This repository is a data pipeline that transforms crowdsourced survey responses about Chinese university quality of life into a structured MkDocs static site. The core ETL script (`questionnaires/main.py`) ingests a CSV of desensitized survey results, applies name normalization via alias mapping, blacklist/whitelist filtering, and splits data into current and archived sets based on a cutoff date. It generates per-university Markdown files with Q&A sections, a navigation structure, and a MkDocs configuration. The pipeline is triggered by GitHub Actions on pushes to master, pushing generated content to a separate branch and to the production website repository. The data is licensed under CC BY-NC-SA 4.0. While the code is functional and well-organized for its purpose, it is tightly coupled to the specific survey schema and data files, limiting direct reuse as a library. The most reusable component is the `FilenameMap` class for generating unique slugs from entity names.
This repository is a curated reference list of academic papers and projects focused on style transfer using diffusion models. It organizes resources into categories including image synthesis (text-driven, image-driven, LoRA-based methods), video synthesis, 3D generation, style datasets, and related surveys. The repository contains no executable code, data, or training scripts; it serves solely as a bibliographic guide for researchers. Content is presented in markdown format with tables linking to arXiv papers and GitHub repositories, supplemented by badges for stars and publication dates.
This repository provides the official code, datasets, and checkpoints for the Timer family of large time series foundation models (Timer, Timer-XL, Sundial) from Tsinghua University. It implements a decoder-only Transformer architecture pre-trained on the Unified Time Series Dataset (UTSD) using the single-series sequence (S3) format. The code supports fine-tuning and evaluation on three downstream tasks: forecasting, segment-level imputation, and predictive anomaly detection. Data loading modules handle multiple time series formats (CSV, NPZ, TXT) with flexible windowing and scaling. The repository includes shell scripts for reproducible experiments across multiple datasets and data scarcity ratios. It also provides HuggingFace integration for zero-shot forecasting without training. While primarily a research codebase rather than a production application, it is well-structured with clean module separation and reusable components.
This repository is an extensive, open-source educational tutorial on building and training large language models (LLMs) from scratch, maintained by the Datawhale community. It covers the full pipeline: NLP fundamentals, Transformer architecture (with a custom Tiny-K implementation), pretraining, supervised fine-tuning (SFT), parameter-efficient fine-tuning (LoRA), reinforcement learning (GRPO, OPD, Search-R1, ReTool), and practical applications like RAG and Agent systems. The code is organized into chapters with clear progression, and each chapter includes runnable scripts, configuration files, and detailed documentation in Chinese. While primarily educational, the repository contains several well-crafted, reusable components such as a modular Transformer implementation, efficient dataset loaders, a RAG pipeline, an Agent framework, and RL training loops with distributed support. The project is actively maintained with recent additions (Extra-Chapter) and has a large community following. However, reproducibility is moderate due to fragmented dependency files, hardcoded paths, and reliance on external services (e.g., DeepSeek Search, SwanLab). The code is not production-grade but serves as an excellent learning resource and a source of reusable patterns for LLM research and development.
PLINDER is a comprehensive protein-ligand interaction dataset and evaluation resource, providing over 400,000 annotated systems with 750+ annotation columns derived from the PDB. The repository contains a full data generation pipeline (Metaflow-based, configurable via YAML) that downloads, annotates, scores, clusters, and splits structures into train/val/test sets with leakage control. It also includes a Python API (PlinderSystem, PlinderDataset) for accessing the dataset, querying similarity scores, and loading structures, as well as an evaluation harness (ModelScores, StratifiedTestSet) for benchmarking docking algorithms using metrics like lDDT-PLI, DockQ, and posebusters checks. The codebase is hybrid: primarily a data research project with strong software engineering practices (Docker, CI/CD, extensive tests, config-driven pipelines), but the core value is the curated dataset and the tools to generate and evaluate it.
TRELLIS.2 is a research codebase for state-of-the-art image-to-3D generation using a novel sparse voxel representation (O-Voxel) and flow matching with 4B parameters. It provides full training and inference pipelines for generating high-fidelity 3D meshes with PBR materials from single images. The repository includes custom CUDA extensions for voxel processing, a flexible sparse tensor library, and comprehensive dataset tooling for large-scale 3D asset curation (Objaverse-XL, Sketchfab). While primarily a research project with training code and data pipelines, it also offers Gradio web demos and Python APIs for inference. The codebase demonstrates strong engineering in its modular design (pipelines, mixins, config-driven training) but suffers from some research-typical issues like code duplication, hardcoded paths, and fragile dependency management.
