This is a research-grade PyTorch implementation of Danijar Hafner's Dreamer v4 model-based reinforcement learning algorithm. The core artifact is the `dreamer4` package, dominated by a monolithic `dreamer4/dreamer4.py` (~3000 lines) that implements `VideoTokenizer`/`SpaceTimeTokenizer`, `DynamicsWorldModel`, `AudioTokenizer`, and `StateTokenizer`. These classes build causal space-time transformers with separate space/time attention blocks, latent flow-matching diffusion with shortcut consistency, BYOL-style tokenizer pre-training, PoPE attention, and a large zoo of experimental loss terms (latent autoregressive loss, decorrelation, orthogonality, SigReg, self-flow distillation, shortcut loss). `dreamer4/trainers.py` provides training loops (VideoTokenizerTrainer, BehaviorCloneTrainer, DreamTrainer, SimTrainer) and `dreamer4/cli.py` exposes CLI entry points for training and serving the world model. The repository is hybrid: primarily research code, but packaged and tested like a library. It ships a pyproject.toml with pinned dependencies, a parametrized test suite (tests/test_dreamer.py, test_e2e.py, test_env_wrapper.py), a browser-based Snake environment with a Flask-style server (`dreamer4/web_env/`) used for human demonstration collection and replay-buffer inspection, and environment wrappers in `dreamer4/env.py` that record agent experience to folders or memmap replay buffers. Root-level scripts are the actual experiment entry points, targeting standard benchmarks: HalfCheetah (imagination RL), CartPole (dynamics + RL), MovingMNIST (tokenizer + dynamics), and Snake (PPO with dynamics model). Craftsmanship is above typical research code: clean package layout, pytest coverage across many hyperparameter combinations, self-contained utilities (LossNormalizer, tensor pack/unpack helpers, with_seed decorator, memmap replay-buffer helpers), and HuggingFace Accelerate integration for distributed/mixed-precision training. However, the single-file core, 100+ hyperparameter flags on trainers, scattered hard-coded constants, sparse docstrings, and reliance on volatile niche packages (x-transformers, h-net-dynamic-chunking, PoPE_pytorch, adam-atan2-pytorch) make long-term stability and code reuse risky. The repository is recently committed and appears maintained, but it is explicitly beta-status experimental research code, not a production system.
Exosphere is a reliability runtime for AI agents, providing a state-based workflow orchestration platform. The system consists of three main components: a Python SDK (`exospherehost`) for defining nodes and execution graphs, a FastAPI state-manager backend that handles state transitions, retries, cron triggers, and secret encryption, and a Next.js dashboard for monitoring and managing workflows. The architecture follows a microservices-like pattern within a monorepo, with the SDK acting as a client library and the state-manager as the central orchestration engine. Key capabilities include dynamic graph execution, infinite parallel fanout/unite patterns, configurable retry policies (exponential, linear, fixed with jitter), signal-based state pruning and re-queuing, and a graph-level key-value store. The system relies on MongoDB for persistence and uses Beanie ODM. The codebase demonstrates solid engineering practices with well-structured controllers, middleware for request ID propagation and exception handling, and a comprehensive test suite, though some tests are incomplete or skipped. The project is in early beta (v0.0.3b1) with active development and monthly releases.
This repository is a curated, structured list of 500+ AI startup market maps from 2024 to 2026, organized by quarter and subcategory (e.g., AI Agents, Compute, Healthcare). The primary artifact is a README.md with human-readable links, supplemented by a machine-readable CSV (`ai_market_maps.csv`), an RSS feed (`feeds/AIMarketMaps.xml`), and a Hugging Face dataset. The RSS feed is programmatically generated by `scripts/generate_rss.py`, which parses the README markdown and uses git history to assign publication dates. The repository serves founders, investors, and analysts by providing a comprehensive, frequently updated landscape of AI startup ecosystems. It is primarily a data curation and distribution project rather than a software application; the only code component is the RSS generation script and its associated CI workflow.
