We are proud to introduce the {mall}. With {mall}, you can use a local LLM to run NLP operations across a data frame. (sentiment, summarization, translation, etc). {mall} has been simultaneusly released to CRAN and PyPi (as an extension to Polars).
We are thrilled to introduce {keras3}, the next version of the Keras R package. {keras3} is a ground-up rebuild of {keras}, maintaining the beloved features of the original while refining and simplifying the API based on valuable insights gathered over the past few years.
Highlights to the most recent updates to `sparklyr` and friends
Interact with Github Copilot and OpenAI's GPT (ChatGPT) models directly in RStudio. The `chattr` Shiny add-in makes it easy for you to interact with these and other Large Language Models (LLMs).
Hugging Face rapidly became a very popular platform to build, share and collaborate on deep learning applications. We have worked on integrating the torch for R ecosystem with Hugging Face tools, allowing users to load and execute language models from their platform.
LoRA (Low Rank Adaptation) is a new technique for fine-tuning deep learning models that works by reducing the number of trainable parameters and enables efficient task switching. In this blog post we will talk about the key ideas behind LoRA in a very minimal torch example.
Implementing a language model from scratch is, arguably, the best way to develop an accurate idea of how its engine works. Here, we use torch to code GPT-2, the immediate successor to the original GPT. In the end, you'll dispose of an R-native model that can make direct use of Hugging Face's pre-trained GPT-2 model weights.
This is a high-level, introductory article about Large Language Models (LLMs), the core technology that enables the much-en-vogue chatbots as well as other Natural Language Processing (NLP) applications. It is directed at a general audience, possibly with some technical and/or scientific background, but no knowledge is assumed of either deep learning or NLP. Having looked at major model ingredients, training workflow, and mechanics of output generation, we also talk about what these models are not.
Announcing safetensors, a new R package allowing for reading and writing files in the safetensors format.
torch v0.11.0 is now on CRAN. This release features much-enhanced support for executing JIT operations. We also amended loading of model parameters, and added a few quality-of-life improvements, like support for temporarily modifying the default torch device, support for specifying data types as strings, and many more.
Implementation and walk-through of LLaMA, a Large Language Model, in R, with TensorFlow and Keras.
Escnn, built on PyTorch, is a library that, in the spirit of Geometric Deep Learning, provides a high-level interface to designing and training group-equivariant neural networks. This post introduces important mathematical concepts, the library's key actors, and essential library use.
luz v0.4.0 is now on CRAN. This release adds support for training models on ARM Mac GPUs, reduces the overhead of using luz, and makes it easier to checkpoint and resume failed runs.
torch v0.10.0 is now on CRAN. This version upgraded the underlying LibTorch to 1.13.1, and added support for Automatic Mixed Precision. As an experimental feature, we now also support pre-built binaries, so you can install torch without having to deal with the CUDA installation.
Currently, in generative deep learning, no other approach seems to outperform the family of diffusion models. Would you like to try for yourself? If so, our torch implementation of de-noising diffusion provides an easy-to-use, easy-to-configure interface.
Please allow us to introduce Deep Learning and Scientific Computing with R torch. Released in e-book format today, and available freely online, this book starts out by introducing torch basics. From there, it moves on to various deep-learning use cases. Finally, it shows how to use torch for more general topics, such as matrix computations and the Fourier Transform.
We code up a simple group-equivariant convolutional neural network (GCNN) that is equivariant to rotation. The world may be upside down, but the network will know.
In this first in a series of posts on group-equivariant convolutional neural networks (GCNNs), meet the main actors — groups — and concepts (equivariance). With GCNNs, we finally revisit the topic of Geometric Deep Learning, a principled, math-driven approach to neural networks that has consistently been rising in scope and impact.
