conceptions architecturales et identifier les inadéquations de dimensions entre les couches. L'outil implémente une analyse structurelle par le biais d'un parcours récursif des modules, d'un suivi des tenseurs basé sur des hooks et d'une inférence de forme pilotée par l'entrée. Ces capacités permettent l'agrégation du nombre de paramètres et le mapping du flux de données entre les opérations successives.
Determines output dimensions by passing dummy tensors through the network to trigger actual layer computations.
IREE is an MLIR-based compiler toolchain and runtime designed to translate machine learning models from various frameworks into optimized binaries for execution across diverse hardware targets. It provides a unified pipeline to ingest models from PyTorch, TensorFlow, JAX, and ONNX, lowering them into a common intermediate representation for deployment on CPUs, GPUs, and bare-metal embedded systems. The project distinguishes itself through a bytecode virtual machine and a hardware abstraction layer that decouple high-level model logic from specific hardware instruction sets. It supports sophis
Processes tensors with varying dimensions while prioritizing static shapes to maintain high performance.
xtensor is a C++ multidimensional array library for numerical computing that provides N-dimensional containers with an interface mirroring the NumPy API. It utilizes a lazy evaluation expression engine to defer numerical computations until assignment, which minimizes memory allocations and intermediate copies. The library features a foreign memory array adaptor that allows it to wrap external buffers, such as NumPy arrays, to perform numerical operations in-place without duplicating data. It further optimizes performance through lazy broadcasting and a system that manages the lifetime of temp
Expands tensor shapes lazily to execute operations without creating intermediate temporary copies.
ComfyUI-GGUF is a memory optimizer and model loader for ComfyUI that enables the execution of large transformer-based generative models using quantized weights. It provides a system for loading GGUF formatted weights within a node-based diffusion interface to reduce GPU memory consumption. The project includes a quantization tool for converting standard model checkpoints into compressed binary formats and a tensor fixer to restore missing keys and correct architectures in binary model files. These utilities ensure that compressed models remain functional during inference on hardware with limi
Corrects tensor shapes and restores missing keys in model architectures to ensure functional inference.
This project is a deep learning model compiler and parser that converts ONNX models into optimized TensorRT engines. It functions as a bridge that maps standardized ONNX operators to vendor-specific kernels to enable high-performance inference on NVIDIA GPUs. The system operates as a GPU inference optimizer, selecting hardware-specific kernels and tuning memory allocation to maximize throughput. It transforms neural network graphs into serialized binary execution plans to reduce runtime overhead. The toolset covers deep learning model deployment and edge AI performance tuning. It includes ca
Determines optimal memory allocation and tensor dimensions during the build process to support variable input sizes.