Awesome GitHub RepositoriesSequence Decoders

This project is a curated educational resource and solution repository for algorithmic challenges, specifically focused on LeetCode problems. It serves as a technical reference for common data structures and algorithmic patterns, providing verified code implementations across multiple programming languages alongside detailed logic and complexity analysis. The repository functions as a comprehensive study guide for competitive programming and technical interview preparation. It includes specialized learning tools such as an Anki flashcard dataset for spaced repetition and a browser extension t

Sglang is a high-performance inference engine and serving system designed for large language and multimodal models. It provides a programmable interface for orchestrating complex generation workflows, enabling developers to coordinate multi-turn dialogues, tool invocations, and reasoning chains through a domain-specific language. The platform is built to support production-scale deployments, offering an OpenAI-compatible API that allows for integration with existing application ecosystems. The system distinguishes itself through a disaggregated architecture that separates compute-intensive pr

This project is an educational platform and research toolkit designed to teach deep learning through a combination of mathematical theory, visual diagrams, and executable code. It provides a comprehensive environment for building, training, and evaluating neural networks, grounding complex concepts in interactive computational notebooks that allow for hands-on experimentation. The framework distinguishes itself by interleaving theoretical foundations—including linear algebra, calculus, and probability—with practical implementations across multiple industry-standard libraries. It supports flex

Tasmota is a universal firmware platform for ESP8266 and ESP32 microcontrollers, designed to provide local control and management of smart home hardware. It functions as an event-driven automation controller that replaces proprietary factory firmware, allowing users to manage relays, sensors, and lighting systems without relying on external cloud services. The system is built on a modular driver architecture that enables dynamic hardware configuration and peripheral support through a web-based management interface. The platform distinguishes itself through a template-driven hardware mapping s

This project is a comprehensive framework and toolkit for developing, optimizing, and deploying transformer-based models across multimodal, document intelligence, and natural language processing tasks. It provides a unified neural architecture that processes text, vision, audio, and document layout data through a shared set of weights, enabling researchers and developers to build foundational models that align cross-modal representations. The platform distinguishes itself through advanced training and inference strategies designed for large-scale deep learning. It incorporates specialized mec

FlashMLA is an LLM attention kernel library and inference acceleration library providing a collection of high-performance CUDA kernels. It implements multi-head latent attention mechanisms designed to reduce memory overhead and increase throughput during the forward and backward passes of large language model inference. The library utilizes quantized cache attention kernels to improve computation efficiency across both sparse and dense token processing. It specifically optimizes the prefill and decoding phases of model inference through these latent attention implementations. The project cov

HRM is an automated reasoning engine and language framework designed to execute complex, multi-scale problem solving. It functions as a reinforcement learning agent that continuously updates internal knowledge representations to improve task performance based on incoming data streams. The system distinguishes itself through a hierarchical architecture that coordinates abstract, long-term planning with granular, low-level logic. By integrating evolutionary algorithms and reinforcement learning, the framework refines model parameters and weights over successive generations, ensuring that intern

Ludwig is a multimodal machine learning platform and low-code framework designed for building, training, and deploying neural networks. It enables the construction of models that process text, images, audio, and tabular data through a unified interface using declarative configuration files rather than custom code. The system features a specialized low-code framework for large language models, supporting supervised fine-tuning, preference alignment, and a constrained decoding tool to force structured data output via logit extraction. It also includes an automated model architecture search to i

GroundingDINO is a deep learning vision model and open-vocabulary object detector designed to map natural language prompts to spatial coordinates. It functions as a text-to-bounding-box framework that enables zero-shot image localization, allowing the system to identify and locate arbitrary objects without requiring predefined classes or specific training for those categories. The project distinguishes itself by matching visual features to natural language descriptions to achieve open-set visual recognition. It supports text-guided image localization and the isolation of specific objects base

The Annotated Transformer is an educational resource that provides annotated code implementations of the Transformer architecture for sequence-to-sequence tasks, built with PyTorch. It serves as a learning tool for understanding attention mechanisms, multi-head parallel attention, and scaled dot-product attention through executable examples that walk through each component of the model. The project covers the full Transformer pipeline, including stacked encoder-decoder layers with residual connections and layer normalization, sinusoidal positional encoding for order-aware representation, and