This repository maintains the curated metadata for GitHub Topics and Collections. It is not a running application but a structured dataset of topic definitions (with aliases, related topics, logos, Wikipedia links) and curated repository lists (collections). The actual 'application' code is a suite of Ruby tests (RSpec/Minimatch) that validate metadata correctness, image dimensions, link validity, and API referential integrity via batched GraphQL queries. CI/CD workflows enforce contribution guidelines, auto-correct renamed repos, and deploy a Jekyll-generated JSON feed. While primarily a data repository, it exhibits strong software engineering practices in validation, caching, and automation. The content is community-curated and licensed under CC-BY 4.0.
This repository is a curated list of MCP (Model Context Protocol) servers and tools specifically designed for the Korean ecosystem. It serves as a community-driven resource that organizes links to various MCP servers across categories such as legal, commerce, finance, real estate, maps, search, tourism, public data, weather, Korean NLP, collaboration, and miscellaneous. The list is maintained in the README.md file and follows specific inclusion criteria to ensure relevance and quality. The project provides clear contributing guidelines in both Korean and English, detailing the eligibility rules, pull request format, quality standards, maintenance policy, and code of conduct. This ensures that contributions are consistent and meet the project's standards. The repository is actively maintained, as evidenced by recent commits and ongoing community engagement (stars, issues, forks). While the repository does not contain any executable code or data processing pipelines, it functions as a valuable discovery tool for developers and researchers seeking MCP servers for Korean services. Its strength lies in its organization, clarity, and the effort to keep the list up-to-date with relevant projects. The lack of code means there is minimal technical debt or dependency concerns, contributing to its high reproducibility and low technical drift.
This repository is a curated list of academic papers on learning-based odometry, covering visual odometry (VO), visual-inertial odometry (VIO), inertial odometry, LiDAR odometry, semantic SLAM, and NeRF SLAM. It is organized as a README with tables detailing paper titles, venues, years, links to papers and code repositories, and short notes on each method. The list includes papers from 2015 to 2026 and is regularly updated with recent contributions. While no code or datasets are provided, the README serves as a comprehensive bibliography for researchers and practitioners in the field. It links to numerous external GitHub repositories and blogs for further exploration.
This repository is the canonical ADC Performance Survey dataset maintained by Boris Murmann. It aggregates published ADC (analog-to-digital converter) performance data from ISSCC and VLSI Symposium papers between 1997 and 2026. The primary data artifact is an Excel workbook (xls/ADCsurvey_latest.xlsx) that is consumed by a set of Jupyter notebooks under plots/ to generate publication-quality figures such as FOMs plots, energy plots, aperture trend plots, and FOM trend plots. The repo also contains PDF resources and a short course presentation.
graph_weather is a research-oriented deep learning library for weather forecasting with graph neural networks. It implements multiple model familiesβGraphCast-inspired encoder/processor/decoder models, the diffusion-based GenCast and GenDA, the ensemble FGN, WeatherMesh, Aurora, FengWu-GHR, and CaFAβalong with a substantial set of data loaders, loss functions, training scripts, and evaluation utilities. The package is structured as a Python library with HuggingFace Hub model integration, PyTorch Lightning training support, and a comprehensive test suite, giving it a hybrid character: it is fundamentally research code, but packaged and tested with application-grade discipline. Entry points include GraphWeatherForecaster, GraphWeatherAssimilator, WeatherStationReader, and various dataset classes, while training scripts live under train/.
