Sept '20 tensorwerk Heartbeat

Hey folks! This month we are presenting a new blogpost with a beginners tutorial on Hangar Machine Learning data loaders and a recap of what happened during the Hangar Public Developers Meeting!

Every month we are sharing news on projects we are working on, conferences and events we attend, what are our plans for the future and everything that might be related to data.

Hangar data loaders tutorial

This month we are featuring a nice introductory blogpost from our community on Hangar data loaders for PyTorch!

Q: How easy is training a model with data saved in Hangar?

A: Super easy! Check out the tutorial from Jithin James on TowardsDataScience 😎

Design Illustration GIF

Hangar open discussion calls recap

If you missed any of the Public Hangar Developers meetings, here is a recap for you! Remember that all the recordings are published on our YouTube channel.

Here is what happened:

September 02, 2020

🔗  Video Link

🗣️ Stockroom submission on the PyTorch Summer Hackathon 2020, Stockroom quick start tutorial

September 09, 2020

🔗  Video Link

🗣️ Hangar commit reference structure new design

September 16, 2020

🔗  Video Link

🗣️ Hangar documentation, Hangar next steps

See you every Wednesday at 9AM EST / 3PM CEST! Please come with questions, we'd love to hear your thoughts and/or issues 🧡

Reach out

If you’d like to have a peek into our vision and our upcoming developments, please send us a note at info@tensorwerk.com. In any case, we will be posting our updates regularly here on Substack. Have fun and stay tuned.


If you want to stay up to date with ideas, projects and plans for the future at tensorwerk, subscribe to our publication and receive the Heartbeat directly in your inbox.

Aug '20 tensorwerk Heartbeat

Hey there! This month we are proud to present our submission to the PyTorch Summer Hackathon 2020 and a glimpse of the Hangar data storage design.

Every month we are sharing news on projects we are working on, conferences and events we attend, what are our plans for the future and everything that might be related to data.

PyTorch Summer Hackathon 2020

We participated in the PyTorch Summer Hackaton 2020 with our Stockroom project! We talked about Stockroom in the last and in the very first heartbeat: it’s a very lightweight ML lifecycle tool based on git and Hangar. If you are not familiar with the project, be sure to check it out.

Our Sherin also made an introductory video, showing a step-by-step walkthrough on how to set up a Stockroom repository, how to import a dataset from torchvision, and how to save experiments metadata and model weights during the training of a neural network.

If you are curious, here is the official submission 👉https://devpost.com/software/stockroom

Did you know that … Data Is Large, We Don’t Waste Space

When a user requests to add data to a Hangar repository, one of the first operations Hangar does is to generate a hash of the array contents. If the hash does not match a piece of data already placed in the Hangar repository, the data is sent to the appropriate storage backend methods.

On the other hand, if a data sample is added to a repository that already has a record of some hash, we don’t even involve the storage backend. All we need to do is just record that a new sample in a column was added with that hash. It makes no sense to write the same data twice.

Read more about the Hangar design on the official documentation 👉 https://hangar-py.readthedocs.io/en/stable/design.html

Reach out

If you’d like to have a peek into our vision and our upcoming developments, please send us a note at info@tensorwerk.com. In any case, we will be posting our updates regularly here on Substack. Have fun and stay tuned.


If you want to stay up to date with ideas, projects and plans for the future at tensorwerk, subscribe to our publication and receive the Heartbeat directly in your inbox.

Jul '20 tensorwerk Heartbeat

Hey folks! The heat of July didn’t stop us ☀️ We successfully started our Hangar open discussion call series, our Sherin published a nice blog post about deep learning models deployment and we released a new version of Stockroom 🚀

Every month we are sharing news on projects we are working on, conferences and events we attend, what are our plans for the future and everything that might be related to data.

All you need is PyTorch, MLflow, RedisAI and a cup of mocha latte

Deploying a Deep Learning model is still a nightmare to you? RedisAI and MLflow are here for you 🤓 Our Sherin Thomas (aka @hhsecond) in his new blog post explains how easy is becoming a MLOps engineer with the right toolbox!

Check it out 👉 https://bit.ly/mlflow-redisai-deployment

Hangar open discussion calls recap

As promised in the last heartbeat, we successfully started our Public Hangar Developers meeting series!

We published all the recordings on our YouTube channel.

Here is a little recap of what happened:

July 07, 2020

🔗 Video Link

🗣️ Team presentation, Hangar introduction, Stockroom introduction, discussion about Hangar Bulk Importer

July 14, 2020

🔗 Video Link

🗣️ Hangar remote system, Users Q&A

July 22, 2020

🔗 Video Link

🗣️ Machine Learning Data Loaders for Hangar

July 28, 2020

🔗 Video Link

🗣️ Hangar Documentation and tutorials

See you every Tuesday at 11AM EST / 5PM CEST! Please come with questions, we'd love to hear your thoughts and/or issues 🧡

Stockroom 0.2.2 release

We presented Stockroom in our very first heartbeat, when it was just at an early stage. With the latest releases, we introduced a lot of interesting features, such as the ability to import torchvision datasets directly into a Stockroom repository. Import CIFAR-10, MNIST and FashionMNIST (more to come!) simply by using this command:

$ stock import [dataset]

For example, to import CIFAR-10, use:

$ stock import torchvision.cifar10

Check out the new Quick Start tutorial and enjoy your new versioning ally 🙌

Reach out

If you’d like to have a peek into our vision and our upcoming developments, please send us a note at info@tensorwerk.com. In any case, we will be posting our updates regularly here on Substack. Have fun and stay tuned.


