Stars
GPU tensor unlock source for RTX 4090 and experimental RTX 4070 Ti SUPER; firmware generated locally
Your GeForce runs AI at half speed. This fixes it for 20/30/40 series cards. Up to 2× FP16 / BF16 / TF32 tensor throughput
An agentic operating system where the kernel is controlled directly by Claude
Method for Long Context RLMs using verifiable Lambda Calculus
Various ML tidbits in Python/PyTorch and C++
Perturbed Saddle-escape Descent (PSD): a first-order optimizer that escapes strict saddle points in nonconvex problems.
Low Level Emulation of classic VA synths & effects of the late 90s/2000s by emulating the used ICs
Real-time terminal monitor for InfiniBand networks - htop for high-speed interconnects
Async web scraping framework on top of Rust. Works with Free-threaded Python (`PYTHON_GIL=0`).
Code for the "Can LLMs Reason Over Non-Text Modalities in a Training-Free Manner? A Case Study with In-Context Representation Learning" paper
Superposition Yields Robust Neural Scaling
Scripts for recreating the Replication Dataset for Fundamental Frequency Estimation. Part of the dissertation "Pitch of Voiced Speech in the Short-Time Fourier Transform". © 2020, Bastian Bechtold.…
🏆 OpenAI Open Model Hackathon Winner - Wildcard category. bota is an autonomous Dota 2 agent powered by GPT-OSS-20B and vLLM, created for OpenAI Open Model Hackathon.
An extension of the PyMARL codebase that includes additional algorithms and environment support
An Ollama-compatible CLI for Apple's Foundation Models
Kyutai's Speech-To-Text and Text-To-Speech models based on the Delayed Streams Modeling framework.
Official repository of DailyTalk: Spoken Dialogue Dataset for Conversational Text-to-Speech, ICASSP 2023
Mirage Persistent Kernel: Compiling LLMs into a MegaKernel
Companion code to the 2020 AISTATS paper: ''Beyond exploding and vanishing gradients: analysing RNN training using attractors and smoothness"