This repository is a collection of practical deep learning implementations and examples built using the TensorFlow framework. It provides a variety of neural network architectures focusing on natural language processing, recommendation systems, reinforcement learning, and time series prediction. The project features a range of specialized models, including sequence-to-sequence and transformer architectures for text processing, and factorization machines for personalized ranking and retrieval. It also includes implementations of reinforcement learning agents using actor-critic and policy gradi

LMCache is a distributed key-value cache manager and tiering system designed to accelerate large language model inference. It functions as a tiered storage layer that offloads tensors from GPU memory to CPU RAM, local disks, or remote object stores, enabling the reuse of cached prefixes across different inference sessions and serving engines. The system differentiates itself through a disaggregated prefill-decode model, which separates prompt processing from token generation by transferring caches between distributed compute nodes. It utilizes peer-to-peer orchestration to share and retrieve

mistral.rs is an inference engine for large language models that runs locally and exposes models behind OpenAI and Anthropic-compatible APIs. It serves as a multi-model serving platform, capable of loading several models in a single server process with per-request routing and on-demand loading and unloading. The engine supports multimodal inference, processing text alongside images, video, audio, and speech inputs, and includes a quantized model deployment runtime that reduces memory use and speeds up inference on consumer hardware. The project distinguishes itself through an agentic tool exe

wav2letter is an automatic speech recognition toolkit and deep learning framework designed to convert audio speech signals into written text. It functions as a distributed training system and an inference engine for building and deploying neural network architectures. The system enables the training of large-scale speech models across multiple compute nodes using custom architecture files and structured recipes. It includes an inference engine that allows these trained models to be executed within Python workflows to transform audio sequences into text. The framework covers the full speech r

Dynamo is a distributed inference orchestration platform designed for large language models. It functions as a system to coordinate prefill and decode phases across GPU nodes, utilizing a multi-backend runtime adapter to connect engines like vLLM and TensorRT-LLM through a unified block-oriented memory interface. An OpenAI-compatible API server provides the frontend for integration with existing tools and clients. The project is distinguished by its disaggregated serving architecture, which separates prompt processing and token generation onto independent GPU pools to optimize throughput and

Neutts is a neural text-to-speech engine designed for real-time streaming output on edge devices such as phones and laptops. It supports voice cloning from short audio references, enabling zero-shot reproduction of a target speaker's voice, and can be fine-tuned or retrained from scratch for custom voices and styles. The system distinguishes itself through a decoder-only architecture that halves memory and accelerates generation on constrained hardware, combined with quantized model inference for reduced memory footprint. Its streaming decoder loop interleaves synthesis with playback, deliver

The official PyTorch implementation of Google's Gemma models

Este es un framework de encoder-decoder basado en TensorFlow y una librería de modelos utilizada para mapear secuencias de entrada a secuencias de salida. Funciona como un mapeador de secuencias de deep learning diseñado para transformar datos secuenciales de un dominio a otro. La librería proporciona herramientas para implementar modelos de secuencia a secuencia en múltiples dominios, incluyendo traducción automática neuronal, resumen automático de texto y generación de subtítulos para imágenes. El framework incorpora redes neuronales recurrentes y utiliza contextualización basada en atención para ponderar las secuencias de entrada. Admite múltiples estrategias de decodificación, incluyendo beam search y decodificación voraz, mientras ejecuta operaciones matemáticas mediante el grafo de cómputo de TensorFlow.

LiteRT-LM es un framework de inferencia de alto rendimiento diseñado para ejecutar modelos de lenguaje de gran tamaño localmente en hardware móvil, de escritorio y IoT. Sirve como un runtime de modelos en el dispositivo que utiliza aceleración de CPU, GPU y NPU para proporcionar un procesamiento de baja latencia. El framework se distingue por su capacidad para procesar entradas de texto, visión y audio a través de un único motor de inferencia multimodal. Cuenta con un servidor HTTP local que emula endpoints de API compatibles con OpenAI y un runtime basado en WebGPU para ejecutar modelos directamente dentro de un navegador web. Para garantizar la fiabilidad de la salida, incluye un generador de texto restringido que impone esquemas JSON o reglas gramaticales en las respuestas del modelo. El proyecto proporciona amplias capacidades para la gestión de conversaciones con estado, decodificación especulativa para aumentar las velocidades de generación de tokens y una interfaz de llamada a herramientas que mapea las solicitudes del modelo a funciones externas. También incluye integración especializada para el ecosistema Apple y un plugin dedicado para ejecutar modelos en Flutter. Los usuarios pueden ejecutar modelos a través de una interfaz de línea de comandos o integrarlos en aplicaciones mediante API nativas.