If you want to stay up to date with ideas, projects and plans for the future at tensorwerk, subscribe to our publication and receive the Heartbeat directly in your inbox.

Jun '20 tensorwerk Heartbeat

This is our sixth Heartbeat (wow, time flies) and here there are some news about what we did (RedisAI plugin for MLFlow), what we are doing now (RedisAI for training infrastructure), and what we will be doing in the next future (Hangar open discussion calls).

Every month we are sharing news on projects we are working on, conferences and events we attend, what are our plans for the future and everything that might be related to data.

RedisAI plugin for MLFlow

With the latest release of MLFlow (1.9) is now possible to use the new RedisAI plugin and deploy an MLFlow model directly on RedisAI without any extra effort from the user!

Installable with:

pip install mlflow-redisai

Check out the plugin and the documentation at https://github.com/RedisAI/mlflow-redisai. Kudos to Sherin for pulling this off 👏

RedisAI for training infrastructure 🔜

As the anticipation of what you will hear a few months from now, we are working as part of a larger team to provide infrastructure for an upcoming (very cool) technology in the deep learning space. This project takes RedisAI into a new territory, where its ability to efficiently serve tensors and computations mixed with Redis’ unique features make it a perfect tool for the job.

We cannot reveal all the details yet but fasten your seatbelts and be prepared to hear more 😉

Hangar open discussion calls

We want the Hangar community to be involved in the development. We want the Hangar community to be curious. We want the Hangar community to ask questions. We know that it will let us deliver a better product.

That’s why we are starting a series of discussion calls open to the community. We are sharing our technical conversations with you, in order to grow a more participative community. There will be calls where we discuss new features we want to integrate into Hangar; more “educational“ calls where we explain some internals to the curious; Q&A sessions dedicated to users; or maybe a combination of them 😉

It’s also a means to get to know you better, and hopefully also the other way around ❤️

Stay tuned on Twitter to know when we are starting, it will be very soon 📞

P.S.: We plan to record them and post them on YouTube, so don’t worry if you miss one.

Reach out

If you’d like to have a peek into our vision and our upcoming developments, please send us a note at info@tensorwerk.com. In any case, we will be posting our updates regularly here on Substack. Have fun and stay tuned.


If you want to stay up to date with ideas, projects and plans for the future at tensorwerk, subscribe to our publication and receive the Heartbeat directly in your inbox.

May '20 tensorwerk Heartbeat

Hey there! This is our fifth heartbeat and … what a month! We published our brand new website, we were at RedisConf 2020 presenting RedisAI GA 1.0 and our Luca has been interviewed twice (for a podcast and during a meetup). Read below to see what happened!

Every month we are sharing news on projects we are working on, conferences and events we attend, what are our plans for the future and everything that might be related to data.

Website

Now, with a fully-fledged professional website (many thanks to the Evoque team for the great work 👏), we can finally say to the world we are ready to be trusted as a solid company, able to compete for real-world challenges.

Now we have a wonderful contact form and we are ready to hear from you. Of course, we are keeping all of the other channels open, so feel free to get in touch also via email or Twitter.

Interview @ Chai Time Data Science

This month our Luca was invited for an interview on the Chai Time Data Science podcast, together with Eli Stevens and Thomas Viehmann. The topic of discussion was their new Manning book “Deep Learning with PyTorch“, expected to be published during August 2020. It has been recognized as the official PyTorch book and here you can read the Essential Excerpts from the book for free.

Talking about the idea and the proposal of writing a technical book on PyTorch, Luca said:

I really wanted to understand how things really work and it was a good chance to have to explain things to others. It’s a good opportunity to learn them yourself.

Here you can find the full interview (also available as only audio):

How OSS is changing the world? @ AWS User Group Pune

Our Luca has been interviewed also during a meetup of the AWS User Group Pune by Jayesh Ahire and the topic of conversation was Open Source Software.

Open-source software is a virtuous mechanism by which a piece of invention gets alive on its own, independently of the commercial exploitation. The viability and the guarantees you get around that software are prolonged over time because they’re not a result of the need of a company at a certain point in time (which can change anytime). And at the same time, it is an opportunity because it’s something that can be created unifying many minds, so it can attract minds from different backgrounds and something like this is impossible in a pure enterprise environment.

Watch the full interview:

Meet the people: Alessia Marcolini

Alessia (@viperale on Twitter) is a Computer Science undergraduate at UniTN and she is helping tensorwerk to grow the community behind the company and its products, writing tutorials, creating these heartbeats (hello 👋) and managing the social networks’ account.

Alessia is a Junior Research Assistant at MPBA lab @ Fondazione Bruno Kessler, working on machine learning / deep learning frameworks to integrate multiple medical imaging modalities and different clinical data to get more precise prognostic/diagnostic cancer biomarkers.

She is keen on contributing to the Python community: she is already a volunteer of the Italian Python Community since 2017, helping with the organisation of PyCon Italy (the national Python Conference, hosting each year 600+ international delegates). Since 2018, she also joined the organisation committee of EuroSciPy, the European Conference for Python in Science.

When not coding, she loves dancing (she has been studying hip hop for 15 years!) and drinking black tea and good gin.

Reach out

If you’d like to have a peek into our vision and our upcoming developments, please send us a note at info@tensorwerk.com. In any case, we will be posting our updates regularly here on Substack. Have fun and stay tuned.


If you want to stay up to date with ideas, projects and plans for the future at tensorwerk, subscribe to our publication and receive the Heartbeat directly in your inbox.

Loading more posts…