Minimal and highly hackable implementation of Looped Transformers with GPT

Minimal Claude Code alternative. Single Python file, zero dependencies, ~250 lines.

A highly compressive and high-quality neural audio codec for speech models.

A Norwegian Language Understanding and Generation Evaluation Benchmark

The official repository for AdaMuon

Fast & memory efficient hashtable based on robin hood hashing for C++11/14/17/20

Trully flash implementation of DeBERTa disentangled attention mechanism.

Efficient optimizers

supporting pytorch FSDP for optimizers

Teaching transformers to play chess

Official implementation of "GPT or BERT: why not both?"

Fast and accurate language identifier

Truly independent web browser

Official implementation of "BERTs are Generative In-Context Learners"

Implementation of the paper "Compositional Generalization with Grounded Language Models", ACL 2024 Findings

Highlight errors in a bib file: missing URLs, capitalization protection, etc

Suplementary code for the NORA large language models

Scripts and documentation on scaling large language model training on the LUMI supercomputer

Enabling easy statistical significance testing for deep neural networks.

LTG-Bert

PyTorch interface for TrueGrad Optimizers

Implementation of the Adan (ADAptive Nesterov momentum algorithm) Optimizer in Pytorch