16 مستودعات
Utilities for inspecting, debugging, and profiling WebGPU applications.
Explore 16 awesome GitHub repositories matching part of an awesome list · Debugging and Profiling Tools. Refine with filters or upvote what's useful.
RenderDoc هو مصحح أخطاء إطارات رسومية ومحلل أداء GPU عبر المنصات يُستخدم لالتقاط وتحليل استدعاءات واجهة برمجة تطبيقات الرسوميات ثلاثية الأبعاد وحالة إطار واحد من العرض. يعمل كأداة تصحيح أخطاء لـ Vulkan وOpenGL وOpenGL ES وDirect3D 11 و12. توفر الأداة قدرات تصحيح أخطاء متخصصة لفحص خطوط أنابيب الرسوميات وحالات الموارد عبر Windows وLinux وAndroid وNintendo Switch. وهي تتيح تحليل تظليل DirectX، وتوصيف تطبيقات OpenGL، وتطوير واجهة برمجة تطبيقات Vulkan للتحقق من سلوك العرض وإصلاح الأخطاء البصرية.
Serves as a cross-platform GPU profiler supporting Windows, Linux, Android, and Nintendo Switch for rendering optimization.
Mage AI is a Python-based data pipeline orchestrator and self-hosted data integrated development environment. It is designed for building, scheduling, and monitoring data workflows using a block-based pipeline design and interactive notebook interface. The platform distinguishes itself by integrating generative AI capabilities, allowing users to connect large language model providers via API to incorporate artificial intelligence into automated data streams. It also functions as an Apache Spark data processor, managing the kernels and infrastructure required for high-volume analytics and larg
Provides step-by-step logs and live data previews to visually identify and fix logic issues in pipelines.
BertViz هي مجموعة أدوات تشخيصية لتصور رؤوس الانتباه وسلوكيات النماذج الداخلية لتفسير كيفية معالجة نماذج اللغات للنصوص. تعمل كأداة قابلية تفسير ومصحح أخطاء لنماذج معالجة اللغات الطبيعية، حيث توفر تحديداً خرائط تفاعلية لآليات الانتباه داخل بنى Transformer. يتيح المشروع تحليل علاقات الرموز من خلال طرق عرض مفصلة لرؤوس وطبقات انتباه محددة. يدعم تصور الانتباه العالمي عبر جميع الطبقات، ورسم خرائط انتباه المشفر-فك المشفر، وفحص الخلايا العصبية الفردية داخل متجهات الاستعلام والمفتاح للكشف عن مساهمتها في حسابات الانتباه. توفر الأداة قدرات لتصفية طرق العرض حسب الطبقة، أو الرأس، أو أزواج الجمل. يتم عرض التصورات مباشرة داخل بيئات دفتر الملاحظات عبر حقن HTML وJavaScript، ويمكن تصديرها كملفات HTML مستقلة للمشاركة الخارجية.
Inspects the internal behavior of NLP models by visualizing encoder and decoder layers.
This is an open-source educational website that translates and localizes MIT's Missing Semester course, teaching practical computing skills for computer science students. The curriculum covers developer tooling, shell scripting, version control, security fundamentals, and open-source collaboration, with a focus on core computing skills including data processing pipelines, workflow automation, secure remote access, shell productivity, Vim editing, and Git version control. The project distinguishes itself by teaching command-line mastery, shell scripting, and automation to boost daily developer
Teaches how to diagnose errors and measure performance using debuggers and profilers.
Data-Juicer is an open-source framework for cleaning, filtering, deduplicating, and transforming multimodal datasets to prepare them for training large language and vision models. It functions as a distributed data pipeline engine that runs processing jobs across Ray clusters, handling billions of samples with automatic operator fusion and adaptive parallelism. The framework provides a library of operators that leverage large language models for semantic extraction, filtering, and data synthesis within processing pipelines. The project distinguishes itself through a YAML-based data recipe sys
Traces sample changes through distributed processing to debug and audit workflows.
Inspects generated code, traces intermediate values, and prints debug output during pipeline execution.
Provides a suite of tools for building, debugging, profiling, and developing software that utilizes GPU hardware.
