Yerzat Dulat

tinker-cookbook. Post-training with Tinker

verifiers. Our library for RL environments + evals

ART. Agent Reinforcement Trainer: train multi-step agents for real-world tasks using GRPO. Give your agents on-the-job training. Reinforcement learning for Qwen3.6, GPT-OSS, Llama, and more!

understand-r1-zero. Understanding R1-Zero-Like Training: A Critical Perspective

DeepEP. DeepEP: an efficient expert-parallel communication library

TinyZero. Minimal reproduction of DeepSeek R1-Zero

al-folio. A beautiful, simple, clean, and responsive Jekyll theme for academics

FinRL. FinRL®: Financial Reinforcement Learning. 🔥

sglang. SGLang is a high-performance serving framework for large language models and multimodal models.

aquakv. Jupyter Notebook

dictionary_learning. Python

switti. (CVPR 2025) Switti: Designing Scale-Wise Transformers for Text-to-Image Synthesis

Regional-Prompting-FLUX. Training-free Regional Prompting for Diffusion Transformers 🔥

llama3_interpretability_sae. A complete end-to-end pipeline for LLM interpretability with sparse autoencoders (SAEs) using Llama 3.2, written in pure PyTorch and fully reproducible.

imitation. Clean PyTorch implementations of imitation and reward learning algorithms

nnsight. The nnsight package enables interpreting and manipulating the internals of deep learned models.

Kolors. Kolors Team

KnowledgeEditingPapers. Must-read Papers on Knowledge Editing for Large Language Models.

CFCQL. Code for NeurIPS2023 accepted paper: Counterfactual Conservative Q Learning for Offline Multi-agent Reinforcement Learning.

V-Express. V-Express aims to generate a talking head video under the control of a reference image, an audio, and a sequence of V-Kps images.

kernl. Kernl lets you run PyTorch transformer models several times faster on GPU with a single line of code, and is designed to be easily hackable.

torchtitan. A PyTorch native platform for training generative AI models

ELLA. ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment

python-horror-show. Strange and odd python snippets explained

vit-vqgan-jax. Jax implementation of VIT-VQGAN

DSTA. This is the code of our paper "Video-Based Human Pose Regression via Decoupled Space-Time Aggregation".

LFM. Official PyTorch implementation of the paper: Flow Matching in Latent Space

scenic. Scenic: A Jax Library for Computer Vision Research and Beyond

micrograd-cuda. Python

fast-llm-security-guardrails. The fastest Trust Layer for AI Agents

ring-attention. ring-attention experiments

ringattention. Large Context Attention

ring-flash-attention. Ring attention implementation with flash attention

FastCkpt. Python package for rematerialization-aware gradient checkpointing

LightSeq. Official repository for DistFlashAttn: Distributed Memory-efficient Attention for Long-context LLMs Training

StreamDiffusion. StreamDiffusion: A Pipeline-Level Solution for Real-Time Interactive Generation

CVQ-VAE. [ICCV 2023] Online Clustered Codebook

Pearl. A Production-ready Reinforcement Learning AI Agent Library brought by the Applied Reinforcement Learning team at Meta.

fmengine. Utilities for Training Very Large Models

AgileRL. Streamlining reinforcement learning with RLOps. State-of-the-art RL algorithms and tools, with 10x faster training through evolutionary hyperparameter optimization.

go-netstat. A netstat implementation written in Go

warp-drive. Extremely Fast End-to-End Deep Multi-Agent Reinforcement Learning Framework on a GPU (JMLR 2022)

evol-instruct.

scope-rl. SCOPE-RL: A python library for offline reinforcement learning, off-policy evaluation, and selection

RRHF. [NIPS2023] RRHF & Wombat

awesome-chatgpt-dataset. Unlock the Power of LLM: Explore These Datasets to Train Your Own ChatGPT!

ralf. Python

SuperAGI. <⚡️> SuperAGI - A dev-first open source autonomous AI agent framework. Enabling developers to build, manage & run useful autonomous agents quickly and reliably.

chain-of-thought-hub. Benchmarking large language models' complex reasoning ability with chain-of-thought prompting

prompt-to-prompt. Jupyter Notebook

llm-cookbook. 面向开发者的 LLM 入门教程，吴恩达大模型系列课程中文版

openai-cookbook. Examples and guides for using the OpenAI API

Prompt-Engineering-Guide. 🐙 Guides, papers, lessons, notebooks and resources for prompt engineering, context engineering, RAG, and AI Agents.