This repository is a curated awesome list and survey companion for diffusion models applied to time series, spatio-temporal, and tabular data. It contains no executable code; the sole content is a large README.md that organizes hundreds of papers into a taxonomy by task (prediction, generation, imputation, anomaly detection, classification/regression, causal inference, event prediction, foundation models) and by application domain (healthcare, sequential recommendation, weather, energy, finance, AIOps, environment, traffic, manufacturing, audio, biology, communication). Each entry links to the paper and, where available, official code repositories. It is the companion to an ACM Computing Surveys paper (DOI 10.1145/3783986) and includes citation information, a Google Sheets taxonomy, and an overview figure. The repository is actively maintained with recent commits and welcomes pull requests. As a research resource, it is highly reproducible: all content is in the markdown file, with no dependencies, datasets, or environment setup required. Although classified as data_research due to its research orientation, it is not a data pipeline or ML codebase. Its value is as a reference index and literature map for researchers; the README itself is the reusable asset.
This repository hosts the Software Carbon Intensity (SCI) specification, a standardized methodology for measuring and reporting the carbon emissions of software systems as a rate (emissions per functional unit). The core artifact is SPEC.md, which defines the SCI calculation framework including formulas for operational and embodied emissions, guidance on system boundaries and functional units, and quantification methods. The repository also includes a static website (generated from SPEC.md) that presents the specification, an example case study demonstrating application of the methodology to a real web application (eShoppen), and automation for building and deploying the site. While the repository does not process data or implement the SCI calculation itself, it provides a complete, reusable framework for software carbon measurement that defines how to collect, transform, and interpret emissions data.
This repository is the source code for the Hugging Face Agents Course, a structured educational curriculum teaching the fundamentals and practical implementation of AI agents. The course is organized into units covering agent theory (ReAct, tool calling, observations), framework-specific tutorials (smolagents, LlamaIndex, LangGraph), agentic RAG, observability and evaluation, and a final project based on the GAIA benchmark. The content is delivered as MDX files with embedded code examples, quizzes, and links to interactive notebooks, and is translated into multiple languages (English, Spanish, French, Chinese, Korean, Russian, Vietnamese). While primarily an educational resource, the repository also contains reusable scripts for automated translation of course content and for pushing quiz data to the Hugging Face Hub, as well as CI/CD workflows for building documentation. The codebase is not a production application or a data pipeline, but rather a hybrid of educational content and supporting tooling for course maintenance.
This repository contains a Jupyter notebook that processes Strava data exports to generate interactive running heatmaps. The core functionality involves parsing GPS and sensor data from .fit.gz files, filtering activities by type and date, detecting a home location, rasterizing GPS tracks into multiple value grids (frequency, pace, heart rate, gradient, elevation change), applying Gaussian blurring and normalization, and rendering the results as overlay layers on a Leaflet basemap via Folium. The output is a single HTML file with six toggleable layers showing different aspects of running patterns. The code demonstrates solid data processing techniques including efficient caching, coordinate transformations, and custom visualization logic, though it remains tightly coupled as a monolithic notebook.
This repository is a research fork of Meta's Habitat-Sim, extended with 3D Gaussian Splatting (3DGS) for photo-realistic rendering in embodied AI navigation tasks. It provides a full simulation platform with physics (Bullet), sensors (RGB, depth, semantic, fisheye), navigation meshes, and a CUDA-accelerated Gaussian splatting rasterizer supporting dynamic avatars. The codebase includes training and evaluation pipelines for PointNav, ImageNav, ObjectNav, Vision-Language Navigation (StreamVLN, Uni-NaVid), and dynamic avatar avoidance. It also contains an LLM-driven agent system (HabitatAgent) with a bridge API, MCP server, and TUI for interactive navigation. The repository is hybrid: primarily a research platform for embodied AI experiments, but also includes application-like components such as web tools (navmesh editor, object annotation helper), a real-time Rerun viewer, and a conda packaging system. The core C++ simulator is well-architected with managed object patterns, replay recording, and PBR rendering, while the Python layer provides episode generation, prompt-based agent control, and analytics dashboards. Data assets (GS scenes, avatars, checkpoints) are hosted externally on Hugging Face, and the setup requires multiple conda environments for different tasks.