MiniMind-O is a research codebase for training and evaluating a small-scale (0.1B parameter) omni-modal language model that processes text, audio, and image inputs and generates text or streaming speech. The architecture uses a Thinker–Talker dual-path design with frozen external encoders (SenseVoice, SigLIP2, Mimi codec, CAM++) and trainable projectors, plus a Multi-Token Prediction (MTP) head for audio codec tokens. The repository includes a complete training pipeline (SFT with DDP, mixed precision, checkpointing), a comprehensive evaluation CLI, a Gradio web demo, and a real-time Flask WebSocket demo with VAD and ASR for near-duplex interaction. Data is stored in parquet format with pre-encoded audio codes, images, and speaker embeddings. The project is primarily research-oriented, aiming to provide a transparent, reproducible baseline for omni-modal models, but also includes application-like components for interactive demos.
Mage is an open-source data pipeline orchestration platform that combines a visual notebook-style editor with a production-grade scheduling and execution engine. The system enables users to build, run, and monitor data pipelines composed of Python, SQL, R, and dbt blocks. Architecturally, it is a full-stack web application with a Python backend (Tornado-based server, SQLAlchemy ORM, REST API) and a Next.js/React frontend with a custom component library. Key entry points include the CLI (`mage` command), the web UI served by `mage_ai/server/server.py`, and the API layer at `mage_ai/api/` which exposes CRUD resources for pipelines, blocks, runs, schedules, users, and integrations. Data flows center around pipeline execution: blocks are connected in a DAG, each block can be a data loader, transformer, or exporter. Execution is handled by a configurable executor (local, K8s, EMR, PySpark) and scheduled via a built-in scheduler or external triggers. The platform integrates with dozens of data sources and destinations through a Singer-based integration framework, supports streaming pipelines, and includes advanced features like RBAC, version control, AI-assisted pipeline generation, and a command center for global search. It depends heavily on external services such as cloud storage (S3, GCS, Azure), databases (PostgreSQL, BigQuery, Snowflake, etc.), and authentication providers (OAuth, LDAP, OIDC).
FastRTC is a Python library that enables developers to build real-time audio/video streaming applications with minimal code. It provides a high-level `Stream` class that integrates with Gradio for automatic UI generation and supports both WebRTC and WebSocket protocols. The library includes built-in handlers for voice activity detection (VAD) via Silero, speech-to-text via Moonshine, text-to-speech via Kokoro/Cartesia, and turn-taking logic. It also offers TURN credential helpers for Cloudflare, Twilio, and Hugging Face. The architecture revolves around defining a handler function that processes media chunks, wrapping it with turn-taking decorators, and creating a Stream object that handles signaling, connection management, and UI rendering. The codebase is actively maintained with frequent commits and an extensive set of demos integrating with various LLMs and AI services.
nlpaug is a Python library for data augmentation across text (character, word, sentence), audio, and spectrogram modalities. It provides a wide array of augmentation techniques including random insertion/substitution/deletion, synonym replacement via WordNet/PPDB, contextual word embeddings (BERT, GPT-2, RoBERTa), back-translation, TF-IDF-based augmentation, keyboard/OCR error simulation, and audio effects (noise, pitch shift, speed, VTLP). The library follows a consistent pattern: each augmenter inherits from a base class, accepts data via a unified `augment()` method, and returns augmented samples. Composition is supported through `Sequential` and `Sometimes` pipeline classes. It relies heavily on external deep learning models (HuggingFace Transformers, word2vec, GloVe, fastText) and audio processing (librosa). The codebase is well-structured with clear separation of concerns (augmenters, models, utilities, flows), extensive unit tests, and documentation. However, it exhibits significant technical drift due to aggressively pinned dependencies (e.g., `transformers>=5.9.0`, `numpy>=2.4.6`) that may not exist or conflict with current stable releases, and some integration code (conda setup for Python 3.5–3.9) is outdated. Despite this, the core augmentation logic and utility modules are sturdy and reusable.
