This repository is a curated knowledge base and learning guide for embodied AI and robotics. It aggregates documentation, paper summaries, interview notes, and learning roadmaps rather than providing executable software. The content is organized into sections covering foundations, classical methods, state-of-the-art techniques, simulation tools, and real-world robotics. Helper scripts for PDF processing support the documentation workflow by converting research papers into text for summarization. The repository serves as an educational resource and reference guide for practitioners and researchers in the field.
This repository is a curated collection of Jupyter notebooks and Python scripts that demonstrate fine-tuning, inference, quantization, and optimization techniques for state-of-the-art vision and multimodal models from the Hugging Face ecosystem. It covers a wide range of tasks including visual question answering, object detection, image classification, video understanding, multimodal retrieval, and preference alignment (DPO). The code is primarily educational, designed to be run interactively in Google Colab, with each notebook or script serving as a self-contained tutorial for a specific model or technique. Notable patterns include LoRA/QLoRA fine-tuning, knowledge distillation via a custom ImageDistilTrainer, any-to-any embedding for retrieval, and video frame downsampling for multimodal input. The repository is not a production library or application; it is a learning resource for engineers and researchers exploring vision-language models.
This repository is a large model zoo (500+ models) focused on converting and quantizing pre-trained computer vision models for edge deployment across multiple frameworks (TFLite, ONNX, OpenVINO, CoreML, TensorRT). It provides hundreds of download scripts for pre-converted model files, conversion pipelines (PyTorch/ TensorFlow to ONNX/TFLite), and a modest set of inference demo scripts for real-time video processing. The repo is primarily a research/engineering resource for edge AI practitioners, not a software application. It exhibits heavy duplication across model directories, relies on an external S3 bucket for model distribution, and uses older framework versions (TensorFlow 1.x/2.x, ONNX opset 11-13). Despite these maintenance concerns, the systematic conversion workflow and post-processing ONNX subgraph generators (e.g., NMS, box decoding) represent reusable patterns for anyone building end-to-end inference pipelines. The repo is hybrid in nature: while classified as data_research due to its focus on pre-trained model artifacts and lack of original training code, it also contains application-like demo code and CLI entry points.
This repository is a curated awesome-list of research papers, projects, datasets, and industry demos related to digital humans: 3D avatar generation and animation, clothed human reconstruction, head avatars, human motion generation, virtual try-on, cloth simulation, and related topics. It contains only a README.md plus LICENSE; there is no application code, no executable pipeline, no tests, no CI, and no dependency manifests. Despite the data_research classification, it is best understood as a manually maintained literature index rather than a data pipeline or software project.
This repository provides the SPICE dataset, a collection of molecular structures and properties for machine learning and molecular simulation. It includes precomputed HDF5 files for amino acid ligands, the DES370K set of small organic molecules, dipeptides, ions, PubChem compounds, solvated amino acids, water clusters, and more. The accompanying Python scripts generate these datasets from raw SDF and PDB files, performing tasks such as molecular filtering, energy computation, and solvation. The repository is primarily a data resource rather than a software application, with scripts that are tightly coupled to the specific dataset generation pipeline. It lacks dependency pinning and formal testing, but the data files are directly usable for training and benchmarking molecular models.
This repository is a comprehensive OSCP pentesting cheatsheet composed of markdown notes and diagrams. It serves as a reference guide for penetration testing methodologies, covering enumeration, privilege escalation, web application testing, Active Directory attacks, and more. While not containing executable code or data processing pipelines, it curates actionable information from various security resources into an organized format. The content is structured as checklists and step-by-step guides for common pentesting tasks, making it immediately applicable for individuals preparing for the OSCP certification or conducting real-world assessments. The repository does not ingest or transform external datasets; instead, it represents synthesized knowledge intended for direct consumption. Despite lacking traditional data engineering components, the repository demonstrates value as a knowledge artifact. Its maintenance is active with recent updates, and the static nature ensures zero technical drift. Reproducibility is trivial as the entire content is self-contained within the repository.
This repository is a collection of domain blocklists designed for use with Pi-hole and similar DNS-based ad-blocking services. It provides multiple specialized text files containing domain names to block, categorized by purpose (e.g., ads and trackers, telemetry, specific applications like TikTok and WhatsApp, malware, ransomware, and spam). The README explains how to integrate these lists into Pi-hole's adblocking configuration. While the repository lacks visible data processing code (showing only the final curated lists), the topical organization and maintenance updates suggest an ongoing curation effort aimed at removing false positives and optimizing the lists for specific use cases.
This repository is a curated collection of all-cause mortality data from 128 countries and territories, spanning 2015–2024, with additional local and preliminary mortality datasets. It is primarily a data asset rather than a software project: the core deliverable is the `world_mortality.csv` file, which aggregates weekly, monthly, or quarterly death counts from official sources such as the Human Mortality Database (STMF), Eurostat, and various national statistical offices. The READMEs provide exhaustive documentation of data sources, transformation notes (e.g., corrections for delayed registration, unknown-month distribution), and methodological caveats. No executable scripts (e.g., R or Python) are present in the provided file tree, meaning the ETL pipeline is not included in this repository. The data supports excess mortality analysis, notably the companion repository `dkobak/excess-mortality`. The project is actively maintained with recent commits and community engagement, but the lack of dependency management, environment definitions, and automated reproducibility infrastructure limits its reusability for other analysts.