El Niño-Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), and Arctic Oscillation (AO) are atmospheric phenomena of global impact that strongly affect people's lives. ENSO, first and foremost, brings with it floods, droughts, and ensuing poverty, in developing countries in the Southern Hemisphere. Here, we use the new torchwavelets package to comparatively inspect patterns in the three series.
torch does not have built-in functionality to do wavelet analysis. But we can efficiently implement what we need, making use of the Fast Fourier Transform (FFT). This post is a very first introduction to wavelets, suitable for readers that have not encountered it before. At the same time, it provides useful starter code, showing an (extensible) way to perform wavelet analysis in torch. It is an excerpt from the corresponding chapter in the forthcoming book, Deep Learning and Scientific Computing with R torch, to be published by CRC Press.
torch v0.9.0 is now on CRAN. This version adds support for ARM systems running macOS, and brings significant performance improvements.
About the Fourier Transform, it has been said that it is one of the greatest wonders of the universe. At the same time, it can be realized in a mere half-dozen lines of code. Even if in the end, you're just going to call torch's built-in functions directly, it helps to understand, and be able to reproduce in code, the ideas that underlie the magic. This post is an excerpt from the forthcoming book, Deep Learning and Scientific Computing with R torch, to be published by CRC Press.
Get to know torch's linalg module, all while learning about different ways to do least-squares regression from scratch. This post is a condensed version of the corresponding chapter in the forthcoming book, Deep Learning and Scientific Computing with R torch, to be published by CRC Press.
Learn how to classify speech utterances with torch, making use of domain knowledge and deep learning. This post is a condensed version of the corresponding chapter in the forthcoming book, Deep Learning and Scientific Computing with R torch, to be published by CRC Press.
The text package attempts to provide user-friendly access and pipelines to HuggingFace's transformer language models in R.
luz version 0.3.0 is now on CRAN. luz is a high-level interface for torch.
New TensorFlow and Keras releases bring improvements big and small.
Announcing the release of "Deep Learning with R, 2nd Edition," a book that shows you how to get started with deep learning in R.
Today, we want to call attention to a highly useful package in the torch ecosystem: torchopt. It extends torch by providing a set of popular optimization algorithms not available in the base library. As this post will show, it is also fun to use!
Sometimes, a software's best feature is the one you've added yourself. This post shows by example why you may want to extend torch, and how to proceed. It also explains a bit of what is going on in the background.
For keras, the last two releases have brought important new functionality, in terms of both low-level infrastructure and workflow enhancements. This post focuses on an outstanding example of the latter category: a new family of layers designed to help with pre-processing, data-augmentation, and feature-engineering tasks.
It's been a while since this blog featured content about Keras for R, so you might've thought that the project was dormant. It's not! In fact, Keras for R is better than ever, with two recent releases adding powerful capabilities that considerably lighten previously tedious tasks. This post provides a high-level overview. Future posts will go into more detail on some of the most helpful new features, as well as dive into the powerful low-level enhancements that make the former possible.
We train a model for image segmentation in R, using torch together with luz, its high-level interface. We then JIT-trace the model on example input, so as to obtain an optimized representation that can run with no R installed. Finally, we show the model being run on Android.
Geometric deep learning is a "program" that aspires to situate deep learning architectures and techniques in a framework of mathematical priors. The priors, such as various types of invariance, first arise in some physical domain. A neural network that well matches the domain will preserve as many invariances as possible. In this post, we present a very conceptual, high-level overview, and highlight a few applications.
Using the torch just-in-time (JIT) compiler, it is possible to query a model trained in R from a different language, provided that language can make use of the low-level libtorch library. This post shows how. In addition, we try to untangle a bit of the terminological jumble surrounding the topic.
The topic of AI fairness metrics is as important to society as it is confusing. Confusing it is due to a number of reasons: terminological proliferation, abundance of formulae, and last not least the impression that everyone else seems to know what they're talking about. This text hopes to counteract some of that confusion by starting from a common-sense approach of contrasting two basic positions: On the one hand, the assumption that dataset features may be taken as reflecting the underlying concepts ML practitioners are interested in; on the other, that there inevitably is a gap between concept and measurement, a gap that may be bigger or smaller depending on what is being measured. In contrasting these fundamental views, we bring together concepts from ML, legal science, and political philosophy.
We are excited to announce the availability of sparklyr.sedona, a sparklyr extension making geospatial functionalities of the Apache Sedona library easily accessible from R.
Sparklyr 1.7 delivers much-anticipated improvements, including R interfaces for image and binary data sources, several new spark_apply() capabilities, and better integration with sparklyr extensions.
