ما هي الميزات الرئيسية لـ leela-zero/leela-zero؟

الميزات الرئيسية لـ leela-zero/leela-zero هي: Deep Learning Game AIs, Deep Reinforcement Learning Implementations, Monte Carlo Tree Search, Neural Network Training Frameworks, Move Selection Networks, Reinforcement Learning, Residual Networks, Self-Play Training Pipelines.

ما هي البدائل مفتوحة المصدر لـ leela-zero/leela-zero؟

تشمل البدائل مفتوحة المصدر لـ leela-zero/leela-zero: leelachesszero/lc0 — Leela Chess Zero is a deep learning game AI and neural network chess engine that uses search algorithms to determine… suragnair/alpha-zero-general — This project is a reinforcement learning framework and game AI engine designed for training adversarial agents in… keras-rl/keras-rl — keras-rl is a reinforcement learning library that enables the training of neural agents using Keras. It serves as a… tensorflow/minigo — Minigo is a TensorFlow-based reinforcement learning engine designed to master the game of Go. It functions as a… tensorlayer/tensorlayer — TensorLayer is a backend-agnostic tensor library and deep learning framework designed for building neural network… tensorpack/tensorpack — Tensorpack is a high-level TensorFlow neural network framework and research library designed for building and training…

leela-zeroleela-zero

Name: leela-zero/leela-zero
Author: leela-zero

View on GitHub

5,579 نجوم1,015 تفرعاتC++GPL-3.07 مشاهدات

Leela Zero

Leela Zero هو محرك Go للتعلم العميق ونظام تعلم تعزيزي ينفذ نهج AlphaGo Zero. يستخدم شبكات تلافيفية متبقية عميقة وبحث شجرة مونت كارلو لتحديد التحركات المثلى وتحليل لعبة Go.

يعمل المشروع كأداة تدريب للشبكة العصبية تولد البيانات من خلال اللعب الذاتي المؤتمت. يستخدم خط تدريب خاضع للإشراف لتحسين أوزان الشبكة، مما يسمح للنظام بتحسين قدراته في اللعب دون الاعتماد على بيانات مقدمة من البشر أو معرفة الخبراء.

يتضمن المحرك منطق تسجيل اللعبة لتحديد الفائزين بناءً على قواعد الإقليم والأسر. كما يوفر أدوات لمعالجة بيانات اللعبة، مثل تحويل ملفات سجل اللعبة إلى تنسيقات رقمية مضغوطة لأطر عمل التعلم العميق.

للتفاعل مع الواجهات الرسومية الخارجية وأدوات التحليل، ينفذ المحرك بروتوكول نص Go (GTP) للاتصال الموحد.

Features

Deep Learning Game AIs - Implements a deep learning AI that masters Go through self-play and neural network training.
Deep Reinforcement Learning Implementations - Implements the AlphaGo Zero approach to learn Go through self-play using deep reinforcement learning.
Monte Carlo Tree Search - Utilizes Monte Carlo Tree Search to simulate potential move sequences and determine optimal play.
Neural Network Training Frameworks - Trains game-playing models using deep learning frameworks and exports the resulting network weights.
Move Selection Networks - Determines optimal moves by combining Monte Carlo Tree Search with deep residual convolutional neural networks.
Reinforcement Learning - Trains neural networks to master Go through a trial-and-error self-play process without human data.
Residual Networks - Uses residual networks with skip connections to predict move probabilities and game win rates.
Self-Play Training Pipelines - Generates training data through automated self-play games to iteratively improve the model.
Neural Network Training Toolkits - Provides a system for generating self-play data and iteratively optimizing deep learning models.
Go Game Rule Engines - Calculates final game scores using territory and capture rules to determine the winner.
أطر عمل تعلم الآلة - Reimplementation of reinforcement learning systems for game play.
Perception and Machine Learning - Reimplementation of reinforcement learning Go engines.

سجل النجوم

leela-zeroleela-zero

View on GitHub

5,579 نجوم1,015 تفرعاتC++GPL-3.07 مشاهدات

Leela Zero

للتفاعل مع الواجهات الرسومية الخارجية وأدوات التحليل، ينفذ المحرك بروتوكول نص Go (GTP) للاتصال الموحد.

Features

Deep Learning Game AIs - Implements a deep learning AI that masters Go through self-play and neural network training.
Deep Reinforcement Learning Implementations - Implements the AlphaGo Zero approach to learn Go through self-play using deep reinforcement learning.
Monte Carlo Tree Search - Utilizes Monte Carlo Tree Search to simulate potential move sequences and determine optimal play.
Neural Network Training Frameworks - Trains game-playing models using deep learning frameworks and exports the resulting network weights.
Move Selection Networks - Determines optimal moves by combining Monte Carlo Tree Search with deep residual convolutional neural networks.
Reinforcement Learning - Trains neural networks to master Go through a trial-and-error self-play process without human data.
Residual Networks - Uses residual networks with skip connections to predict move probabilities and game win rates.
Self-Play Training Pipelines - Generates training data through automated self-play games to iteratively improve the model.
Neural Network Training Toolkits - Provides a system for generating self-play data and iteratively optimizing deep learning models.
Go Game Rule Engines - Calculates final game scores using territory and capture rules to determine the winner.
أطر عمل تعلم الآلة - Reimplementation of reinforcement learning systems for game play.
Perception and Machine Learning - Reimplementation of reinforcement learning Go engines.

