Do you have lots of data and are unsure of how to gain value from it? Do you think computer vision models require a large investment of hours of tagging and labeling data in order to work correctly? Well, this post will shed some light on how to avoid those extensive hours and loss of valuable time by using embeddings.

AI research is moving fast to make the creation of ML models faster and less dependent on humans. How? By moving towards machine learning models purely based on existing data without the need of hand labeling in order to create datasets.

Building Python Machine Learning projects that are both maintainable and easy to deploy is a hard job. To name a few topics, things like managing data pipelines, training multiple models, not to mention production deploys, and versioning. As you can imagine, everything can go out of control very quickly if not handled properly.

Thankfully, today we have so many options to choose from when it comes to cloud services for training and deploying your models into production, namely AWS SageMaker, Google’s AI Platform, or Azure’s Machine Learning.

TorchServe is a great tool to deploy trained PyTorch models, there is no denying that. But, as with any relatively new project, it is still creating a community around it to help with the more niche aspects of its implementation. As part of this community, we want to do our part. So today, we will be discussing how to develop advanced custom handlers with PyTorch’s TorchServe. We will also be reviewing the process of saving your PyTorch model with torch-model-archiver and how to include all the new artifacts created while we are at it.

We embarked on this journey specifically because, as great as the included inference handlers can be, you will need to tailor them at some point to fit your needs.

We can’t believe Streamlit is barely one year old now. Today, we can’t think of any better tool for creating a fast and interactive dashboard, demos, and visualizations. On top of that, they recently launched a new awesome feature called Streamlit Components. This new feature enables developers to create custom visualizations using your favorite front-end framework, namely React, Vue, or just plain HTML.

Especially, Streamlit is really convenient when starting new machine learning projects, making it super easy and fast to go from nothing to a proof of concept of a ML solution. Because of this, we use Streamlit pretty frequently: for project prototyping, dynamic reports, and for building cool demos 💅

Computer vision has seen tremendous change in the last decade. Along this journey, the nature of the techniques used has shifted from fixed-pipeline algorithms based mostly in heuristics and little data to resource-intensive general-purpose models that require both heavy computing power and a lot of data. This shift, while unlocking new capabilities, presents some drawbacks.

Most of the work done at the research level is fragmented because researchers, by the very nature of good research, need to focus on one aspect at a time when planning their experiments. And while there are papers on unifying techniques, the difficulty to evaluate, and the requirements on what a novel contribution is being so stringent, these are few and apart. Most of these works come from non-academic initiatives such as http://fast.ai/.

We will shortly introduce the Google Cloud service Dataflow, and how it can be used to run predictions on millions of images in a serverless way. No cluster creation, no maintenance, and you only pay for what you use. We will start by providing a context on why we think this is important, a brief introduction to a couple of concepts and then go directly into a use case.

Scaling Machine Learning models is hard and expensive. It is costly mainly for two reasons: because the infrastructure is expensive, and because your experimentation pipeline is slow causing your ML team to be constantly waiting for the results.