Audiocraft is a research codebase from Meta (Facebook Research) for state-of-the-art audio generation, including MusicGen, AudioGen, MAGNeT, JASCO, MultiBand Diffusion, and AudioSeal watermarking. It provides training and inference pipelines for text-to-music, text-to-sound, music continuation, style transfer, and symbolic conditioning (chords, drums, melody). The repository is primarily a research library with extensive configuration-driven training (Hydra/Dora), evaluation metrics (FAD, KLD, CLAP, ViSQOL), and interactive Gradio demos. It is not a production application but a collection of models, solvers, and utilities for audio generation research. The codebase is well-structured with modular components (conditioners, codebook patterns, audio effects) and includes unit tests, but relies on internal datasets and cluster paths that limit full reproducibility. Pretrained models are available via HuggingFace Hub for inference.
SkinTokens is a research codebase for a novel 3D character rigging method (TokenRig) that learns a compact discrete representation (SkinTokens) for skinning weights. The system ingests 3D meshes with skeleton and skinning data (from ArticulationXL, VRoid Hub, ModelsResource), transforms them through a pipeline of augmentation, order arrangement, vertex group extraction, and sampling, then trains an FSQ-CVAE to compress skinning weights into discrete tokens. A Qwen3-0.6B-based autoregressive transformer (TokenRig) generates the full rig (skeleton hierarchy + skin tokens) as a single token sequence. The repo includes a Blender-based server for mesh loading and export, a Gradio demo, a CLI, and extensive data processing modules. The code is well-structured with dataclass-based configs, abstract interfaces, and modular design, but remains primarily a research artifact with some incomplete training implementations (e.g., TokenRig training_step raises NotImplementedError). Its hybrid nature leans strongly toward data research, as the core contribution is a learned representation and generative model for animation rigging.
This repository is a large curated dataset rather than a software application or data pipeline. It contains a multilingual, awesome-list style directory of 3,527 generative AI applications organized into 80 categories, with README.md as the canonical content store. Each entry follows a reproducible markdown record schema: application name, screenshot URL, short description, visit redirect link, and a separate pricing/information link. Categories span the generative AI ecosystem, including AI agents, chatbots, image/video generation, code assistants, legal, accounting, and gaming. The data_research classification reflects the fact that the primary asset is structured content, not executable code; there are no notebooks, model files, or ETL scripts in the repository.
Python-100-Days (Python 100倩δ»ζ°ζε°ε€§εΈ) is one of the most widely referenced Chinese-language Python learning curricula on GitHub (~186k stars). Despite its scale, it is fundamentally an educational artifact rather than a deployable software system: the dominant content is a 100-day sequence of Markdown tutorials (Day01-100) augmented with teaching-oriented code examples, SQL schema scripts, and Jupyter notebooks. The closest software classification is data_research, driven by the substantial data-analysis track β Day66-80 covers NumPy, pandas, and Matplotlib against real bundled datasets (scraped BOSSη΄θ job postings across nine Chinese cities, 2020 sales data, 2023 Beijing hukou data), and Day81-90 covers scikit-learn machine learning with classic datasets (iris, Titanic, MPG, Boston housing). The repo's character is hybrid: it also carries a Django web-development track (Day46-60), a web-scraping track with Scrapy/Selenium (Day61-65), and database/DevOps chapters (Day36-45, Day91-100), but none of these constitute a runnable production application β they are teaching material. Ground truth from the code: the only real end-to-end 'data pipeline' is the job-market dataset pipeline. Raw CSVs under Day66-80/code/res/jobs/ contain scraped listings from zhipin.com with a consistent schema (company_name, uri, salary, site, year, edu, job_name, city, pos_count); the notebook workflow produces a cleaned artifact (cleaned_jobs-checkpoint.csv) with salary ranges parsed into salary_lower/salary_upper bounds. This demonstrates a reusable salary-parsing and job-market normalization pattern. The machine-learning chapters are textbook-style reproductions of classical algorithms for pedagogy, not original research, and the scraping data is a timestamp-less snapshot. Maintenance is active: the author continues to revise the curriculum (a 2025 version reset of the introductory series) and the last commit is recent. Technical drift is medium β older chapters reference Django 2.2, Vue 2.6.10, ECharts 4.2.1, and Python 3.7-era idioms, while some third-party API examples (TianAPI, chaojiying captcha OCR, SMS gateways) depend on external services that may have changed. Reproducibility is partial: dataset files needed by notebooks are checked in, but the main curriculum has no pinned dependency manifest (only the opencourse subproject has requirements.txt), no conda environment file, and the scraped data has no crawl timestamp.