Perfetto is a platform for system-level performance tracing and analysis on Linux and Android. It combines a high-throughput trace recorder, a SQL-based query engine, and a browser-based visualizer into a single toolchain. The platform covers CPU scheduling and call-stack profiling, native and Java heap memory allocation tracking, GPU and graphics events, and system-wide counters such as CPU frequency and power consumption. The architecture decouples trace recording from offline analysis, using a compact protobuf format for event encoding and columnar storage for efficient SQL queries. The we
Captures GPU events, counters, and render stages to debug graphics performance and frame timing.
dlt هي أداة لاستيعاب البيانات بلغة Python وإطار عمل لخط أنابيب ETL مصمم لجلب البيانات من مصادر متنوعة وحفظها في وجهات مهيكلة. تعمل كمحرك لاستنتاج المخطط (schema inference) يكتشف تلقائياً أنواع البيانات ويسطح هياكل JSON المتداخلة في جداول علائقية، ناقلاً البيانات من المصادر إلى بحيرات البيانات، أو المستودعات، أو قواعد بيانات المتجهات. يتميز المشروع بتوليد خط أنابيب مدعوم بالذكاء الاصطناعي، باستخدام نماذج لغات كبيرة لسقالات كود الاستخراج والموصلات لـ REST APIs. كما يدعم تخزين المتجهات متعدد الوسائط والتعبئة المتخصصة لقواعد بيانات المتجهات لدعم تطبيقات الذكاء الاصطناعي والتعلم الآلي. يغطي إطار العمل مجموعة واسعة من القدرات بما في ذلك تطور المخطط المؤتمت، وتحميل البيانات التزايدي عبر تتبع الحالة، والتحقق من جودة البيانات من خلال فرض عقود البيانات. يوفر أدوات لتطبيع البيانات العلائقية، وتحويلات ما قبل وما بعد التحميل، ومجموعة متنوعة من محولات الوجهة لقواعد بيانات SQL ومخازن الكائنات السحابية. تتم إدارة المراقبة من خلال لوحات معلومات تنفيذ خط الأنابيب، وتتبع نسب الأعمدة، والتحقق من إصدار المخطط باستخدام التجزئات القائمة على المحتوى.
Retrieves execution traces, timing, and error logs to troubleshoot and debug pipeline runs.
Jitsu is an open-source customer data platform designed to orchestrate event data pipelines. It captures, transforms, and routes behavioral data from web and server sources into data warehouses and analytics tools, providing a unified infrastructure for managing event streams. The platform distinguishes itself through its focus on self-hosted, containerized operations that grant users full control over their data security and privacy. It features a robust identity resolution engine that stitches disparate user identifiers into persistent profiles across sessions and devices, alongside program
Provides an interactive editor and debugger to test data processing functions against sample payloads.
SHADERed is a cross-platform shader development environment and debugger used for authoring vertex, pixel, and compute shaders. It functions as an integrated editor for writing and compiling shader code while providing live visual previews and the ability to configure render pipeline states. The tool features a GLSL shader debugger that enables line-by-line execution with breakpoints and variable watches to resolve visual logic errors. It includes a shader asset manager for importing 3D models and textures as data inputs and supports an extensible plugin system to integrate new languages and
Provides a comprehensive suite for debugging and profiling GPU programs via breakpoints and variable watches.
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,
Generates operator delegation tables, memory profiles, and time profiles to diagnose exported LLM performance.
Devtool is a suite of developer tools designed to execute server-side JavaScript code within a browser environment. It functions as a bridge that allows Node.js programs to run inside a Chromium instance, enabling the use of built-in browser developer tools for visual debugging and performance profiling. The project provides a read-eval-print loop for interactive command execution with support for relative imports and piped input. It also includes a file watcher that monitors source files for changes and automatically reloads programs to accelerate the iteration cycle. The system integrates
Provides a visual interface for setting breakpoints and running performance profiles to identify execution bottlenecks.
Habitat-sim is a high-performance 3D simulation platform designed for training and benchmarking embodied AI agents within photorealistic indoor and outdoor environments. It serves as a simulator for AI and robotics, providing a system for generating synthetic data and simulating physical interactions. The project is distinguished by a native C++ core that enables high-throughput simulation and a rendering pipeline using physically based rendering and baked global illumination. It features a navigation system based on pre-computed navigation meshes to ensure collision-free traversal and a rigi
Provides verbose logging and validation for GPU operations to resolve graphics context issues.
WebGPU devtools is a Web browser extension that helps debug WebGPU applications by monitoring WebGPU usage.
Browser extension for WebGPU development.
Install node. Note: I recommend using nvm(mac/linux) or nvm-windows(windows). run these commands:
Early-stage debugging tool for the API.