This repository hosts the Hugging Face Deep Reinforcement Learning Course (v2.0), a free, open-source educational curriculum covering RL from fundamentals (Q-Learning, Policy Gradients) to advanced topics (PPO, A2C, Multi-Agent RL, Curiosity-Driven RL). It is primarily a collection of theoretical lessons (MDX files) and hands-on Jupyter notebooks designed to run on Google Colab. The notebooks train agents using libraries like Stable-Baselines3, CleanRL, ML-Agents, and Sample Factory, and push trained models to the Hugging Face Hub with evaluation videos and leaderboard comparison. While not a production application or data pipeline, the repo contains reusable training loops, evaluation helpers, and Hub packaging functions that can be extracted for other RL projects. The course is in low-maintenance mode, with some broken features (Unit 7 AI vs AI) and dependency drift (gym vs gymnasium), but the core content remains functional and well-structured.
This repository is primarily a collection of educational Jupyter notebooks and course materials accompanying the book 'Understanding Deep Learning' by Simon J.D. Prince. The content includes hands-on exercises implementing deep learning concepts from scratch, covering topics such as neural networks, optimization, convolutional networks, attention mechanisms, generative models, and reinforcement learning. The repository also contains a React-based website that serves as a frontend for the book, featuring sections for hero content, coding exercises, instructor resources, media, and further reading. While the website is a functional application, the dominant component is the educational notebook collection, which constitutes the majority of files and represents the core learning resource. The notebooks follow a consistent structure of theoretical explanations, code implementations with TODOs for students to complete, and visualizations to build intuition.
This repository provides Simplified Chinese language packs for WHMCS (Web Host Manager Complete Solution), a web hosting billing and automation platform. It contains PHP array files mapping English interface strings to Chinese translations for multiple versions of WHMCS (from 7.x to 9.0.4). The translations are generated using AI-assisted machine translation followed by human proofreading and optimization, as described in the accompanying documentation. The repository is actively maintained, with updates aligned to WHMCS releases, and uses the override mechanism for compatibility with newer WHMCS versions. While the repository does not contain reusable data processing components, it serves as a stable, versioned resource for deploying WHMCS in Chinese-speaking environments.
This repository implements a hierarchical controller for quadruped locomotion that combines a high-level reinforcement learning policy with a low-level model predictive controller (MPC). The RL policy dynamically predicts weight parameters for the MPC, which computes optimal ground reaction forces and joint torques. The codebase includes Isaac Gym training environments for the Unitree A1, Go1, and Aliengo robots, along with a custom C++ MPC solver (OSQP-based) and a finite state machine for robust transient handling. While primarily a research project that bridges RL and MPC for legged robots, it also provides reusable robotics components such as state estimators, leg controllers, gait pattern generators, and foot swing trajectory planners. The work is based on the MIT Cheetah software and includes detailed development logs and configuration files for reproducible experiments.
OpenClaw RTT is a data repository that stores normalized round-trip time (RTT) performance test results for the OpenClaw multi-channel messaging integration platform. It contains structured JSONL files under `data/channels/` for Telegram, Slack, Discord, and WhatsApp, alongside raw `result.json` files under `runs/`. The data is collected from automated QA canary and release harness runs executed via `pnpm openclaw qa <channel>`, with all tests using a mock OpenAI provider. The repository's primary function is to enable version-to-version performance regression tracking across chat channels, with summaries displayed in the README dashboard. The codebase includes import and summary scripts, validation logic, and CI workflows, but the majority of content is historical test result data. Notable patterns include consistent JSONL schemas, polling configurations, and resource measurement (max RSS, elapsed time). Many channel tests, especially Slack and WhatsApp, show persistent failures, indicating integration issues or broken test infrastructure.
This repository is a static collection of approximately 850 bitmap font images sourced from demoscene archives, retro games, and Amiga/Atari ST scene productions. It contains zero application code, no scripts, no data pipelines, and no notebooks — it is purely a curated asset archive. The fonts are stored as PNG files with descriptive filenames (often indicating the originating game, demo, or cracking group), organized in a flat fonts/ directory plus a font-pack/ subdirectory that preserves an original download from totorman.com including its bundled license.txt. The only non-image files are README.md, previews.md (a generated catalog embedding every font as an image preview), and the font-pack license/index artifacts. The repository's primary value is preservation: many of the original demoscene sites hosting these fonts have died, and this collection keeps the assets available. The previews.md file provides a browsable visual index of every font, which was deliberately moved out of the README to avoid GitHub throttling. The font-pack subdirectory retains its original Common Public License 1.0 terms from Marc Russell's spicypixel.net release, but the vast majority of fonts carry no provenance, attribution, or licensing metadata — the README explicitly disclaims knowledge of their origins. As a data repository, there are no transformations, analytical schemas, or research questions. It is best understood as a dataset for game developers, pixel artists, and retro-computing enthusiasts seeking authentic bitmap fonts for use in projects or reference. The lack of metadata, categorization, or machine-readable indexing limits its usability for automated consumption, but the raw assets are complete, immediately accessible by cloning, and visually verifiable through the preview catalog. Maintenance consists of sporadic font additions rather than structured development, and technical drift is effectively nil since the repository contains only stable PNG files with no dependencies.