This repository is a collection of C++ sample programs for deploying deep learning models on AXERA AI SoCs (NPU chips). It provides ready-to-run inference examples for a wide range of computer vision tasks including object detection (YOLOv5/6/7/8/9/10/11/26, YOLOX, YOLO-World, PPYOLOE+, RT-DETR, SCRFD), instance segmentation, oriented object detection (OBB), pose estimation, depth estimation, semantic segmentation, stereo matching, and more. Each example is a standalone command-line executable that loads a pre-compiled .axmodel, runs inference via the AX_ENGINE API, post-processes the outputs, and saves a result image. The code is organized per target chip (ax650, ax637, ax620e, ax615) with near-identical copies of examples across platforms, differing only in system initialization and minor chip-specific logic. The architecture is not a library or framework but a set of reference implementations demonstrating the integration of AXERA's NPU SDK. The primary data flow is: load image → preprocess (letterbox resize, color conversion) → push to engine → run synchronous inference → decode model outputs (e.g., anchor-based proposals, DFL, NMS) → draw bounding boxes/keypoints/masks → save output. External dependencies include the AX_ENGINE runtime, OpenCV for image processing, and standard C++ libraries. The build system uses CMake with per-chip toolchain files and GitHub Actions CI for cross-compilation verification. While the code is functional and covers a broad model zoo, it exhibits significant duplication across chip directories and hardcoded configuration (thresholds, image sizes, class names). The monolithic detection.hpp (~3500 lines) bundles many model-specific post-processing functions, increasing maintenance burden. There are no automated tests, and manual memory management is prevalent. Despite these quality issues, the repository serves its purpose as a practical reference for developers integrating AXERA hardware, and it stays current with recent model releases (YOLOv13, YOLO26, Depth Anything V2).
This repository is a curated list of papers, datasets, and code repositories for multi-turn interactions with large language models, accompanied by a static website that serves as a companion to a survey paper published in Transactions on Machine Learning Research (TMLR) 2026. The website includes sections on tasks (instruction following and conversational engagement), improvement methods, open challenges, and a searchable benchmark explorer. The repository also includes a script for processing survey data from Excel files and the README, though this script is not required to run the website.
This repository is the companion code for the O'Reilly book 'Hands-On Large Language Models' by Jay Alammar and Maarten Grootendorst. It provides 12 chapters of Jupyter notebooks covering fundamental and advanced LLM topics including tokenization, embeddings, text classification, clustering, prompt engineering, semantic search, multimodal models, fine-tuning, and generation. All notebooks are designed to run on Google Colab with free GPU resources and rely heavily on Hugging Face libraries (transformers, datasets, sentence-transformers, TRL) and external APIs (OpenAI, Cohere). The code is educational in nature, with clear step-by-step explanations and visualizations. While it is not a production application or a novel research pipeline, it serves as a high-quality, reproducible tutorial resource for practitioners learning LLM workflows. The repository includes pinned dependencies, Conda environment setup, and local installation guides, making it relatively easy to reproduce in a controlled environment, though API keys are required for some sections.
This repository is a comprehensive collection of educational tutorials that implement machine learning algorithms and applications entirely from scratch using only NumPy. It covers a vast range of topics including core ML algorithms (linear regression, SVM, K-means, PCA, etc.), neural networks (CNN, RNN, LSTM, GAN, transformer), computer vision (object detection, face recognition, OCR), natural language processing (sentiment analysis, chatbots, text summarization, BPE tokenizer), time series forecasting (ARIMA, LSTM, Prophet), recommendation systems, anomaly detection, and reinforcement learning. Each tutorial is a self-contained Python script that generates synthetic data with planted structure, trains a hand-coded model, and validates against naive baselines. The code is heavily commented and designed for teaching, not production. There are no real datasets, no external dependencies beyond NumPy, and no shared library—each file is independent, leading to significant code duplication across tutorials. The repository serves as an excellent learning resource for understanding the inner workings of ML algorithms, but it is not suitable for direct reuse in production systems without substantial refactoring.
