Implementation of a Transformer that Ponders, using the scheme from the PonderNet paper

Related tags

Deep Learningdeep-learningtransformersartificial-intelligenceadaptive-computation-time
Overview

Ponder(ing) Transformer

Implementation of a Transformer that learns to adapt the number of computational steps it takes depending on the difficulty of the input sequence, using the scheme from the PonderNet paper. Will also try to abstract out a pondering module that can be used with any block that returns an output with the halting probability.

This repository would not have been possible without repeated viewings of Yannic's educational video

Install

$ pip install ponder-transformer

Usage

import torch
from ponder_transformer import PonderTransformer

model = PonderTransformer(
    num_tokens = 20000,
    dim = 512,
    max_seq_len = 512
)

mask = torch.ones(1, 512).bool()

x = torch.randint(0, 20000, (1, 512))
y = torch.randint(0, 20000, (1, 512))

loss = model(x, labels = y, mask = mask)
loss.backward()

Now you can set the model to .eval() mode and it will terminate early when all samples of the batch have emitted a halting signal

import torch
from ponder_transformer import PonderTransformer

model = PonderTransformer(
    num_tokens = 20000,
    dim = 512,
    max_seq_len = 512,
    causal = True
)

x = torch.randint(0, 20000, (2, 512))
mask = torch.ones(2, 512).bool()

model.eval() # setting to eval makes it return the logits as well as the halting indices

logits, layer_indices = model(x,  mask = mask) # (2, 512, 20000), (2)

# layer indices will contain, for each batch element, which layer they exited

Citations

@misc{banino2021pondernet,
    title   = {PonderNet: Learning to Ponder}, 
    author  = {Andrea Banino and Jan Balaguer and Charles Blundell},
    year    = {2021},
    eprint  = {2107.05407},
    archivePrefix = {arXiv},
    primaryClass = {cs.LG}
}
You might also like...
Implementation of the Transformer variant proposed in
Implementation of the Transformer variant proposed in "Transformer Quality in Linear Time"

FLASH - Pytorch Implementation of the Transformer variant proposed in the paper Transformer Quality in Linear Time Install $ pip install FLASH-pytorch

Phil Wang 209 Dec 28, 2022
Third party Pytorch implement of Image Processing Transformer (Pre-Trained Image Processing Transformer arXiv:2012.00364v2)

ImageProcessingTransformer Third party Pytorch implement of Image Processing Transformer (Pre-Trained Image Processing Transformer arXiv:2012.00364v2)

null 61 Jan 1, 2023
Episodic Transformer (E.T.) is a novel attention-based architecture for vision-and-language navigation. E.T. is based on a multimodal transformer that encodes language inputs and the full episode history of visual observations and actions.
Episodic Transformer (E.T.) is a novel attention-based architecture for vision-and-language navigation. E.T. is based on a multimodal transformer that encodes language inputs and the full episode history of visual observations and actions.

Episodic Transformers (E.T.) Episodic Transformer for Vision-and-Language Navigation Alexander Pashevich, Cordelia Schmid, Chen Sun Episodic Transform

Alex Pashevich 62 Dec 24, 2022
CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped
CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped

CSWin-Transformer This repo is the official implementation of "CSWin Transformer: A General Vision Transformer Backbone with Cross-Shaped Windows". Th

Microsoft 409 Jan 6, 2023
3D-Transformer: Molecular Representation with Transformer in 3D Space

3D-Transformer: Molecular Representation with Transformer in 3D Space

null 55 Dec 19, 2022
This repository builds a basic vision transformer from scratch so that one beginner can understand the theory of vision transformer.

vision-transformer-from-scratch This repository includes several kinds of vision transformers from scratch so that one beginner can understand the the

null 1 Dec 24, 2021
Transformer - Transformer in PyTorch

Transformer 完成进度 Embeddings and PositionalEncoding with example. MultiHeadAttent

Tianyang Li 1 Jan 6, 2022
Transformer Huffman coding - Complete Huffman coding through transformer

Transformer_Huffman_coding Complete Huffman coding through transformer 2022/2/19

null 3 May 19, 2022
The official repository for our paper "The Devil is in the Detail: Simple Tricks Improve Systematic Generalization of Transformers". We significantly improve the systematic generalization of transformer models on a variety of datasets using simple tricks and careful considerations.

