Command-line toolkit for interacting with the Naya Create split keyboard over USB CDC.

CLI text optimizer built on GEPA. Uses Agentic Coding CLI's as mutator and observer -- no api keys required

Stable Looped Models

The 7-Zip derivative intended for macOS

Grant private entitlements to OSX apps

FlexTensor is a tensor offloading and management library for PyTorch that enables running large models on limited GPU memory by intelligently offloading tensors between GPU and CPU memory.

TriAttention — Efficient long reasoning with trigonometric KV cache compression. Enables OpenClaw local deployment on memory-constrained GPUs.

🎨 NeMo Data Designer: Generate high-quality synthetic data from scratch or from seed data.

turboquant-based compression engine for LLM KV cache

MathCode: A Frontier Mathematical Coding Agent

Python package for LLM compression

Code for the papers: “Four Over Six: More Accurate NVFP4 Quantization with Adaptive Block Scaling” and “Adaptive Block-Scaled Data Types”

Inspects nsys dumps and measures NCCL collective launch skew

common in-memory tensor structure

Region-level profiling for CUDA kernels with trace, NVBit, CUPTI, NSys, and an interactive Explorer.

Rust implementation of protobuf with editions support, JSON serialization, and zero-copy views

APOLLO: SGD-like Memory, AdamW-level Performance; MLSys'25 Oustanding Paper Honorable Mention

Compile programs directly into transformer weights. Includes a 2D convex-hull KV cache with O(log n) inference.

Repository for the blog post JAX-LM: Language Modeling and Distributed Training in JAX

🔥 LeetCode for PyTorch — practice implementing softmax, attention, GPT-2 and more from scratch with instant auto-grading. Jupyter-based, self-hosted or try online.

A framework for verifiable reasoning with language models.

Running a big model on a small laptop

REAP: Router-weighted Expert Activation Pruning for SMoE compression

Your personal intelligence agent. Watches the world from multiple data sources and pings you when something changes.