Implementation of OmniNet, Omnidirectional Representations from Transformers, in Pytorch

Last update: Nov 21, 2022

Overview

Omninet - Pytorch

Implementation of OmniNet, Omnidirectional Representations from Transformers, in Pytorch. The authors propose that we should be attending to all the tokens of the previous layers, leveraging recent efficient attention advances to achieve this goal.

Install

$ pip install omninet-pytorch

Usage

import torch
from omninet_pytorch import Omninet

omninet = Omninet(
    dim = 512,                     # model dimension
    depth = 6,                     # depth
    dim_head = 64,                 # dimension per head
    heads = 8,                     # number of heads
    pool_layer_tokens_every = 3,   # key to this paper - every N layers, omni attend to all tokens of all layers
    attn_dropout = 0.1,            # attention dropout
    ff_dropout = 0.1,              # feedforward dropout
    feature_redraw_interval = 1000 # how often to redraw the projection matrix for omni attention net - Performer
)

x = torch.randn(1, 1024, 512)
mask = torch.ones(1, 1024).bool()

omninet(x, mask = mask) # (1, 1024, 512)

Causal case, just use the class OmninetCausal. At the moment, it isn't faithful to the paper (I am using layer axial attention with layer positional embeddings to draw up information), but will fix this once I rework the linear attention CUDA kernel.

import torch
from omninet_pytorch import OmninetCausal

omninet = OmninetCausal(
    dim = 512,                     # model dimension
    depth = 6,                     # depth
    dim_head = 64,                 # dimension per head
    heads = 8,                     # number of heads
    pool_layer_tokens_every = 3,   # key to this paper - every N layers, omni attend to all tokens of all layers
    attn_dropout = 0.1,            # attention dropout
    ff_dropout = 0.1               # feedforward dropout
)

x = torch.randn(1, 1024, 512)
mask = torch.ones(1, 1024).bool()

omninet(x, mask = mask) # (1, 1024, 512)

Citations

@misc{tay2021omninet,
    title   = {OmniNet: Omnidirectional Representations from Transformers}, 
    author  = {Yi Tay and Mostafa Dehghani and Vamsi Aribandi and Jai Gupta and Philip Pham and Zhen Qin and Dara Bahri and Da-Cheng Juan and Donald Metzler},
    year    = {2021},
    eprint  = {2103.01075},
    archivePrefix = {arXiv},
    primaryClass = {cs.CV}
}

PyTorch implementation for paper StARformer: Transformer with State-Action-Reward Representations.

StARformer This repository contains the PyTorch implementation for our paper titled StARformer: Transformer with State-Action-Reward Representations.

14 Dec 9, 2022

Official PyTorch implementation of BlobGAN: Spatially Disentangled Scene Representations

BlobGAN: Spatially Disentangled Scene Representations Official PyTorch Implementation Paper | Project Page | Video | Interactive Demo BlobGAN.mp4 This

148 Dec 29, 2022

ChatBot-Pytorch - A GPT-2 ChatBot implemented using Pytorch and Huggingface-transformers

ChatBot-Pytorch A GPT-2 ChatBot implemented using Pytorch and Huggingface-transf

42 Dec 9, 2022

Here is the implementation of our paper S2VC: A Framework for Any-to-Any Voice Conversion with Self-Supervised Pretrained Representations.

S2VC Here is the implementation of our paper S2VC: A Framework for Any-to-Any Voice Conversion with Self-Supervised Pretrained Representations. In thi

81 Dec 15, 2022

[ICCV'21] Official implementation for the paper Social NCE: Contrastive Learning of Socially-aware Motion Representations

CrowdNav with Social-NCE This is an official implementation for the paper Social NCE: Contrastive Learning of Socially-aware Motion Representations by

125 Dec 23, 2022

Tensorflow 2 implementation of the paper: Learning and Evaluating Representations for Deep One-class Classification published at ICLR 2021

Deep Representation One-class Classification (DROC). This is not an officially supported Google product. Tensorflow 2 implementation of the paper: Lea

137 Dec 23, 2022

Public Implementation of ChIRo from "Learning 3D Representations of Molecular Chirality with Invariance to Bond Rotations"