Codebase for training transformers on systematic generalization datasets. The official repository for our EMNLP 2021 paper The Devil is in the Detail:

Csordás Róbert 57 Nov 21, 2022
Comments
  • Evaluating ponder-net on more pondering-steps than trained on.

    Evaluating ponder-net on more pondering-steps than trained on.

    As the paper says,

    In evaluation, and under known temporal or computational limitations, N can be set naively as a constant (or not set any limit, i.e. N → ∞). For training, we found that a more effective (and interpretable) way of parameterizing N is by defining a minimum cumulative probability of halting. N is then the smallest value of n such that sum( p_sub_ j > 1 − ε)over(j=1, n) , with the hyper-parameter ε positive near 0 (in our experiments 0.05).

    from that I infer that pondering can be done to more steps than trained on. How can be done so with this implementation?

    edit: I was going through the paper again,and I think what the paper means is that the max_num_pondering_steps:N should be re evaluated at every training-step, the model should be run till the condition is met or a pre-defined num of max steps is reached, and where the cumsum_probs condition will be met will be set as 'N', with the cumsum_probs normalised with one of the methods. Then that value of 'N' will be used to calc prior geom for the kl_div (and not normalising the prior geom term).

    i.e. if the num of pondering steps are initially set to 'M', then the model will recur for 'k' steps - i.e. till the condition is met or for 'M' num of max steps; then 'N' will be calculated by first calculating the probabilities - p_0 to p_k - then normalizing through one of the methods, then calculate cumulative-sum of those probabilities, and checking where the sum is greater than threshold, and assigning it the value 'N'. After that, calculating prior geometric values with the defined hyper-parameter, for 'N' seq-len, and using this in the kl-div term against the halting probs truncated to 'N' steps.

    λp is a hyper-parameter that defines a geometric prior distribution pG(λp) on the halting policy (truncated at N)

    opened by Vbansal21 0
  • Can pondernet used for imagenet?

    Can pondernet used for imagenet?

    I plan to do a project on the complexity of tasks on image dataset like imagenet, cifar 100. If I use a vision transformer, then can I implement my project?

    opened by fryegg 2
Releases(0.0.8)
Owner
Phil Wang
Working with Attention. It's all we need
Phil Wang
GitHub Repository
Fastshap: A fast, approximate shap kernel

fastshap: A fast, approximate shap kernel fastshap was designed to be: Fast Calculating shap values can take an extremely long time. fastshap utilizes

Samuel Wilson 22 Sep 24, 2022
Differentiable rasterization applied to 3D model simplification tasks

nvdiffmodeling Differentiable rasterization applied to 3D model simplification tasks, as described in the paper: Appearance-Driven Automatic 3D Model

NVIDIA Research Projects 336 Dec 30, 2022
The code release of paper Low-Light Image Enhancement with Normalizing Flow

[AAAI 2022] Low-Light Image Enhancement with Normalizing Flow Paper | Project Page Low-Light Image Enhancement with Normalizing Flow Yufei Wang, Renji

Yufei Wang 176 Jan 06, 2023
Anomaly detection analysis and labeling tool, specifically for multiple time series (one time series per category)

taganomaly Anomaly detection labeling tool, specifically for multiple time series (one time series per category). Taganomaly is a tool for creating la

Microsoft 272 Dec 17, 2022
Mengzi Pretrained Models

中文 | English Mengzi 尽管预训练语言模型在 NLP 的各个领域里得到了广泛的应用，但是其高昂的时间和算力成本依然是一个亟需解决的问题。这要求我们在一定的算力约束下，研发出各项指标更优的模型。 我们的目标不是追求更大的模型规模，而是轻量级但更强大，同时对部署和工业落地更友好的模型。

Langboat 424 Jan 04, 2023
Boost learning for GNNs from the graph structure under challenging heterophily settings. (NeurIPS'20)

Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs Jiong Zhu, Yujun Yan, Lingxiao Zhao, Mark Heimann, Leman Akoglu,