WarriorJS is a terminal-based educational game that teaches JavaScript and TypeScript through interactive coding challenges. Players write a `Player` class with a `playTurn(warrior)` method to control a warrior navigating grid-based towers filled with enemies, captives, and stairs. The game runs as a CLI application using Ink (React for terminal UI) and relies on a sandboxed code execution environment via Node's `vm` module and esbuild. The monorepo structure includes a CLI app, game engine (`@warriorjs/core`), abilities library (`@warriorjs/abilities`), units library (`@warriorjs/units`), scoring module, spatial math helpers, effects, and two sample towers. The codebase demonstrates clean architecture with well-separated concerns, extensive test coverage, and reusable patterns such as the ability/effect binding factory and sensed space wrappers. However, the project has been inactive since 2018 (last release v0.14.0), the documentation website uses the deprecated Docusaurus v1, and there are some mixed module system usages. Despite being stale, the game engine and ability system remain functional and could be extracted for similar turn-based coding games or sandboxed code execution.
This is an educational repository of Python scripts, mini-projects, and markdown guides for building LLM-based AI agents applied to cybersecurity. The code is organized into five progressive parts: foundational LLM/embeddings usage, prompt templates, prompt chaining, RAG over security data, and agentic workflows. It demonstrates multiple agent frameworks (LangChain, LangGraph, CrewAI, DeepAgents) and MCP server integrations for tools like nmap and Shodan. The dominant character is application/tutorial code rather than data research, although it bundles small datasets (CVE embeddings, sample RAG corpora) and commits Chroma vector stores for offline exercises. The repo is not a single deployable service; it is a collection of entry-point scripts and small frameworks. Architecture is a set of standalone examples with shared patterns: state graphs for orchestration, YAML-configured crews, MCP adapters, and RAG ingestion/query pipelines. Most scripts require OpenAI-compatible API keys and/or local tools such as nmap/Ollama. Simulated data and placeholder responses in some examples make it a learning environment rather than production security tooling. Maintenance is current but drift is moderate: some examples use deprecated langchain vectorstore imports, documentation is duplicated across subdirectories, several referenced files are missing, and hardcoded model strings like gpt-5.4-mini appear dated or fictional. There is no root-level automated test suite; only the Shodan MCP subproject contains tests. The strongest reusable patterns are the MCP server wrapping, LangGraph conditional-state orchestration, CrewAI declarative crew configuration, and the hierarchical SOC agent supervisor.
Craftium is a voxel-based game engine (Luanti/Minetest fork) extended with reinforcement learning capabilities. The core system provides a moddable 3D world with block-based gameplay, Lua scripting API, client-server architecture, and procedural map generation. The craftium add-on supplies Python bindings, Gymnasium environment interfaces, and RL training scripts to enable machine learning research within the game world. The system supports both traditional gameplay and AI experimentation through environment-specific mods that define reward functions, observations, and termination conditions for agents.
EchoTorch is a PyTorch-based research library for reservoir computing and echo state networks (ESNs). It implements multiple ESN variants, including a self-predicting ESN (SPESNCell), conceptor-based pattern manipulation with logic operations (AND, OR, NOT, PHI), a custom TimeTensor time-series wrapper, and supporting datasets such as Lorenz, Roessler, Mackey-Glass, NARMA, and MNIST. The repository is a library rather than an application: there are no web servers, CLIs, or service entry points; the primary interface is the Python package `echotorch` with submodules for neural network cells, datasets, transforms, costs, and utilities.
This repository is a curated list of resources for practicing data science with Python, organized in the style of 'awesome' lists. It contains links to libraries, tutorials, blog posts, talks, code snippets, and other resources across various data science domains including core libraries, machine learning, NLP, computer vision, and more. The repository serves as a reference guide for data scientists and practitioners rather than containing executable code or performing data processing tasks.