Today, we're introducing luz, a high-level interface to torch that lets you train neural networks in a concise, declarative style. In some sense, it is to torch what Keras is to TensorFlow: It provides both a streamlined workflow and powerful ways for customization.
Torch is not just for deep learning. Its L-BFGS optimizer, complete with Strong-Wolfe line search, is a powerful tool in unconstrained as well as constrained optimization.
The sparklyr 1.6 release introduces weighted quantile summaries, an R interface to power iteration clustering, spark_write_rds(), as well as a number of dplyr-related improvements.
We conclude our mini-series on time-series forecasting with torch by augmenting last time's sequence-to-sequence architecture with a technique both immensely popular in natural language processing and inspired by human (and animal) cognition: attention.
In our overview of techniques for time-series forecasting, we move on to sequence-to-sequence models. Architectures in this family are commonly used in natural language processing (NLP) tasks, such as machine translation. With NLP, however, significant pre-processing is required before proceeding to model definition and training. In staying with our familiar numerical series, we can fully concentrate on the concepts.
We continue our exploration of time-series forecasting with torch, moving on to architectures designed for multi-step prediction. Here, we augment the "workhorse RNN" by a multi-layer perceptron (MLP) to extrapolate multiple timesteps into the future.
This post is an introduction to time-series forecasting with torch. Central topics are data input, and practical usage of RNNs (GRUs/LSTMs). Upcoming posts will build on this, and introduce increasingly involved architectures.
Last month, we conducted our first survey on mlverse software, covering topics ranging from area of application through software usage to user wishes and suggestions. In addition, the survey asked about thoughts on social impacts of AI/ML. This post presents the results, and tries to address some of the things that came up.
Today we introduce tabnet, a torch implementation of "TabNet: Attentive Interpretable Tabular Learning" that is fully integrated with the tidymodels framework. Per se, already, tabnet was designed to require very little data pre-processing; thanks to tidymodels, hyperparameter tuning (so often cumbersome in deep learning) becomes convenient and even, fun!
This article translates Daniel Falbel's post on "Simple Audio Classification" from TensorFlow/Keras to torch/torchaudio.
El Niño-Southern Oscillation (ENSO) is an atmospheric phenomenon, located in the tropical Pacific, that greatly affects ecosystems as well as human well-being on a large portion of the globe. We use the convLSTM introduced in a prior post to predict the Niño 3.4 Index from spatially-ordered sequences of sea surface temperatures.
In forecasting spatially-determined phenomena (the weather, say, or the next frame in a movie), we want to model temporal evolution, ideally using recurrence relations. At the same time, we'd like to efficiently extract spatial features, something that is normally done with convolutional filters. Ideally then, we'd have at our disposal an architecture that is both recurrent and convolutional. In this post, we build a convolutional LSTM with torch.
The torch 0.2.0 release includes many bug fixes and some nice new features like initial JIT support, multi-worker dataloaders, new optimizers and a new print method for nn_modules.
Unlike all three previous sparklyr releases, the recent release of sparklyr 1.5 placed much more emphasis on enhancing existing sparklyr features rather than creating new ones. As a result, many valuable suggestions from sparklyr users were taken into account and were successfully addressed in a long list of bug fixes and improvements.
The need to segment images arises in various sciences and their applications, many of which are vital to human (and animal) life. In this introductory post, we train a U-Net to mark lesioned regions on MRI brain scans.
How not to die from poisonous mushrooms. Also: How to use torch for deep learning on tabular data, including a mix of categorical and numerical features.
We learn about transfer learning, input pipelines, and learning rate schedulers, all while using torch to tell apart species of beautiful birds.
We are excited to announce a number of powerful, new functionalities and improvements which are now part of sparklyr.flint 0.2!
Today, we wrap up our mini-series on torch basics, adding to our toolset two abstractions: loss functions and optimizers.
In this third installment of our mini-series introducing torch basics, we replace hand-coded matrix operations by modules, considerably simplifying our toy network's code.