بدائل مفتوحة المصدر لـ Leela Zero

مشاريع مفتوحة المصدر مشابهة، مرتبة حسب عدد الميزات المشتركة مع Leela Zero.

leelachesszero/lc0
LeelaChessZero/lc0
2,991عرض على GitHub
Leela Chess Zero is a deep learning game AI and neural network chess engine that uses search algorithms to determine optimal moves and evaluate game states. It functions as a UCI chess engine, implementing the Universal Chess Interface standard for compatibility with various graphical user interfaces. The system acts as a hardware-accelerated move calculator, leveraging GPU and CPU backends to accelerate neural network inference. It supports the generation and submission of self-play games to training clients to improve the strength of its neural network models. The engine provides capabilit
C++alphazeroalphazero-inspiredchess
عرض على GitHub2,991
suragnair/alpha-zero-general
suragnair/alpha-zero-general
4,471عرض على GitHub
This project is a reinforcement learning framework and game AI engine designed for training adversarial agents in two-player turn-based games. It implements a training loop that utilizes self-play and Monte Carlo Tree Search to produce neural networks capable of predicting board strength and move probabilities. The system decouples the reinforcement learning engine from specific game rules through an abstract game logic interface, allowing for the definition of custom game rules, win conditions, and board representations. It supports integration with various deep learning frameworks to serve
Jupyter Notebook
عرض على GitHub4,471
keras-rl/keras-rl
keras-rl/keras-rl
5,551عرض على GitHub
keras-rl is a reinforcement learning library that enables the training of neural agents using Keras. It serves as a framework for implementing deep reinforcement learning agents that interact with simulated environments to discover optimal behaviors and maximize cumulative rewards. The library provides a system for configuring, training, and managing neural network agents. It handles the interaction loop between agents and environments, allowing models to learn through direct experience and gradient-based optimization. The framework includes capabilities for model weight management, allowing
Python
عرض على GitHub5,551
tensorflow/minigo
tensorflow/minigo
3,531عرض على GitHub
Minigo is a TensorFlow-based reinforcement learning engine designed to master the game of Go. It functions as a comprehensive system for training neural networks to predict board policies and game outcomes, utilizing a model trainer to generate self-play data and optimize weights. The project is distinguished by its ability to perform large-scale game simulations using Kubernetes to distribute worker nodes across CPU, GPU, and TPU hardware. It employs a Monte Carlo Tree Search implementation to identify optimal moves and supports specialized hardware acceleration, including inference on Edge
C++
عرض على GitHub3,531

عرض جميع البدائل الـ 30 لـ Leela Zero

الأسئلة الشائعة

ما هي وظيفة leela-zero/leela-zero؟

سجل النجوم

بدائل مفتوحة المصدر لـ Leela Zero

مشاريع مفتوحة المصدر مشابهة، مرتبة حسب عدد الميزات المشتركة مع Leela Zero.

leelachesszero/lc0
LeelaChessZero/lc0
2,991عرض على GitHub
Leela Chess Zero is a deep learning game AI and neural network chess engine that uses search algorithms to determine optimal moves and evaluate game states. It functions as a UCI chess engine, implementing the Universal Chess Interface standard for compatibility with various graphical user interfaces. The system acts as a hardware-accelerated move calculator, leveraging GPU and CPU backends to accelerate neural network inference. It supports the generation and submission of self-play games to training clients to improve the strength of its neural network models. The engine provides capabilit
C++alphazeroalphazero-inspiredchess
عرض على GitHub2,991
suragnair/alpha-zero-general
suragnair/alpha-zero-general
4,471عرض على GitHub
This project is a reinforcement learning framework and game AI engine designed for training adversarial agents in two-player turn-based games. It implements a training loop that utilizes self-play and Monte Carlo Tree Search to produce neural networks capable of predicting board strength and move probabilities. The system decouples the reinforcement learning engine from specific game rules through an abstract game logic interface, allowing for the definition of custom game rules, win conditions, and board representations. It supports integration with various deep learning frameworks to serve
Jupyter Notebook
عرض على GitHub4,471
keras-rl/keras-rl
keras-rl/keras-rl
5,551عرض على GitHub
keras-rl is a reinforcement learning library that enables the training of neural agents using Keras. It serves as a framework for implementing deep reinforcement learning agents that interact with simulated environments to discover optimal behaviors and maximize cumulative rewards. The library provides a system for configuring, training, and managing neural network agents. It handles the interaction loop between agents and environments, allowing models to learn through direct experience and gradient-based optimization. The framework includes capabilities for model weight management, allowing
Python
عرض على GitHub5,551
tensorflow/minigo
tensorflow/minigo
3,531عرض على GitHub
Minigo is a TensorFlow-based reinforcement learning engine designed to master the game of Go. It functions as a comprehensive system for training neural networks to predict board policies and game outcomes, utilizing a model trainer to generate self-play data and optimize weights. The project is distinguished by its ability to perform large-scale game simulations using Kubernetes to distribute worker nodes across CPU, GPU, and TPU hardware. It employs a Monte Carlo Tree Search implementation to identify optimal moves and supports specialized hardware acceleration, including inference on Edge
C++
عرض على GitHub3,531

عرض جميع البدائل الـ 30 لـ Leela Zero

Leela Zero

Features

سجل النجوم

Leela Zero

Features

بدائل مفتوحة المصدر لـ Leela Zero

LeelaChessZero/lc0

suragnair/alpha-zero-general

keras-rl/keras-rl

tensorflow/minigo

الأسئلة الشائعة

سجل النجوم

بدائل مفتوحة المصدر لـ Leela Zero

LeelaChessZero/lc0

suragnair/alpha-zero-general

keras-rl/keras-rl

tensorflow/minigo

الأسئلة الشائعة