- All languages
- Ada
- Arduino
- Assembly
- Astro
- AutoIt
- BQN
- C
- C#
- C++
- COBOL
- CSS
- Clojure
- CoffeeScript
- Common Lisp
- Coq
- Crystal
- Cuda
- Cython
- Dhall
- Dockerfile
- Elixir
- Emacs Lisp
- Erlang
- F#
- F*
- Fennel
- Fortran
- FreeMarker
- Futhark
- GAMS
- Go
- Groovy
- HCL
- HTML
- Haskell
- Haxe
- Janet
- Java
- JavaScript
- Jinja
- Julia
- Jupyter Notebook
- Kotlin
- Lean
- Lua
- MATLAB
- MDX
- Makefile
- Markdown
- Mercury
- MiniZinc
- MoonScript
- Nim
- Nunjucks
- OCaml
- Objective-C
- OpenSCAD
- PHP
- Pascal
- Perl
- Perl 6
- Processing
- Prolog
- PureScript
- Python
- R
- Racket
- Raku
- Reason
- Roff
- Ruby
- Rust
- SCSS
- Sass
- Scala
- Scheme
- Shell
- Shen
- Smalltalk
- Stan
- Standard ML
- Svelte
- Swift
- TSQL
- Tcl
- TeX
- Toit
- TypeScript
- VHDL
- Vala
- Vim Script
- Vue
- WebAssembly
- Wren
- Zig
Starred repositories
Generate, parse, validate, error-correct and present IBAN (and IIBAN) bank account information in PHP.
Create ACH, SEPA, FedNow, SWIFT, RTP payment initiations and process bank statements. Used in production to move hundreds of thousands of dollars a month.
Official code implementation of General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Model
Searches for lost YouTube videos in archives
📚 Process PDFs, Word documents and more with spaCy
Repo for "LoLCATs: On Low-Rank Linearizing of Large Language Models"
Agentless🐱: an agentless approach to automatically solve software development problems
Build GUI for your Python program with JavaScript, HTML, and CSS
On the Theoretical Limitations of Embedding-Based Retrieval
A library of reinforcement learning components and agents
This dataset code generates mathematical question and answer pairs, from a range of question types at roughly school-level difficulty.
Code to our ICML 2025 Paper "Calibrated Language Models and How to Find Them with Label Smoothing"
Code for Paper (Preserving Diversity in Supervised Fine-tuning of Large Language Models)
Deep Reinforcement Learning for Continuous Control in PyTorch
A general model-free off-policy actor-critic implementation. Continuous and Discrete Soft Actor-Critic with multimodal observations, data augmentation, offline learning and behavioral cloning.
Advantage-Filtered Behavioral Cloning for Offline Continuous Control
Pocket Flow: 100-line LLM framework. Let Agents build Agents!
The Hunspell binding for NodeJS that exposes as much of Hunspell as possible and also adds new features. Hunspell is a first class spellcheck library used by Google, Apple, and Mozilla.
Simple text proofreader based on 'write-good' (hemingway-app-like suggestions) and 'nodehun' (spelling).
Democratizing Reinforcement Learning for LLMs
A series of technical report on Slow Thinking with LLM
Write programs like message passing graphs and get parallelism for free. Statically typed and compiled to machine code!