With torch, there is hardly ever a reason to code backpropagation from scratch. Its automatic differentiation feature, called autograd, keeps track of operations that need their gradients computed, as well as how to compute them. In this second post of a four-part series, we update our simple, hand-coded network to make use of autograd.
In this first installment of a four-part miniseries, we present the main things you will want to know about torch tensors. As an illustrative example, we'll code a simple neural network from scratch.
Sparklyr 1.4 is now available! This release comes with delightful new features such as weighted sampling and tidyr verbs support for Spark dataframes, robust scaler for standardizing data based on median and interquartile range, spark_connect interface for RAPIDS GPU acceleration plugin, as well as a number of dplyr-related improvements.
Today, we are excited to introduce torch, an R package that allows you to use PyTorch-like functionality natively from R. No Python installation is required: torch is built directly on top of libtorch, a C++ library that provides the tensor-computation and automatic-differentiation capabilities essential to building neural networks.
We are pleased to announce that sparklyr.flint, a sparklyr extension for analyzing time series at scale with Flint, is now available on CRAN. Flint is an open-source library for working with time-series in Apache Spark which supports aggregates and joins on time-series datasets.
A few weeks ago, we showed how to forecast chaotic dynamical systems with deep learning, augmented by a custom constraint derived from domain-specific insight. Global weather is a chaotic system, but of much higher complexity than many tasks commonly addressed with machine and/or deep learning. In this post, we provide a practical introduction featuring a simple deep learning baseline for atmospheric forecasting. While far away from being competitive, it serves to illustrate how more sophisticated and compute-intensive models may approach that formidable task by means of methods situated on the "black-box end" of the continuum.
This post explores how to train large datasets with TensorFlow and R. Specifically, we present how to download and repartition ImageNet, followed by training ImageNet across multiple GPUs in distributed environments using TensorFlow and Apache Spark.
A couple of months ago, Amazon, Facebook, Microsoft, and other contributors initiated a challenge consisting of telling apart real and AI-generated ("fake") videos. We show how to approach this challenge from R.
In the last part of this mini-series on forecasting with false nearest neighbors (FNN) loss, we replace the LSTM autoencoder from the previous post by a convolutional VAE, resulting in equivalent prediction performance but significantly lower training time. In addition, we find that FNN regularization is of great help when an underlying deterministic process is obscured by substantial noise.
Nowadays, Microsoft, Google, Facebook, and OpenAI are sharing lots of state-of-the-art models in the field of Natural Language Processing. However, fewer materials exist how to use these models from R. In this post, we will show how R users can access and benefit from these models as well.
How can the seemingly iterative process of weighted sampling without replacement be transformed into something highly parallelizable? Turns out a well-known technique based on exponential variates accomplishes exactly that.
In a recent post, we showed how an LSTM autoencoder, regularized by false nearest neighbors (FNN) loss, can be used to reconstruct the attractor of a nonlinear, chaotic dynamical system. Here, we explore how that same technique assists in prediction. Matched up with a comparable, capacity-wise, "vanilla LSTM", FNN-LSTM improves performance on a set of very different, real-world datasets, especially for the initial steps in a multi-step forecast.
Sparklyr 1.3 is now available, featuring exciting new functionalities such as integration of Spark higher-order functions and data import/export in Avro and in user-defined serialization formats.
In nonlinear dynamics, when the state space is thought to be multidimensional but all we have for data is just a univariate time series, one may attempt to reconstruct the true space via delay coordinate embeddings. However, it is not clear a priori how to choose dimensionality and time lag of the reconstruction space. In this post, we show how to use an autoencoder architecture to circumvent the problem: Given just a scalar series of observations, the autoencoder directly learns to represent attractors of chaotic systems in adequate dimensionality.
PixelCNN is a deep learning architecture - or bundle of architectures - designed to generate highly realistic-looking images. To use it, no reverse-engineering of arXiv papers or search for reference implementations is required: TensorFlow Probability and its R wrapper, tfprobability, now include a PixelCNN distribution that can be used to train a straightforwardly-defined neural network in a parameterizable way.
Compared to other applications, deep learning models might not seem too likely as victims of privacy attacks. However, methods exist to determine whether an entity was used in the training set (an adversarial attack called member inference), and techniques subsumed under "model inversion" allow to reconstruct raw data input given just model output (and sometimes, context information). This post shows an end-to-end example of model inversion, and explores mitigation strategies using TensorFlow Privacy.