awesome-langchain. 😎 Awesome list of tools and projects with the awesome LangChain framework

llm-numbers. Numbers every LLM developer should know

arith. Demo on how to build parsers with Alex and Happy

Prompt-Diffusion. Official PyTorch implementation of the paper "In-Context Learning Unlocked for Diffusion Models"

LLM-VM. irresponsible innovation. Try now at https://chat.dev/

RasaGPT. 💬 RasaGPT is the first headless LLM chatbot platform built on top of Rasa and Langchain. Built w/ Rasa, FastAPI, Langchain, LlamaIndex, SQLModel, pgvector, ngrok, telegram

PaLM. An open-source implementation of Google's PaLM models

RWKV-LM. RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RNN and transformer - great performance, linear time, constant space (no kv-cache), fast training, infinite ctx_len, and free sentence embedding.

mlops-zoomcamp. Free MLOps course from DataTalks.Club

superagent. Superagent protects your AI applications against prompt injections, data leaks, and harmful outputs. Embed safety directly into your app and prove compliance to your customers.

GPTeam. GPTeam: An open-source multi-agent simulation

gpt-neox. An implementation of model parallel autoregressive transformers on GPUs, based on the Megatron and DeepSpeed libraries

guidance. A guidance language for controlling large language models.

developer. the first library to let you embed a developer agent in your own app!

dify. Production-ready platform for agentic workflow development.

LocalAI. LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.

quivr. Opiniated RAG for integrating GenAI in your apps 🧠 Focus on your product rather than the RAG. Easy integration in existing products with customisation! Any LLM: GPT4, Groq, Llama. Any Vectorstore: PGVector, Faiss. Any Files. Anyway you want.

E2B. Open-source, secure environment with real-world tools for enterprise-grade agents.

progen. Official release of the ProGen models

langflow. Langflow is a powerful tool for building and deploying AI-powered agents and workflows.

llm-foundry. LLM training code for Databricks foundation models

sol-did. Specification for the did:sol method

zkdrops. Private airdrops for Ethereum.

chainlink-solana-demo. Showing how to deploy a Solana program using Chainlink Price Feeds

neodyme-breakpoint-workshop. Solana Security Workshop for Breakpoint by Neodyme

svoice. We provide a PyTorch implementation of the paper Voice Separation with an Unknown Number of Multiple Speakers In which, we present a new method for separating a mixed audio sequence, in which multiple voices speak simultaneously. The new method employs gated neural networks that are trained to separate the voices at multiple processing steps, while maintaining the speaker in each output channel fixed. A different model is trained for every number of possible speakers, and the model with the largest number of speakers is employed to select the actual number of speakers in a given sample. Our method greatly outperforms the current state of the art, which, as we show, is not competitive for more than two speakers.

owncast. Take control over your live stream video by running it yourself. Streaming + chat out of the box.

nio. 💬 Nio is an upcoming matrix client for iOS.

open-source-ios-apps. :iphone: Collaborative List of Open-Source iOS Apps

CEGIS. Counter-example guided inductive synthesis (CEGIS) implementation for the SMT solver Z3 by Microsoft Research

pylo2. Python wrapper around several Prolog engines. Hoping to make symbolic AI a part of standard AI toolkit.

engineering-sw-hw-model-checking-letures. Lectures in MIPT: intro to engineering software using model-checking.

arq. Fast job queuing and RPC in python with asyncio and redis.

hydra. Hydra is a framework for elegantly configuring complex applications

differential-datalog. DDlog is a programming language for incremental computation. It is well suited for writing programs that continuously update their output in response to input changes. A DDlog programmer does not write incremental algorithms; instead they specify the desired input-output mapping in a declarative manner.

aws-lambda-haskell-runtime. ⚡Haskell runtime for AWS Lambda

s2i-lisp. Source-to-Image builder for Common Lisp applications on OpenShift

react-redux-links. Curated tutorial and resource links I've collected on React, Redux, ES6, and more

RecordRTC. RecordRTC is WebRTC JavaScript library for audio/video as well as screen activity recording. It supports Chrome, Firefox, Opera, Android, and Microsoft Edge. Platforms: Linux, Mac and Windows.

simple-peer. 📡 Simple WebRTC video, voice, and data channels

btfs. A bittorrent filesystem based on FUSE.

web-audio-recorder-js. .wav/.ogg/.mp3 recorder with Web Audio API

janus-gateway. Janus WebRTC Server

coturn. coturn TURN server project

acme. A library of reinforcement learning components and agents

torchlambda. Lightweight tool to deploy PyTorch models to AWS Lambda

postbird. Open source PostgreSQL GUI client for macOS, Linux and Windows

leveldb. LevelDB is a fast key-value storage library written at Google that provides an ordered mapping from string keys to string values.

ideas. :rocket: Ideas for everyone under a CC licence. Feel free to use. I'll send you a postcard if you build anything on this list.

sod. An Embedded Computer Vision & Machine Learning Library (CPU Optimized & IoT Capable)

textshot. Python tool for grabbing text via screenshot

boolector. A Satisfiability Modulo Theories (SMT) solver for the theories of fixed-size bit-vectors, arrays and uninterpreted functions.

