Skip to content

@HazyResearch HazyResearch

Change the repository type filter

All

    Repositories list

    124 repositories

    • cartridges

      Public
      Storing long contexts in tiny caches with self-study
      Python
      •
      Apache License 2.0
      •7•105•4•0•Updated Jul 21, 2025Jul 21, 2025

    • ThunderKittens-HIP

      Public
      Tile primitives for speedy kernels
      Cuda
      •
      MIT License
      •0•6•0•0•Updated Jul 19, 2025Jul 19, 2025

    • minions

      Public
      Big & Small LLMs working together
      Python
      •
      MIT License
      •121•1.1k•1•11•Updated Jul 18, 2025Jul 18, 2025

    • bwler

      Public
      Official repo for BWLer: Barycentric Weight Layer
      Python
      •
      Other
      •2•17•0•0•Updated Jul 17, 2025Jul 17, 2025

    • ThunderKittens

      Public
      Tile primitives for speedy kernels
      Cuda
      •
      MIT License
      •160•2.5k•40•13•Updated Jul 15, 2025Jul 15, 2025

    • zoology

      Public
      Understand and test language model architectures on synthetic tasks.
      Python
      •
      Apache License 2.0
      •37•221•2•1•Updated Jul 13, 2025Jul 13, 2025

    • scaling-verification

      Public
      Python
      •
      MIT License
      •0•19•0•0•Updated Jun 24, 2025Jun 24, 2025

    • based

      Public
      Code for exploring Based models from "Simple linear attention language models balance the recall-throughput tradeoff"
      Python
      •
      Apache License 2.0
      •17•237•3•0•Updated Jun 6, 2025Jun 6, 2025

    • Megakernels

      Public
      kernels, of the mega variety
      Python
      •
      MIT License
      •22•453•3•1•Updated Jun 2, 2025Jun 2, 2025

    • hyena-dna

      Public
      Official implementation for HyenaDNA, a long-range genomic foundation model built with Hyena
      genomicslanguage-modelsfoundation-models
      Assembly
      •
      Apache License 2.0
      •100•704•32•7•Updated Apr 22, 2025Apr 22, 2025

    • precision-ls

      Public
      Python
      •
      MIT License
      •5•7•0•0•Updated Mar 18, 2025Mar 18, 2025

    • lolcats

      Public
      Repo for "LoLCATs: On Low-Rank Linearizing of Large Language Models"
      Python
      •
      Apache License 2.0
      •25•244•8•0•Updated Jan 31, 2025Jan 31, 2025

    • aioli

      Public
      Aioli: A unified optimization framework for language model data mixing
      Jupyter Notebook
      •
      Apache License 2.0
      •4•27•1•0•Updated Jan 17, 2025Jan 17, 2025

    • flash-fft-conv

      Public
      FlashFFTConv: Efficient Convolutions for Long Sequences with Tensor Cores
      C++
      •
      Apache License 2.0
      •29•323•18•4•Updated Dec 28, 2024Dec 28, 2024

    • m2

      Public
      Repo for "Monarch Mixer: A Simple Sub-Quadratic GEMM-Based Architecture"
      Assembly
      •
      Apache License 2.0
      •42•555•25•2•Updated Dec 28, 2024Dec 28, 2024

    • meerkat

      Public
      Creative interactive views of any dataset.
      data-sciencemlpandasfoundation-modelsmachine-learning
      Python
      •
      Apache License 2.0
      •44•843•8•3•Updated Dec 24, 2024Dec 24, 2024

    • smoothie

      Public
      Jupyter Notebook
      •
      MIT License
      •3•9•0•0•Updated Dec 10, 2024Dec 10, 2024

    • train-tk

      Public
      train with kittens!
      Python
      •7.2k•61•0•0•Updated Oct 25, 2024Oct 25, 2024

    • wonderbread

      Public
      WONDERBREAD benchmark + dataset for BPM tasks
      Jupyter Notebook
      •6•26•1•0•Updated Oct 20, 2024Oct 20, 2024

    • vllm

      Public
      A high-throughput and memory-efficient inference and serving engine for LLMs
      Python
      •
      Apache License 2.0
      •8.8k•1•0•0•Updated Oct 14, 2024Oct 14, 2024

    • eclair-agents

      Public
      Automating enterprise workflows with multimodal agents
      Jupyter Notebook
      •
      Apache License 2.0
      •13•108•0•0•Updated Oct 9, 2024Oct 9, 2024

    • legalbench

      Public
      An open science effort to benchmark legal reasoning in foundation models
      Python
      •68•457•9•6•Updated Aug 25, 2024Aug 25, 2024

    • hgcn

      Public
      Hyperbolic Graph Convolutional Networks in PyTorch.
      Python
      •115•631•20•3•Updated Jul 25, 2024Jul 25, 2024

    • manifest

      Public
      Prompt programming with FMs.
      Python
      •
      Apache License 2.0
      •45•443•6•2•Updated Jul 22, 2024Jul 22, 2024

    • prefix-linear-attention

      Public
      Python
      •2•56•0•0•Updated Jul 9, 2024Jul 9, 2024

    • safari

      Public
      Convolutions for Sequence Modeling
      Assembly
      •
      Apache License 2.0
      •71•893•25•1•Updated Jun 13, 2024Jun 13, 2024

    • based-evaluation-harness

      Public
      A framework for few-shot evaluation of language models.
      Python
      •
      MIT License
      •2.6k•10•0•0•Updated Jun 8, 2024Jun 8, 2024

    • olive-evaluation-harness

      Public
      A framework for few-shot evaluation of language models.
      Python
      •
      MIT License
      •2.6k•8•0•0•Updated Jun 3, 2024Jun 3, 2024

    • axolive

      Public
      Go ahead and axolotl questions
      Python
      •
      Apache License 2.0
      •1.1k•2•0•0•Updated Jun 3, 2024Jun 3, 2024

    • olive-recipes

      Public
      Scripts for fine-tuning Meta Llama3 with composable FSDP & PEFT methods to cover single/multi-node GPUs. Supports default & custom datasets for applications such as summarization and Q&A. Supporting a number of candid inference solutions such as HF TGI, VLLM for local or cloud deployment. Demo apps to showcase Meta Llama3 for WhatsApp & Messenger.
      Jupyter Notebook
      •2.6k•1•0•0•Updated Jun 3, 2024Jun 3, 2024