DemoGPT is an open-source Python framework that provides two core capabilities: (1) an automatic Streamlit application generator that converts natural language descriptions into complete, runnable Streamlit apps through a multi-stage LLM pipeline (planning, task decomposition, code generation, assembly), and (2) the DemoGPT AgentHub, a modular library for building AI agents with built-in tools (web search, code execution, Wikipedia, weather, etc.), two agent types (ToolCallingAgent for single-step tasks, ReactAgent for multi-step reasoning), and a RAG (Retrieval Augmented Generation) system supporting Chroma, FAISS, and Pinecone vector stores. The codebase is primarily a library/application hybrid, with the Streamlit app acting as both a user interface and a demonstration of the generation pipeline. The architecture separates concerns into chains, prompts, task definitions, and controllers, but many components are tightly coupled to the specific generation pipeline. The AgentHub portion is more modular and reusable. Technical drift is medium: the code uses deprecated LangChain APIs (LLMChain, Chain.run), hardcodes older models like gpt-3.5-turbo-16k in some generated code, and has a broad dependency range. Maintenance appears active with recent commits and responsive issue triage, but the codebase would benefit from modernizing to LangChain's new LCEL syntax and updating default models.
This repository is primarily an extensive educational compendium covering mathematics, computer science, and artificial intelligence from first principles to state-of-the-art topics. It contains over 100 markdown files organized into 20 chapters, each with detailed explanations, diagrams, and interactive coding tasks (primarily in Python/JAX). The repo also includes auxiliary tooling: a MkDocs-based documentation site with MathJax and a Pyodide in-browser Python runner that includes a JAX compatibility shim, and a TypeScript MCP (Model Context Protocol) server that provides programmatic access to the compendium content for AI assistants. While the vast majority of the repository is educational content rather than runnable software, the MCP server and interactive runner are functional application components. The compendium is well-structured and current, but the code examples are pedagogical (simplified, not production-ready) and the JAX shim is a fragile polyfill. The project is not a data research pipeline—it contains no datasets, experiment logs, or trained models.
This repository is the archived reference implementation of the CAI (Cybersecurity AI) framework by Alias Robotics. It is a comprehensive, open-source platform for building and deploying AI-powered offensive and defensive security automation. The codebase provides a full-stack agent framework including a Python SDK (Agent, Runner, tools, handoffs, guardrails, tracing, voice, MCP), a CLI/REPL with rich interactive features, a Textual-based TUI, a FastAPI REST API server, and a benchmarking suite (CAIBench) with cyber ranges and CTF challenges. The framework supports multiple LLM providers (OpenAI, LiteLLM, custom) and includes specialized agents for red teaming, blue teaming, bug bounty, DFIR, network analysis, and more. It also features continuous operations (24/7 autonomous loops), session management, cost tracking, and extensive refusal testing (400+ scenarios). While the repository is now archived and will not receive further updates, the SDK components (function_tool decorator, Handoff abstraction, tracing processors, voice pipeline) are well-crafted and could be extracted for reuse in other agent-based projects. The codebase is large and monolithic, with some global mutable state and test duplication, but overall demonstrates strong engineering practices in the cybersecurity AI domain.
This repository is a curated 'awesome list' for Federated Learning (FL) research, maintained as a Jekyll static site. It aggregates hundreds of linked GitHub repositories and DBLP search strategies. The core data asset is a machine-generated go.mod file that pseudo-pins the linked repositories as Go dependencies, produced by the add_dependent.py script which extracts GitHub URLs from README.md, queries the GitHub API for commit hashes, and writes a go.mod. However, this is not a functional Go module; it is a novelty lockfile. The dblp-search-keywords.md document provides a meticulous analysis of DBLP search syntax to discover FL papers without the 'federate' keyword. The repository has significant hybrid character: while it is primarily a content aggregation site (application-adjacent), its technical substance lies in the data exploration and search strategy documentation (research-oriented). The actual code is minimal and tightly coupled to the repository's specific structure, limiting reusability. The add_dependent.py script is a one-time scraper with hardcoded rate limiting and token placeholder, not a general-purpose tool.