Deep learning need not be irreconcilable with privacy protection. Federated learning enables on-device, distributed model training; encryption keeps model and gradient updates private; differential privacy prevents the training data from leaking. As of today, private and secure deep learning is an emerging technology. In this post, we introduce Syft, an open-source framework that integrates with PyTorch as well as TensorFlow. In an example use case, we obtain private predictions from a Keras model.
A new sparklyr release is now available. This sparklyr 1.2 release features new functionalities such as support for Databricks Connect, a Spark backend for the 'foreach' package, inter-op improvements for working with Spark 3.0 preview, as well as a number of bug fixes and improvements addressing user-visible pain points.
A new release of pins is available on CRAN today. This release adds support to time travel across dataset versions, which improves collaboration and protects your code from breaking when remote resources change unexpectedly.
The term "federated learning" was coined to describe a form of distributed model training where the data remains on client devices, i.e., is never shipped to the coordinating server. In this post, we introduce central concepts and run first experiments with TensorFlow Federated, using R.
This blog just got a new title: RStudio AI Blog. We explain why.
Kullback-Leibler divergence is not just used to train variational autoencoders or Bayesian networks (and not just a hard-to-pronounce thing). It is a fundamental concept in information theory, put to use in a vast range of applications. Most interestingly, it's not always about constraint, regularization or compression. Quite on the contrary, sometimes it is about novelty, discovery and surprise.
Broadcasting, as done by Python's scientific computing library NumPy, involves dynamically extending shapes so that arrays of different sizes may be passed to operations that expect conformity - such as adding or multiplying elementwise. In NumPy, the way broadcasting works is specified exactly; the same rules apply to TensorFlow operations. For anyone who finds herself, occasionally, consulting Python code, this post strives to explain.
TensorFlow 2.1, released last week, allows for mixed-precision training, making use of the Tensor Cores available in the most recent NVidia GPUs. In this post, we report first experimental results and provide some background on what this is all about.
Differential Privacy guarantees that results of a database query are basically independent of the presence in the data of a single individual. Applied to machine learning, we expect that no single training example influences the parameters of the trained model in a substantial way. This post introduces TensorFlow Privacy, a library built on top of TensorFlow, that can be used to train differentially private deep learning models from R.
TensorFlow Hub is a library for the publication, discovery, and consumption of reusable parts of machine learning models. A module is a self-contained piece of a TensorFlow graph, along with its weights and assets, that can be reused across different tasks in a process known as transfer learning.
Continuing our tour of applications of TensorFlow Probability (TFP), after Bayesian Neural Networks, Hamiltonian Monte Carlo and State Space Models, here we show an example of Gaussian Process Regression. In fact, what we see is a rather "normal" Keras network, defined and trained in pretty much the usual way, with TFP's Variational Gaussian Process layer pulling off all the magic.
Looking for materials to get started with deep learning from R? This post presents useful tutorials, guides, and background documentation on the new TensorFlow for R website. Advanced users will find pointers to applications of new release 2.0 (or upcoming 2.1!) features alluded to in the recent TensorFlow 2.0 post.
In a Bayesian neural network, layer weights are distributions, not tensors. Using tfprobability, the R wrapper to TensorFlow Probability, we can build regular Keras models that have probabilistic layers, and thus get uncertainty estimates "for free". In this post, we show how to define, train and obtain predictions from a probabilistic convolutional neural network.
Part of the r-tensorflow ecosystem, tfprobability is an R wrapper to TensorFlow Probability, the Python probabilistic programming framework developed by Google. We take the occasion of tfprobability's acceptance on CRAN to give a high-level introduction, highlighting interesting use cases and applications.
Is society ready to deal with challenges brought about by artificially-generated information - fake images, fake videos, fake text? While this post won't answer that question, it should help form an opinion on the threat exerted by fake text as of this writing, autumn 2019. We introduce gpt2, an R package that wraps OpenAI's public implementation of GPT-2, the language model that early this year surprised the NLP community with the unprecedented quality of its creations.