This repository is not a software application, data pipeline, or research codebase. It is a content-distribution repository for a single educational artifact: a 'VIP cheatsheet' PDF condensing Stanford CME 295 (Transformers & Large Language Models). The repo contains 16 language folders, each holding one PDF and one translated README, plus a root README and an MIT license. There is zero executable code, no tests, no CI/CD, no dependency manifests, and no data processing logic of any kind. It is classified as data_research only by process of elimination — binary data files dominate and there are no application entry points — but in reality it is educational content, a hybrid that fits neither software nor data-engineering categories cleanly.
This repository is a curated collection of research papers on Multi-Agent Reinforcement Learning (MARL), organized by categories such as framework, cooperation, communication, and applications. It serves as a reference for researchers and practitioners in the field. The repository contains no code, data, or executable components; it is purely a markdown file with links to papers. The list is actively maintained and includes recent papers up to 2025. It is not a software application or data pipeline, but a static paper list.
This repository is a large personal collection of wallpaper images (primarily JPG/PNG, with some GIF and MP4 animated wallpapers) curated by the owner and shared publicly. It contains zero source code, no build system, no dependencies, and no automation anywhere in the tree. It is classified as data_research only by convention because its contents are raw data files rather than executable code, but it has no ETL pipeline, notebooks, analytics, or models — it is fundamentally a static asset repository with no hybrid software character. The collection is organized into thematic directories — Abstract, ChadVault, DistroLure, KawaiiWaifu (the largest, dominated by anime fan art), Minimalist, Mobile (with a WaifuSpecial subfolder), Pixelated, PureShots (photography), PurrfectCats, TimelessPaintings (classical art), VariantScenes, VideoWallpapers, and WideScreen — each containing a readme.md. Hundreds of additional images sit uncategorized at the repository root. Content spans AI-generated art (filenames carrying Midjourney-style prompt suffixes), Unsplash photography, wallhaven downloads, video game and anime fan art, and scans of classical paintings. Many filenames encode the original artist or generation prompt, which serves as basic provenance metadata. As an engineering artifact the repo offers nothing reusable: no scripts, clean abstractions, or data-processing logic exist to extract, and the highlighted_components list is therefore empty. Its value is purely as a curated image dataset — roughly a thousand-plus hand-selected wallpapers across aesthetic categories, with 306 stars and 22 forks indicating genuine community utility. Reproduction is trivial since all assets are bundled in-repo and no external services or credentials are required. The main limitations are the lack of machine-readable metadata (no resolution, source, or license tags) and the large number of uncategorized root-level files, which make automated discovery and licensing audits harder than the folder structure alone suggests.
This repository is a static asset dump of extracted Pokémon GO game files, not a software application. It contains 3D models (FBX/OBJ), textures (PNG), app icons, and text strings organized under top-level directories such as '3D Assets', 'Images', and 'Texts'. The assets are grouped by category (Pokémon, costumes, addressable assets) and by in-game event (e.g., Halloween 2021, New Years 2020/2021), with consistent naming conventions based on Pokémon IDs (pm####) and costume descriptors. There is no source code, ETL pipeline, or analytical schema; the repository serves as a raw data source for fan projects, 3D artists, and game data miners.
This repository is a curated collection of machine unlearning research, combining a comprehensive survey paper (in the form of a Kaggle notebook) with executable implementations of several unlearning algorithms across different domains. It includes code for graph neural network unlearning (GraphProjector), federated learning unlearning (FedEraser), and approximate nearest neighbor search with unlearning (UnfoldedSelf-ReconstructionHashing, a fork of ann-benchmarks). The repo serves both as a literature index (taxonomy, evaluation metrics, datasets) and a benchmark hub for reproducing and comparing unlearning methods. While the survey is actively maintained, the algorithm codebases are research prototypes with varying levels of polish and portability.
This is a research repository from the SCIR-HI lab at Harbin Institute of Technology, focused on instruction-tuning large language models (LLaMA, Alpaca, Bloom, Huozi) with Chinese medical knowledge. The core contribution is the construction of Chinese medical instruction datasets using structured knowledge bases (cMeKG) and medical literature via GPT3.5, followed by LoRA-based fine-tuning. The repository provides scripts for fine-tuning, inference, and model export, along with pre-trained LoRA weights. While it demonstrates a complete research pipeline (data generation, training, evaluation), it is primarily a research artifact rather than a production application. The code is functional and includes a clean, reusable Prompter class for template management, but lacks software engineering rigor (no tests, no CI). Reproducibility is moderate: dependencies are pinned, but the data generation process is not fully automated, and external base models and LoRA weights must be obtained separately.