I'm not sure if the machine learning engineer role is very well-defined. IMO, a good MLE does "full-stack" work -- owning ML end-to-end, from model development to integration in production pipelines. I interview for both MLE and data science roles. Here's what I look for:

A thing I hear sometimes: "what if my loss doesn't match the signature loss = fn(y_true, y_pred)?" This is not a requirement in Keras -- it's only the default setting. If you have a loss with multiple inputs/targets, here are your options, in order of preference: (a thread)

We now have 44 high-quality, short code examples demonstrating common workflows at keras.io/examples/ Most of them are contributed by the community. They cover computer vision, NLP, timeseries, audio, generative deep learning, RL...

Amazing article from Eugene Yan describing developments in NLP from RNNs(1985) to Big Bird(2020). Really helps get a bird's-eye view of this rapidly progressing field. eugeneyan.com/writing/nlp-su…

Wrote a short post along with a colab notebook showing how you can use MarianMT models in Hugging Face for text data augmentation. amitness.com/back-translati…

I recently discovered an approach to run VSCode on Google Colab directly. With the setup, you can use the notebook and the IDE side-by-side. amitness.com/vscode-on-cola…

1/ "Compound interest is the eighth wonder of the world. He who understands it, earns it. He who doesn't, pays it." — Albert Einstein It affects: Your health Your money Your relationships Your reputation Your abilities

⭐New blogpost⭐: Hyperparameter tuning with Keras and Ray Tune A practical tutorial on using HyperOpt’s Bayesian optimization with HyperBand scheduler to choose the best hyperparameters for machine learning models. Check out the blog: medium.com/@himanshurawla… ray

This is one of the main tricks I recommend when trying to improve current model performance. Don't start with hyperparameter search, start by looking at individual examples with large losses, and most of the time you'll understand what feature your model is missing.

Our coaching classes spoon feeding culture and copy paste xerox of notes has resulted in a entire with terrible research skills.

If you're looking for the best resources to prepare for ML interviews in 2020, Here's a wiki from the fast.ai forums: forums.fast.ai/t/wiki-2020-ml… Also, all contributions are welcomed!

Publishing proof > sending resumes.

Work on things that keep working after you stop working.

This is how you make a Rangoli !!!

I wish I had a third person view of my life. Since it's easy to point out flaws than to realize doing the same.

If you get a hard time saving money, 💵 Think of the savings amount ~as~ 💸 Money spent in buying FREEDOM💪 for your future You'll become more interested in saving a good amount now, and live financially free just a few years later. #PersonalFinanceUnder25

13/n The mind, by its very nature, is unruly, untamed and wild. It loves to dwell in imaginary futures or live in the past. It detests the present moment. Because when the mind is truly in the moment, it dies. It disappears. A state of no-mind.

The first person invested 12,000. In the first six years. The second person invested 74,000. After chilling for six years. At the end, both are about the same. The magic of being the early bird!

Many machine learning papers and tutorials start from labeled datasets. But what if we don't have labels? How do we collect them? Here we discuss semi-supervised learning, active learning, and weak supervision to bootstrap labels for your ML system. eugeneyan.com/writing/bootst…

Announcement: my book Deep Learning with Python (2nd edition) has been released. 500 pages of code examples, theory, context, practical tips... If you want to really understand how deep learning works, why it matters, and how to use it, this is your book! manning.com/books/deep-lea…