ML Frameworks

A WebGPU library that treats shader files like typed TypeScript modules and runs identically in the browser, headless Node, or a deterministic mock for tests.

Official code for a book that ships PicoAgents, a full multi-agent framework written from scratch so you can see how the pieces fit together without vendor abstraction getting in the way.

It rounds up pre-built Windows binaries for AI libraries that typically force users into complicated, error-prone source builds.

A curated collection of Jupyter notebooks and R Markdown files covering the standard ML curriculum, useful mostly as a reference for how to structure your own.

It trains multi-step LLM agents by treating every turn as a step-level MDP transition instead of a single growing prompt-response sequence.

A production-hardened fork of slime that keeps massive MoE models from collapsing by obsessing over bit-wise alignment between rollout and training.

RLinf exists because fine-tuning policies for robots and agents still requires rewriting your training stack for every new simulator, world model, or hardware rig.

It exists to give individual quants an integrated, locally hosted alternative to cloud platforms by packing Qlib, LLM-driven factor mining, and broker-grade execution into a single Docker stack.

PySR hunts for human-readable equations that fit your data, using a Julia-backed genetic search engine wrapped in a scikit-learn interface.

rLLM lets you train any LLM agent with RL by swapping the client and adding a decorator—no framework lock-in required.

RF-DETR is Roboflow’s bet that a DINOv2 transformer backbone can finally beat YOLO on both speed and accuracy in real-world detection and segmentation tasks.

It squeezes massive PyTorch language models into a fraction of their usual memory using 8-bit and 4-bit quantization, enabling inference and fine-tuning on consumer hardware.

OpenMed packages clinical entity extraction and HIPAA-grade de-identification into models small enough for Apple Silicon and impatient DevOps teams.

SpeechBrain exists to stop researchers from stitching together a dozen different libraries every time they build a conversational AI system.

It turns LLM theory into a stack of runnable notebook exercises—problem sets, answers, and hardware benchmarks included.

loralib lets you specialize billion-parameter PyTorch models by training tiny low-rank matrices instead of touching the frozen base weights.

It exists so you can train PyTorch models against the noise and non-idealities of analog in-memory hardware before the silicon even exists.

A Jupyter Book that rebuilds classic algorithms in NumPy and shows you the math actually working.

Genblaze orchestrates generative video, image, and audio pipelines across providers like OpenAI, Runway, and NVIDIA NIM while generating a SHA-256-backed provenance manifest for every run.

A library of 98 structured skill packs that turns coding agents into end-to-end AI researchers, from distributed training to LaTeX submission.