Marketing-for-Engineers. A curated collection of marketing articles & tools to grow your product.

Speaker-independent-emotional-voice-conversion-based-on-conditional-VAW-GAN-and-CWT. This is the implementation of our Interspeech 2020 paper "Converting anyone's emotion: towards speaker-independent emotional voice conversion".

texts. Articles about various things

gotestwaf. An open-source project in Golang to asess different API Security tools and WAF for detection logic and bypasses

tla-rust. writing correct lock-free and distributed stateful systems in Rust, assisted by TLA+

practical-tla-plus. Source Code for 'Practical TLA+' by Hillel Wayne

bayou. System for synthesizing Java API idioms, powered by Neural Sketch Learning

HaVSA. HaVSA (Have-Saa) is a Haskell implementation of the Version Space Algebra Machine Learning technique described by Tessa Lau.

learn2branch. Exact Combinatorial Optimization with Graph Convolutional Neural Networks (NeurIPS 2019)

guiding-synthesizers. Guiding Program Synthesis by Learning to Generate Examples

neural_sketch. neural sketch project, currently in generative regex, list transformation (deepcoder), and text editing (robustfill) domains

rulesynthesis. Code for "Learning Compositional Rules via Neural Program Synthesis"

CoT. (Beta Version!) Experiment Code for Paper ``CoT: Cooperative Training for Generative Modeling of Discrete Data''

cs323-notes. Course notes for CS323: Automated Reasoning

asp-game-strategies. Learning game strategies in ASP (clingo) using minimax and inductive logic programming (ILASP) :clubs: :game_die:

learn-physics. A library of functions for vector calculus, calculation of electric field, electric flux, magnetic field, and other quantities in mechanics and electromagnetic theory.

logic-life-search. Cellular automata search program

mrcl. Code for the NeurIPS19 paper "Meta-Learning Representations for Continual Learning"

ANML. A Neuromodulated Meta-Learning algorithm

hexlite. Python-based HEX implementation for a fragment of the HEX language and a subset of features.

pytorch-lightning. Pretrain, finetune ANY AI model of ANY size on 1 or 10,000+ GPUs with zero code changes.

causal_inference_python_code. Python code for part 2 of the book Causal Inference: What If, by Miguel Hernán and James Robins

Upside-Down-Reinforcement-Learning. Upside-Down Reinforcement Learning (⅂ꓤ) implementation in PyTorch. Based on the paper published by Jürgen Schmidhuber.

system-design-primer. Learn how to design large-scale systems. Prep for the system design interview. Includes Anki flashcards.

clutrr. Diagnostic benchmark suite to explicitly test logical relational reasoning on natural language

InvariantRiskMinimization. PyTorch code to run synthetic experiments.

meta_seq2seq. Compositional generalization through meta sequence-to-sequence learning

Polygames. The project is a platform of zero learning with a library of games.

EGG. EGG: Emergence of lanGuage in Games

pysequitur. Python Sequitur (Nevill-Manning) algorithm

Re-Pair-VF. C

Re-Pair. A space-efficient compressor based on the Re-Pair grammar

MR-Repair. C

Financial-Models-Numerical-Methods. Collection of notebooks about quantitative finance, with interactive python code.

LanczosNetwork. Lanczos Network, Graph Neural Networks, Deep Graph Convolutional Networks, Deep Learning on Graph Structured Data, QM8 Quantum Chemistry Benchmark, ICLR 2019

elpi. Embeddable Lambda Prolog Interpreter

gpt-2-output-dataset. Dataset of GPT-2 outputs for research in detection, biases, and more

pyxel. A retro game engine for Python

multiagent-gail. Python

stacknn-core. Pip-installable differentiable stacks in PyTorch!

hasktorch. Tensors and neural networks in Haskell

higher. higher is a pytorch library allowing users to obtain higher order gradients over losses spanning training loops rather than individual training steps.

node. Neural Oblivious Decision Ensembles for Deep Learning on Tabular Data

imo-coq. International Mathematics Olympiad in Coq

pneumothorax-segmentation. SIIM-ACR Pneumothorax Segmentation first place solution

industry-machine-learning. A curated list of applied machine learning and data science notebooks and libraries across different industries (by @firmai)

Apollo-11. Original Apollo 11 Guidance Computer (AGC) source code for the command and lunar modules.

adaptive-span. Transformer training code for sequential tasks

DXNN. Topology and Parameter Evolving Universal Learning Network Platform

STALINGRAD. Stalin∇, a brutally optimizing compiler for the VLAD language

awesome-haskell-deep-learning. In the tradition of "awesome" (curated) lists, this is a list of references and code for doing deep learning in Haskell.

