Rui CarmoThat grand, spacious (and long ignored) branch of computing that I never really found a reason to graduate in (big mistake I only realized 30 years later) and that, these days, deals not just with thinking machines (so-called “hard” AI), but mostly with enhancing the usefulness of machines in general by exploring mechanisms to express and manipulate harvested knowledge.
Or just pretty pictures and sophisticated text generation. Either way, it seems we’re currently between AI winters, with a lot of investment in GPU hardware that still needs to prove itself consistently, measurably useful in real life use cases.
Note: Items pertaining to Large Language Models, Model Context Protocol, Speech and Stable Diffusion/Image Generation are now in their own section, as this page was getting too crowded. I will be moving other sections out of here as well as they grow over time, so that this page can focus on the more general aspects of AI and ML.
Resources
| Field | Category | Date | Link | Notes |
|---|---|---|---|---|
| General | Examples | 2024 | flash-attention-minimal | a minimal implementation of Flash Attention |
| Frameworks | tinygrad | A 8000 LOC deep learning framework |
||
| 2023 | marvin | A generic wrapper for various AI APIs |
||
| MLX | An array framework for Apple silicon |
|||
| mlx-examples | MLX examples |
|||
| Jupyter | jupyter-ai | an official Jupyter plugin that can handle multiple AI back-ends (although it seems less flexible than the others right now) |
||
| Libraries | unstructured | a library for handling and segmenting unstructured data of various kinds, from text to common file formats |
||
| ml-ane-transformers | Apple’s transformers library, optimized for the Neural Engine |
|||
| 2009 | Alchemy | A toolkit providing a series of algorithms for statistical relational learning and probabilistic logic inference, based on Markov logic representation |
||
| Networking | 2024 | scuda | a GPU over IP bridge that allows remote GPUs to be utilized by CPU-only machines |
|
| Tools | scrape-it-now | A flexible scraper implemented in Azure |
||
| treescope | An interactive tensor visualizer for IPython notebooks. |
|||
| pico-tflmicro | a port of TensorFlow Lite Micro to the Raspberry Pi Pico |
|||
| 2023 | basaran | An Open-Source alternative to the OpenAI text completion API, with a compatible streaming API for privately hosted models. |
||
| explainerdashboard | a web app that explains the workings of a (scikit-learn compatible) machine learning model |
|||
| NVIDIA Triton Inference Server | A high-performance inference server |
|||
| 2022 | Project Bumblebee | Pre-trained and transformer neural models in Elixir. |
||
| Vector Databases | 2024 | txtai | embeddings database for semantic search |
|
| tinkerbird | Vector Database atop IndexedDB |
|||
| Large Language Models | Frameworks | 2025 | NLWeb | an approach to make websites easier to integrate into conversational interfaces |
| Multimodal Models | Libraries | 2024 | zerox | a library that performs OCR on documents and converts them to Markdown |
| Models | Hybrid-Net | Real-time audio to chords, lyrics, beat, and melody. |
||
| Music Generation | 2025 | NotaGen | a tool for generating music from text prompts |
|
| Tools | 2024 | swift-ocr-llm-powered-pdf-to-markdown | a tool that processes PDF files into structured Markdown format |
|
| Time Series Forecasting | Models | 2025 | moment | a time series forecasting model |
| Vision | Tools | 2024 | machina | a CCTV viewer that connects to RTSP streams and performs real-time object tagging using YOLO and ollama |