GEMS Lab: Graph Exploration & Mining at Scale, University of Michigan 70 Dec 18, 2022
gtfs2vec - Learning GTFS Embeddings for comparing PublicTransport Offer in Microregions

gtfs2vec This is a companion repository for a gtfs2vec - Learning GTFS Embeddings for comparing PublicTransport Offer in Microregions publication. Vis

Politechnika Wrocławska - repozytorium dla informatyków 5 Oct 10, 2022
OSLO: Open Source framework for Large-scale transformer Optimization

O S L O Open Source framework for Large-scale transformer Optimization What's New: December 21, 2021 Released OSLO 1.0. What is OSLO about? OSLO is a

TUNiB 280 Nov 24, 2022
An integration of several popular automatic augmentation methods, including OHL (Online Hyper-Parameter Learning for Auto-Augmentation Strategy) and AWS (Improving Auto Augment via Augmentation Wise Weight Sharing) by Sensetime Research.

An integration of several popular automatic augmentation methods, including OHL (Online Hyper-Parameter Learning for Auto-Augmentation Strategy) and AWS (Improving Auto Augment via Augmentation Wise

45 Dec 08, 2022
Tools for manipulating UVs in the Blender viewport.

UV Tool Suite for Blender A set of tools to make editing UVs easier in Blender. These tools can be accessed wither through the Kitfox - UV panel on th

35 Oct 29, 2022
UMEC: Unified Model and Embedding Compression for Efficient Recommendation Systems

[ICLR 2021] "UMEC: Unified Model and Embedding Compression for Efficient Recommendation Systems" by Jiayi Shen, Haotao Wang*, Shupeng Gui*, Jianchao Tan, Zhangyang Wang, and Ji Liu

VITA 39 Dec 03, 2022
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
Code and Experiments for ACL-IJCNLP 2021 Paper Mind Your Outliers! Investigating the Negative Impact of Outliers on Active Learning for Visual Question Answering.

Code and Experiments for ACL-IJCNLP 2021 Paper Mind Your Outliers! Investigating the Negative Impact of Outliers on Active Learning for Visual Question Answering.

Sidd Karamcheti 50 Nov 16, 2022
Code for the CIKM 2019 paper "DSANet: Dual Self-Attention Network for Multivariate Time Series Forecasting".

Dual Self-Attention Network for Multivariate Time Series Forecasting 20.10.26 Update: Due to the difficulty of installation and code maintenance cause

Kyon Huang 223 Dec 16, 2022
A denoising diffusion probabilistic model synthesises galaxies that are qualitatively and physically indistinguishable from the real thing.

Realistic galaxy simulation via score-based generative models Official code for 'Realistic galaxy simulation via score-based generative models'. We us

Michael Smith 32 Dec 20, 2022
Implementation of popular bandit algorithms in batch environments.

batch-bandits Implementation of popular bandit algorithms in batch environments. Source code to our paper "The Impact of Batch Learning in Stochastic

Danil Provodin 2 Sep 11, 2022
WTTE-RNN a framework for churn and time to event prediction

WTTE-RNN Weibull Time To Event Recurrent Neural Network A less hacky machine-learning framework for churn- and time to event prediction. Forecasting p

Egil Martinsson 727 Dec 28, 2022
On Nonlinear Latent Transformations for GAN-based Image Editing - PyTorch implementation

On Nonlinear Latent Transformations for GAN-based Image Editing - PyTorch implementation On Nonlinear Latent Transformations for GAN-based Image Editi

Valentin Khrulkov 22 Oct 24, 2022
Monitora la qualità della ricezione dei segnali radio nelle province siciliane.

FMap-server Monitora la qualità della ricezione dei segnali radio nelle province siciliane. Conversion data Frequency - StationName maps are stored in

Triglie 5 May 24, 2021
Official implementation of Few-Shot and Continual Learning with Attentive Independent Mechanisms

Few-Shot and Continual Learning with Attentive Independent Mechanisms This repository is the official implementation of Few-Shot and Continual Learnin

Chikan_Huang 25 Dec 08, 2022