whatsinfinitum
Computer Science Ph.D., AI Researcher, founder, entrepreneur. I'm a tech lead at Copilot Model team working on Copilot Custom Models for GitHub Enterprise.
@github Tokyo, Japan
Pinned Repositories
bullet3
Bullet Physics SDK: real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning etc.
FastChat
An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.
LLaVA
[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.
minbpe
Minimal, clean, code for the Byte Pair Encoding (BPE) algorithm commonly used in LLM tokenization.
ml-ferret
mlx
MLX: An array framework for Apple silicon
public-test
to test copilot on windows
RWKV-LM
RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding.
tabby
Self-hosted AI coding assistant
whatsinfinitum
Config files for my GitHub profile.
whatsinfinitum's Repositories
whatsinfinitum/whatsinfinitum
Config files for my GitHub profile.
whatsinfinitum/bullet3
Bullet Physics SDK: real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning etc.
whatsinfinitum/minbpe
Minimal, clean, code for the Byte Pair Encoding (BPE) algorithm commonly used in LLM tokenization.
whatsinfinitum/ml-ferret
whatsinfinitum/mlx
MLX: An array framework for Apple silicon
whatsinfinitum/tabby
Self-hosted AI coding assistant
whatsinfinitum/FastChat
An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.
whatsinfinitum/LLaVA
[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.
whatsinfinitum/public-test
to test copilot on windows
whatsinfinitum/RWKV-LM
RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding.