Example repository for custom C++/CUDA operators for TorchScript

Overview

Custom TorchScript Operators Example

This repository contains examples for writing, compiling and using custom TorchScript operators. See here for the accompanying tutorial.

Contents

There a few monuments in this repository you can visit. They are described in context in the tutorial, which you are encouraged to read. These monuments are:

  • example_app/warp_perspective/op.cpp: The custom operator implementation,
  • example_app/main.cpp: An example application that loads and executes a serialized TorchScript model, which uses the custom operator, in C++,
  • script.py: Example of using the custom operator in a scripted model,
  • trace.py: Example of using the custom operator in a traced model,
  • eager.py: Example of using the custom operator in vanilla eager PyTorch,
  • load.py: Example of using torch.utils.cpp_extension.load to build the custom operator,
  • load.py: Example of using torch.utils.cpp_extension.load_inline to build the custom operator,
  • setup.py: Example of using setuptools to build the custom operator,
  • test_setup.py: Example of using the custom operator built using setup.py.

To execute the C++ application, first run script.py to serialize a TorchScript model to a file called example.pt, then pass that file to the example_app/build/example_app binary.

Setup

For the smoothest experience when trying out these examples, we recommend building a docker container from this repository's Dockerfile. This will give you a clean, isolated Ubuntu Linux environment in which we guarantee everything to work perfectly. These steps should get you started:

$ git clone https://github.com/pytorch/extension-script

$ cd extension-script

$ docker build -t extension-script .

$ docker run -v $PWD:/home -it extension-script

$ [email protected]:/home# source /activate # Activate the Conda environment

$ cd example_app && mkdir build && cd build

$ cmake -DCMAKE_PREFIX_PATH=/libtorch ..
-- The C compiler identification is GNU 5.4.0
-- The CXX compiler identification is GNU 5.4.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Found torch: /libtorch/lib/libtorch.so
-- Configuring done
-- Generating done
-- Build files have been written to: /home/example_app/build

$ make -j
Scanning dependencies of target warp_perspective
[ 25%] Building CXX object warp_perspective/CMakeFiles/warp_perspective.dir/op.cpp.o
[ 50%] Linking CXX shared library libwarp_perspective.so
[ 50%] Built target warp_perspective
Scanning dependencies of target example_app
[ 75%] Building CXX object CMakeFiles/example_app.dir/main.cpp.o
[100%] Linking CXX executable example_app
[100%] Built target example_app

This will create a shared library under /home/example_app/build/warp_perspective/libwarp_perspective.so containing the custom operator defined in example_app/warp_perspective/op.cpp. Then, you can run the examples, e.g.:

(base) [email protected]:/home# python script.py
graph(%x.1 : Dynamic
      %y : Dynamic) {
  %20 : int = prim::Constant[value=1]()
  %16 : int[] = prim::Constant[value=[0, -1]]()
  %14 : int = prim::Constant[value=6]()
  %2 : int = prim::Constant[value=0]()
  %7 : int = prim::Constant[value=42]()
  %z.1 : int = prim::Constant[value=5]()
  %z.2 : int = prim::Constant[value=10]()
  %13 : int = prim::Constant[value=3]()
  %4 : Dynamic = aten::select(%x.1, %2, %2)
  %6 : Dynamic = aten::select(%4, %2, %2)
  %8 : Dynamic = aten::eq(%6, %7)
  %9 : bool = prim::TensorToBool(%8)
  %z : int = prim::If(%9)
    block0() {
      -> (%z.1)
    }
    block1() {
      -> (%z.2)
    }
  %17 : Dynamic = aten::eye(%13, %14, %2, %16)
  %x : Dynamic = my_ops::warp_perspective(%x.1, %17)
  %19 : Dynamic = aten::matmul(%x, %y)
  %21 : Dynamic = aten::add(%19, %z, %20)
  return (%21);
}

tensor([[11.6196, 12.0056, 11.6122, 12.9298,  7.0649],
        [ 8.5063,  9.0621,  9.9925,  6.3741,  8.9668],
        [12.5898,  6.5872,  8.1511, 10.0806, 11.9829],
        [ 4.9142, 11.6614, 15.7161, 17.0538, 11.7243],
        [10.0000, 10.0000, 10.0000, 10.0000, 10.0000],
        [10.0000, 10.0000, 10.0000, 10.0000, 10.0000],
        [10.0000, 10.0000, 10.0000, 10.0000, 10.0000],
        [10.0000, 10.0000, 10.0000, 10.0000, 10.0000]])
This code provides a PyTorch implementation for OTTER (Optimal Transport distillation for Efficient zero-shot Recognition), as described in the paper.

Data Efficient Language-Supervised Zero-Shot Recognition with Optimal Transport Distillation This repository contains PyTorch evaluation code, trainin

Meta Research 45 Dec 20, 2022
QueryInst: Parallelly Supervised Mask Query for Instance Segmentation

QueryInst is a simple and effective query based instance segmentation method driven by parallel supervision on dynamic mask heads, which outperforms previous arts in terms of both accuracy and speed.

Hust Visual Learning Team 386 Jan 08, 2023
This repository contains various models targetting multimodal representation learning, multimodal fusion for downstream tasks such as multimodal sentiment analysis.

Multimodal Deep Learning 🎆 🎆 🎆 Announcing the multimodal deep learning repository that contains implementation of various deep learning-based model

Deep Cognition and Language Research (DeCLaRe) Lab 398 Dec 30, 2022
A tutorial on DataFrames.jl prepared for JuliaCon2021

JuliaCon2021 DataFrames.jl Tutorial This is a tutorial on DataFrames.jl prepared for JuliaCon2021. A video recording of the tutorial is available here

Bogumił Kamiński 106 Jan 09, 2023
Unofficial Pytorch Lightning implementation of Contrastive Syn-to-Real Generalization (ICLR, 2021)

Unofficial Pytorch Lightning implementation of Contrastive Syn-to-Real Generalization (ICLR, 2021)

Gyeongjae Choi 17 Sep 23, 2021
Official repository for: Continuous Control With Ensemble DeepDeterministic Policy Gradients

Continuous Control With Ensemble Deep Deterministic Policy Gradients This repository is the official implementation of Continuous Control With Ensembl

4 Dec 06, 2021
This repo contains the code for the paper "Efficient hierarchical Bayesian inference for spatio-temporal regression models in neuroimaging" that has been accepted to NeurIPS 2021.

Dugh-NeurIPS-2021 This repo contains the code for the paper "Efficient hierarchical Bayesian inference for spatio-temporal regression models in neuroi

Ali Hashemi 5 Jul 12, 2022
An Efficient Training Approach for Very Large Scale Face Recognition or F²C for simplicity.

Fast Face Classification (F²C) This is the code of our paper An Efficient Training Approach for Very Large Scale Face Recognition or F²C for simplicit

33 Jun 27, 2021
Automatically align face images 🙃→🙂. Can also do windowing and warping.

Automatic Face Alignment (AFA) Carl M. Gaspar & Oliver G.B. Garrod You have lots of photos of faces like this: But you want to line up all of the face

Carl Michael Gaspar 15 Dec 12, 2022
Losslandscapetaxonomy - Taxonomizing local versus global structure in neural network loss landscapes

Taxonomizing local versus global structure in neural network loss landscapes Int

Yaoqing Yang 8 Dec 30, 2022
[BMVC 2021] Official PyTorch Implementation of Self-supervised learning of Image Scale and Orientation Estimation

Self-Supervised Learning of Image Scale and Orientation Estimation (BMVC 2021) This is the official implementation of the paper "Self-Supervised Learn

Jongmin Lee 17 Nov 10, 2022
This repository contains the official implementation code of the paper Transformer-based Feature Reconstruction Network for Robust Multimodal Sentiment Analysis

This repository contains the official implementation code of the paper Transformer-based Feature Reconstruction Network for Robust Multimodal Sentiment Analysis, accepted at ACMMM 2021.

Ziqi Yuan 10 Sep 30, 2022
TipToiDog - Tip Toi Dog With Python

TipToiDog Was ist dieses Projekt? Meine 5-jährige Tochter spielt sehr gerne das

1 Feb 07, 2022
Optimizing Deeper Transformers on Small Datasets

DT-Fixup Optimizing Deeper Transformers on Small Datasets Paper published in ACL 2021: arXiv Detailed instructions to replicate our results in the pap

16 Nov 14, 2022
Checking fibonacci - Generating the Fibonacci sequence is a classic recursive problem

Fibonaaci Series Generating the Fibonacci sequence is a classic recursive proble

Moureen Caroline O 1 Feb 15, 2022
Python package for visualizing the loss landscape of parameterized quantum algorithms.

orqviz A Python package for easily visualizing the loss landscape of Variational Quantum Algorithms by Zapata Computing Inc. orqviz provides a collect

Zapata Computing, Inc. 75 Dec 30, 2022
Implementation of QuickDraw - an online game developed by Google, combined with AirGesture - a simple gesture recognition application

QuickDraw - AirGesture Introduction Here is my python source code for QuickDraw - an online game developed by google, combined with AirGesture - a sim

Viet Nguyen 89 Dec 18, 2022
[CVPRW 21] "BNN - BN = ? Training Binary Neural Networks without Batch Normalization", Tianlong Chen, Zhenyu Zhang, Xu Ouyang, Zechun Liu, Zhiqiang Shen, Zhangyang Wang

BNN - BN = ? Training Binary Neural Networks without Batch Normalization Codes for this paper BNN - BN = ? Training Binary Neural Networks without Bat

VITA 40 Dec 30, 2022
This is the workbook I created while I was studying for the Qiskit Associate Developer exam. I hope this becomes useful to others as it was for me :)

A Workbook for the Qiskit Developer Certification Exam Hello everyone! This is Bartu, a fellow Qiskitter. I have recently taken the Certification exam

Bartu Bisgin 66 Dec 10, 2022
Jittor Medical Segmentation Lib -- The assignment of Pattern Recognition course (2021 Spring) in Tsinghua University

THU模式识别2021春 -- Jittor 医学图像分割 模型列表 本仓库收录了课程作业中同学们采用jittor框架实现的如下模型: UNet SegNet DeepLab V2 DANet EANet HarDNet及其改动HarDNet_alter PSPNet OCNet OCRNet DL

48 Dec 26, 2022