Welcome to the February Heartbeat! This month’s featured image is a DVC pipeline\ncreated by one of our users,\nwhich we think resembles a…
","html":"Welcome to the February Heartbeat! This month’s featured image is a DVC pipeline\ncreated by one of our users,\nwhich we think resembles a valentine. Here are some more highlights from our\nteam and our community:
\nOur team is growing! In early January, DVC gained two new folks: engineer\nSaugat Pachhai and data scientist\nElle O’Brien. Saugat, based in Nepal, will\nbe contributing to core DVC. Elle (that’s me!), currently in San Francisco, will\nbe leading data science projects and outreach with DVC.
\nWe’re gearing up for a spring full of talks about DVC projects, including\nnew up-and-coming features for data cataloging and continuous integration. Here\nare just a few events that have been added to our schedule:
\n![\"Machine](\"/uploads/images/2020-02-10/mlprague.jpg\")
\n ![\"DivOps](\"/uploads/images/2020-02-10/divops_logo.png\"){kind=logo}
\n ![\"Women](\"/uploads/images/2020-02-10/wids.jpeg\")
\n -Elle O’Brien was recently accepted to give a keynote at\nWomen in Data Science San Diego on May 9. The\ntalk is called “Packaging data and machine learning models for sharing.”
\n-Elle will also be speaking at Div Ops, a new online\nconference about (you guessed it) DevOps, on March 27.
\nLook out for more conference announcements soon- in our brand new community\npage! We’ve just launched a new hub for sharing\nevents, goings-ons, and ways to contribute to DVC.
\nOur users continue to put awesome things on the internet. Like this AI blogger\nwho isn’t afraid to wear his heart on his sleeve.
\n![\"My](\"/uploads/images/2020-02-10/musa_atlihan.jpeg\")
\n Musa Atlihan writes:
\n\n\nFrom my experience, whether it is a real-world data science project or it is a\ndata science competition, there are two major key components for success.\nThose components are API simplicity and reproducible pipelines. Since data\nscience means experimenting a lot in a limited time frame, first, we need\nmachine learning tools with simplicity and second, we need\nreliable/reproducible machine learning pipelines. Thanks to tools like Keras,\nLightGBM, and fastai we already have simple yet powerful tools for rapid model\ndevelopment. And thanks to DVC, we are building large projects with\nreproducible pipelines very easily.
\n
It’s cool how Musa puts DVC in context with libraries for model building. In a\nway, the libraries that have made it easier than ever to iterate through\ndifferent model architectures have increased the need for reproducibility in\nproportion.
\nMeanwhile in Germany, superusers Marcel Mikl and Bert Besser wrote\nanother seriously\ncomprehensive article about DVC for Codecentric. Marcel and Bert walk readers\nthrough the steps to build a custom machine learning training pipeline with\nremote computing resources like GCP and AWS. It’s an excellent guide to\nconfiguring model training with attention to automation and collaboration.\nWe give them 🦉🦉🦉🦉🦉 out of 5.
\n![\"Remote](\"/uploads/images/2020-02-10/marcel.png\")
\n Here are a few more stories on our radar:
\nDVC at the University. A recent ..\npre-print from natural language processing researchers at Université Laval\nexplains how DVC facilitated dataset access for collaborators.
\n\n\n“In our case, the original dataset takes up to 6 Gigabytes. The previous way\nof retrieving the dataset over the network with a standard 20 Mbits/sec\ninternet connexion took up to an hour to complete (including uncompressing\nthe data). Using DVC reduced the retrieval time of the dataset to 3 minutes\nover the network with the same internet connexion.”
\n
Thanks for sharing- this is a lovely result. Oh, and last…
\n![\"elvie\"](\"/static/f0e8a9d4e7525ba2c56504833e14c3cd/4362d/elvie.png\" "\\"elvie\\"")
\n