TensorFlow 2.0 was finally released last week. As R users we have two kinds of questions. First, will my keras code still run? And second, what is it that changes? In this post, we answer both and, then, give a tour of exciting new developments in the r-tensorflow ecosystem.
TensorFlow Probability, and its R wrapper tfprobability, provide Markov Chain Monte Carlo (MCMC) methods that were used in a number of recent posts on this blog. These posts were directed to users already comfortable with the method, and terminology, per se, which readers mainly interested in deep learning won't necessarily be. Here we try to make up leeway, introducing Hamitonian Monte Carlo (HMC) as well as a few often-heard "buzzwords" accompanying it, always striving to keep in mind what it is all "for".
A deep learning model - BERT from Google AI Research - has yielded state-of-the-art results in a wide variety of Natural Language Processing (NLP) tasks. In this tutorial, we will show how to load and train the BERT model from R, using Keras.
Have you ever wondered why you can call TensorFlow - mostly known as a Python framework - from R? If not - that's how it should be, as the R packages keras and tensorflow aim to make this process as transparent as possible to the user. But for them to be those helpful genies, someone else first has to tame the Python.
In image segmentation, every pixel of an image is assigned a class. Depending on the application, classes could be different cell types; or the task could be binary, as in "cancer cell yes or no?". Area of application notwithstanding, the established neural network architecture of choice is U-Net. In this post, we show how to preprocess data and train a U-Net model on the Kaggle Carvana image segmentation data.
In this post we use tfprobability, the R interface to TensorFlow Probability, to model censored data. Again, the exposition is inspired by the treatment of this topic in Richard McElreath's Statistical Rethinking. Instead of cute cats though, we model immaterial entities from the cold world of technology: This post explores durations of CRAN package checks, a dataset that comes with Max Kuhn's parsnip.
TensorFlow feature columns provide useful functionality for preprocessing categorical data and chaining transformations, like bucketization or feature crossing. From R, we use them in popular "recipes" style, creating and subsequently refining a feature specification. In this post, we show how using feature specs frees cognitive resources and lets you focus on what you really want to accomplish. What's more, because of its elegance, feature-spec code reads nice and is fun to write as well.
Previous posts featuring tfprobability - the R interface to TensorFlow Probability - have focused on enhancements to deep neural networks (e.g., introducing Bayesian uncertainty estimates) and fitting hierarchical models with Hamiltonian Monte Carlo. This time, we show how to fit time series using dynamic linear models (DLMs), yielding posterior predictive forecasts as well as the smoothed and filtered estimates from the Kálmán filter.
As of today, there is no mainstream road to obtaining uncertainty estimates from neural networks. All that can be said is that, normally, approaches tend to be Bayesian in spirit, involving some way of putting a prior over model weights. This holds true as well for the method presented in this post: We show how to use tfprobability, the R interface to TensorFlow Probability, to add uncertainty estimates to a Keras model in an elegant and conceptually plausible way.
This post builds on our recent introduction to multi-level modeling with tfprobability, the R wrapper to TensorFlow Probability. We show how to pool not just mean values ("intercepts"), but also relationships ("slopes"), thus enabling models to learn from data in an even broader way. Again, we use an example from Richard McElreath's "Statistical Rethinking"; the terminology as well as the way we present this topic are largely owed to this book.
This post is a first introduction to MCMC modeling with tfprobability, the R interface to TensorFlow Probability (TFP). Our example is a multi-level model describing tadpole mortality, which may be known to the reader from Richard McElreath's wonderful "Statistical Rethinking".
Continuing from the recent introduction to bijectors in TensorFlow Probability (TFP), this post brings autoregressivity to the table. Using TFP through the new R package tfprobability, we look at the implementation of masked autoregressive flows (MAF) and put them to use on two different datasets.
Sometimes in deep learning, architecture design and hyperparameter tuning pose substantial challenges. Using Auto-Keras, none of these is needed: We start a search procedure and extract the best-performing model. This post presents Auto-Keras in action on the well-known MNIST dataset.