NumEr. Numeric Erlang - vector and matrix operations with CUDA. Heavily inspired by Pteracuda - https://github.com/kevsmith/pteracuda

AlphaNPI. Adapting the AlphaZero algorithm to remove the need of execution traces to train NPI.

xlnet. XLNet: Generalized Autoregressive Pretraining for Language Understanding

differentiable_sorting. Differentiable bitonic sorting

neuralkanren. Neural Guided Constraint Logic Programming for Program Synthesis

e-SNLI. Python

dl2. DL2 is a framework that allows training neural networks with logical constraints over numerical values in the network (e.g. inputs, outputs, weights) and to query networks for inputs fulfilling a logical formula.

kaggle-imaterialist. The First Place Solution of Kaggle iMaterialist (Fashion) 2019 at FGVC6

gym-sokoban. Sokoban environment for OpenAI Gym

clpz. Constraint Logic Programming over Integers

AdaptiveNeuralTrees. Adaptive Neural Trees

PlaNet. Deep Planning Network: Control from pixels by latent planning with learned dynamics

SATNet. Bridging deep learning and logical reasoning using a differentiable satisfiability solver.

graphics. TensorFlow Graphics: Differentiable Graphics Layers for TensorFlow

MAAC. Code for "Actor-Attention-Critic for Multi-Agent Reinforcement Learning" ICML 2019

sequential_social_dilemma_games. Repo for reproduction of sequential social dilemmas

multiagent. Using RLLib and PycoLab to explore intelligent cooperative behavior in sequential social dilemmas

neural-logic-machines. Implementation for the Neural Logic Machines (NLM).

craft-env. Python

AmpliGraph. Python library for Representation Learning on Knowledge Graphs https://docs.ampligraph.org

KG-DQN. Python

structural-probes. Codebase for testing whether hidden states of neural networks encode discrete structures.

ConvE. Convolutional 2D Knowledge Graph Embeddings resources

MinAtar. Python

ConvKB. A Novel Embedding Model for Knowledge Base Completion Based on Convolutional Neural Network (NAACL 2018) (Pytorch and Tensorflow)

CapsGNN. A PyTorch implementation of "Capsule Graph Neural Network" (ICLR 2019).

lola. Code release for Learning with Opponent-Learning Awareness and variations.

GraphRNN. Python

Ape-X. PyTorch Implementation of Distributed Prioritized Experience Replay(Ape-X)

SimGNN. A PyTorch implementation of "SimGNN: A Neural Network Approach to Fast Graph Similarity Computation" (WSDM 2019).

meta-dataset. A dataset of datasets for learning to learn from few examples

APPNP. A PyTorch implementation of "Predict then Propagate: Graph Neural Networks meet Personalized PageRank" (ICLR 2019).

spreadingvectors. Open source implementation of "Spreading Vectors for Similarity Search"

awesome-graph-classification. A collection of important graph embedding, classification and representation learning papers with implementations.

SinkhornAutoDiff. Toolbox to integrate optimal transport loss functions using automatic differentiation and Sinkhorn's algorithm

SEAL. SEAL (learning from Subgraphs, Embeddings, and Attributes for Link prediction). "M. Zhang, Y. Chen, Link Prediction Based on Graph Neural Networks, NeurIPS 2018 spotlight".

SimplE. Implementation of SimplE Embedding for Link Prediction in Knowledge Graphs in PyTorch

AttentionWalk. A PyTorch Implementation of "Watch Your Step: Learning Node Embeddings via Graph Attention" (NeurIPS 2018).

One-shot-Relational-Learning. Code for One-shot Relational Learning for Knowledge Graphs (EMNLP18)

MixHop-and-N-GCN. An implementation of "MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing" (ICML 2019).

XWorld. A C++/Python simulator package for reinforcement learning

abstract-reasoning-matrices. Progressive matrices dataset, as described in: Measuring abstract reasoning in neural networks (Barrett*, Hill*, Santoro*, Morcos, Lillicrap), ICML2018

modular-assemblies. [NeurIPS 2019] Code for the paper "Learning to Control Self-Assembling Morphologies: A Study of Generalization via Modularity"

gpt-2. Code for the paper "Language Models are Unsupervised Multitask Learners"