Skip to content

@pytorch-labs PyTorch Labs

Change the repository type filter

All

    Repositories list

    16 repositories

    • tritonbench

      Public
      Tritonbench is a collection of PyTorch custom operators with example inputs to measure their performance.
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •3•21•7•4•Updated Nov 29, 2024Nov 29, 2024

    • torchft

      Public
      PyTorch per step fault tolerance (actively under development)
      trainingfault-tolerancemltorchpytorch
      Python
      •
      Other
      •1•7•0•1•Updated Nov 28, 2024Nov 28, 2024

    • attention-gym

      Public
      Helpful tools and examples for working with flex-attention
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •24•491•40•1•Updated Nov 24, 2024Nov 24, 2024

    • ethos-u-core-driver-mirror

      Public
      This is a mirror of https//git.mlplatform.org/ml/ethos-u/ethos-u-core-driver.git/
      C++
      •
      Apache License 2.0
      •0•0•0•2•Updated Nov 21, 2024Nov 21, 2024

    • tosa_serialization_lib-mirror

      Public
      This is a mirror of https//git.mlplatform.org/tosa/serialization_lib.git/
      C++
      •
      Apache License 2.0
      •0•0•0•2•Updated Nov 21, 2024Nov 21, 2024

    • torchvision-extra-decoders

      Public
      An extension of `torchvision` for decoding AVIF and HEIC images.
      C++
      •
      GNU Lesser General Public License v2.1
      •1•3•0•2•Updated Nov 21, 2024Nov 21, 2024

    • torchfix

      Public
      TorchFix - a linter for PyTorch-using code with autofix support
      pythonstatic-code-analysislinterstatic-analysispytorchflake8hacktoberfestflake8-plugin
      Python
      •
      Other
      •16•105•10•4•Updated Nov 15, 2024Nov 15, 2024

    • test-infra

      Public
      Other
      •1•0•0•0•Updated Nov 13, 2024Nov 13, 2024

    • applied-ai

      Public
      Applied AI experiments and examples for PyTorch
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •14•172•8•1•Updated Oct 31, 2024Oct 31, 2024

    • LeanRL

      Public
      LeanRL is a fork of CleanRL, where selected PyTorch scripts optimized for performance using compile and cudagraphs.
      Python
      •
      Other
      •17•466•6•1•Updated Oct 26, 2024Oct 26, 2024

    • gpt-fast

      Public
      Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •515•5.7k•66•42•Updated Oct 18, 2024Oct 18, 2024

    • segment-anything-fast

      Public
      A batched offline inference oriented version of segment-anything
      Python
      •
      Apache License 2.0
      •72•1.2k•27•5•Updated Sep 13, 2024Sep 13, 2024

    • FACTO

      Public
      Framework for Algorithmic Correctness Testing of Operators
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •1•1•0•0•Updated Aug 27, 2024Aug 27, 2024

    • superblock

      Public archive
      A block oriented training approach for inference time optimization.
      Python
      •
      MIT License
      •3•30•0•1•Updated Aug 19, 2024Aug 19, 2024

    • float8_experimental

      Public archive
      This repository contains the experimental PyTorch native float8 training UX
      Python
      •
      BSD 3-Clause "New" or "Revised" License
      •20•212•0•22•Updated Aug 1, 2024Aug 1, 2024

    • triton-cpu

      Public
      An experimental CPU backend for Triton (https//github.com/openai/triton)
      C++
      •
      MIT License
      •2•35•1•0•Updated May 13, 2024May 13, 2024