Normalizing flows are one of the lesser known, yet fascinating and successful architectures in unsupervised deep learning. In this post we provide a basic introduction to flows using tfprobability, an R wrapper to TensorFlow Probability. Upcoming posts will build on this, using more complex flows on more complex data.
Not everybody who wants to get into deep learning has a strong background in math or programming. This post elaborates on a concepts-driven, abstraction-based way to learn what it's all about.
Sometimes, deep learning is seen - and welcomed - as a way to avoid laborious preprocessing of data. However, there are cases where preprocessing of sorts does not only help improve prediction, but constitutes a fascinating topic in itself. One such case is audio classification. In this post, we build on a previous post on this blog, this time focusing on explaining some of the non-deep learning background. We then link the concepts explained to updated for near-future releases TensorFlow code.
Mostly when thinking of Variational Autoencoders (VAEs), we picture the prior as an isotropic Gaussian. But this is by no means a necessity. The Vector Quantised Variational Autoencoder (VQ-VAE) described in van den Oord et al's "Neural Discrete Representation Learning" features a discrete latent space that allows to learn impressively concise latent representations. In this post, we combine elements of Keras, TensorFlow, and TensorFlow Probability to see if we can generate convincing letters resembling those in Kuzushiji-MNIST.
TensorFlow Probability offers a vast range of functionality ranging from distributions over probabilistic network layers to probabilistic inference. It works seamlessly with core TensorFlow and (TensorFlow) Keras. In this post, we provide a short introduction to the distributions layer and then, use it for sampling and calculating probabilities in a Variational Autoencoder.
As shown in a previous post, naming and locating a single object in an image is a task that may be approached in a straightforward way. This is not the same with general object detection, though - naming and locating several objects at once, with no prior information about how many objects are supposed to be detected. In this post, we explain the steps involved in coding a basic single-shot object detector: Not unlike SSD (Single-shot Multibox Detector), but simplified and designed not for best performance, but comprehensibility.
Embedding layers are not just useful when working with language data. As "entity embeddings", they've recently become famous for applications on tabular, small-scale data. In this post, we exemplify two possible use cases, also drawing attention to what not to expect.
In deep learning, there is no obvious way of obtaining uncertainty estimates. In 2016, Gal and Ghahramani proposed a method that is both theoretically grounded and practical: use dropout at test time. In this post, we introduce a refined version of this method (Gal et al. 2017) that has the network itself learn how uncertain it is.
Object detection (the act of classifying and localizing multiple objects in a scene) is one of the more difficult, but very relevant in practice deep learning tasks. We'll build up to it in several posts. Here we start with the simpler tasks of naming and locating a single object.
Like GANs, variational autoencoders (VAEs) are often used to generate images. However, VAEs add an additional promise: namely, to model an underlying latent space. Here, we first look at a typical implementation that maximizes the evidence lower bound. Then, we compare it to one of the more recent competitors, MMD-VAE, from the Info-VAE (information maximizing VAE) family.
Why do we use the activations we use, and how do they relate to the cost functions they tend to co-appear with? In this post we provide a conceptual introduction.
Advanced applications like generative adversarial networks, neural style transfer, and the attention mechanism ubiquitous in natural language processing used to be not-so-simple to implement with the Keras declarative coding paradigm. Now, with the advent of TensorFlow eager execution, things have changed. This post explores using eager execution with R.
Embeddings are not just for use in natural language processing. Here we apply embeddings to a common task in collaborative filtering - predicting user ratings - and on our way, strive for a better understanding of what an embedding layer really does.
Conditional GANs (cGANs) may be used to generate one type of object based on another - e.g., a map based on a photo, or a color video based on black-and-white. Here, we show how to implement the pix2pix approach with Keras and eager execution.
Image captioning is a challenging task at intersection of vision and language. Here, we demonstrate using Keras and eager execution to incorporate an attention mechanism that allows the network to concentrate on image features relevant to the current state of text generation.
Continuing our series on combining Keras with TensorFlow eager execution, we show how to implement neural style transfer in a straightforward way. Based on this easy-to-adapt example, you can easily perform style transfer on your own images.
Many fields are benefiting from the use of deep learning, and with the R keras, tensorflow and related packages, you can now easily do state of the art deep learning in R. In this post, we want to give some orientation as to how to best get started.