Learning 3D Representations of Molecular Chirality with Invariance to Bond Rotations This directory contains the model architectures and experimental

35 Dec 5, 2022

Official implementation of the paper WAV2CLIP: LEARNING ROBUST AUDIO REPRESENTATIONS FROM CLIP

Wav2CLIP 🚧 WIP 🚧 Official implementation of the paper WAV2CLIP: LEARNING ROBUST AUDIO REPRESENTATIONS FROM CLIP 📄 🔗 Ho-Hsiang Wu, Prem Seetharaman

240 Dec 13, 2022

Implementation of the method proposed in the paper "Neural Descriptor Fields: SE(3)-Equivariant Object Representations for Manipulation"

Neural Descriptor Fields (NDF) PyTorch implementation for training continuous 3D neural fields to represent dense correspondence across objects, and u

167 Jan 6, 2023

Implementation of OmniNet, Omnidirectional Representations from Transformers, in Pytorch

Related tags

Overview

Omninet - Pytorch

Install

Usage

Citations

You might also like...

PyTorch implementation for paper StARformer: Transformer with State-Action-Reward Representations.

Official PyTorch implementation of BlobGAN: Spatially Disentangled Scene Representations

ChatBot-Pytorch - A GPT-2 ChatBot implemented using Pytorch and Huggingface-transformers

Here is the implementation of our paper S2VC: A Framework for Any-to-Any Voice Conversion with Self-Supervised Pretrained Representations.

[ICCV'21] Official implementation for the paper Social NCE: Contrastive Learning of Socially-aware Motion Representations

Tensorflow 2 implementation of the paper: Learning and Evaluating Representations for Deep One-class Classification published at ICLR 2021

Public Implementation of ChIRo from "Learning 3D Representations of Molecular Chirality with Invariance to Bond Rotations"

Official implementation of the paper WAV2CLIP: LEARNING ROBUST AUDIO REPRESENTATIONS FROM CLIP

Implementation of the method proposed in the paper "Neural Descriptor Fields: SE(3)-Equivariant Object Representations for Manipulation"

Releases(0.0.6)

0.0.6(Mar 19, 2021)

0.0.5(Mar 2, 2021)

0.0.4(Mar 2, 2021)

0.0.3(Mar 2, 2021)

0.0.2(Mar 2, 2021)

0.0.1(Mar 2, 2021)

Owner

Phil Wang

Official implementation of Representer Point Selection via Local Jacobian Expansion for Post-hoc Classifier Explanation of Deep Neural Networks and Ensemble Models at NeurIPS 2021

JFB: Jacobian-Free Backpropagation for Implicit Models

Efficient Sharpness-aware Minimization for Improved Training of Neural Networks

Speedy Implementation of Instance-based Learning (IBL) agents in Python

Torch implementation of various types of GAN (e.g. DCGAN, ALI, Context-encoder, DiscoGAN, CycleGAN, EBGAN, LSGAN)

Enabling Lightweight Fine-tuning for Pre-trained Language Model Compression based on Matrix Product Operators

A practical ML pipeline for data labeling with experiment tracking using DVC.

SegNet-like Autoencoders in TensorFlow

Generative Query Network (GQN) in PyTorch as described in "Neural Scene Representation and Rendering"

Public repository containing materials used for Feed Forward (FF) Neural Networks article.

Semantic-aware Grad-GAN for Virtual-to-Real Urban Scene Adaption

Cockpit is a visual and statistical debugger specifically designed for deep learning.

A3C LSTM Atari with Pytorch plus A3G design

Scalable and Elastic Deep Reinforcement Learning Using PyTorch. Please star. 🔥

Official implementation of deep-multi-trajectory-based single object tracking (IEEE T-CSVT 2021).

Reducing Information Bottleneck for Weakly Supervised Semantic Segmentation (NeurIPS 2021)

YOLOv7 - Framework Beyond Detection

Reducing Information Bottleneck for Weakly Supervised Semantic Segmentation (NeurIPS 2021)

Computer Vision Paper Reviews with Key Summary of paper, End to End Code Practice and Jupyter Notebook converted papers

Ros2-voiceroid2 - ROS2 wrapper package of VOICEROID2