Generative adversarial networks (GANs) are a popular deep learning approach to generating new entities (often but not always images). We show how to code them using Keras and TensorFlow eager execution.
As sequence to sequence prediction tasks get more involved, attention mechanisms have proven helpful. A prominent example is neural machine translation. Following a recent Google Colaboratory notebook, we show how to implement attention in R.
Using Keras to train a convolutional neural network to classify physical activity. The dataset was built from the recordings of 30 subjects performing basic activities and postural transitions while carrying a waist-mounted smartphone with embedded inertial sensors.
In this post we will examine making time series predictions using the sunspots dataset that ships with base R. Sunspots are dark spots on the sun, associated with lower temperature. Our post will focus on both how to apply deep learning to time series forecasting, and how to properly apply cross validation in this domain.
In this tutorial we will build a deep learning model to classify words. We will use the Speech Commands dataset which consists of 65,000 one-second audio files of people saying 30 different words.
If you don't have local access to a modern NVIDIA GPU, your best bet is typically to run GPU intensive training jobs in the cloud. Paperspace is a cloud service that provides access to a fully preconfigured Ubuntu 16.04 desktop environment equipped with a GPU.
A new major release of lime has landed on CRAN. lime is an R port of the Python library of the same name by Marco Ribeiro that allows the user to pry open black box machine learning models and explain their outcomes on a per-observation basis
The aim of this post is to illustrate how deep learning is being applied in cancer immunotherapy (Immuno-oncology or Immunooncology) - a cancer treatment strategy, where the aim is to utilize the cancer patient's own immune system to fight the cancer.
In this post we will train an autoencoder to detect credit card fraud. We will also demonstrate how to train Keras models in the cloud using CloudML. The basis of our model will be the Kaggle Credit Card Fraud Detection dataset.
The kerasformula package offers a high-level interface for the R interface to Keras. It’s main interface is the kms function, a regression-style interface to keras_model_sequential that uses formulas and sparse matrices. We use kerasformula to predict how popular tweets will be based on how often the tweet was retweeted and favorited.
Using Keras to predict customer churn based on the IBM Watson Telco Customer Churn dataset. We also demonstrate using the lime package to help explain which features drive individual model predictions. In addition, we use three new packages to assist with Machine Learning: recipes for preprocessing, rsample for sampling data and yardstick for model metrics.
We are excited to announce the availability of the cloudml package, which provides an R interface to Google Cloud Machine Learning Engine. CloudML provides a number of services including on-demand access to training on GPUs and hyperparameter tuning to optimize key attributes of model architectures.
In this post we will use Keras to classify duplicated questions from Quora. Our implementation is inspired by the Siamese Recurrent Architecture, with modifications to the similarity measure and the embedding layers (the original paper uses pre-trained word vectors)
Word embedding is a method used to map words of a vocabulary to dense vectors of real numbers where semantically similar words are mapped to nearby points. In this example we'll use Keras to generate word embeddings for the Amazon Fine Foods Reviews dataset.
In this post, we'll review three advanced techniques for improving the performance and generalization power of recurrent neural networks. We'll demonstrate all three concepts on a temperature-forecasting problem, where you have access to a time series of data points coming from sensors installed on the roof of a building.
Having to train an image-classification model using very little data is a common situation, in this article we review three techniques for tackling this problem including feature extraction and fine tuning from a pretrained network.
Two-class classification, or binary classification, may be the most widely applied kind of machine-learning problem. In this excerpt from the book Deep Learning with R, you'll learn to classify movie reviews as positive or negative, based on the text content of the reviews.
The tfruns package provides a suite of tools for tracking, visualizing, and managing TensorFlow training runs and experiments from R.
We are excited to announce that the keras package is now available on CRAN. The package provides an R interface to Keras, a high-level neural networks API developed with a focus on enabling fast experimentation.
The tfestimators package is an R interface to TensorFlow Estimators, a high-level API that provides implementations of many different model types including linear models and deep neural networks.
The final release of TensorFlow v1.3 is now available. This release marks the initial availability of several canned estimators including DNNClassifier and DNNRegressor.