Machine Learning Basics
A practical set of notebooks on machine learning basics, implemented in both TF2.0 + Keras and PyTorch.
deep-learning natural-language-processing tensorflow pytorch basics foundation madewithml code keras tutorial


📓 Notebooks 🐍 Python 🔢 NumPy
🐼 Pandas TensorFlow PyTorch
📈 Linear Regression
📊 Logistic Regression
️🎛 Multilayer Perceptrons
🔎 Data & Models
🛠 Utilities
️✂️ Preprocessing
️🖼 Convolutional Neural Networks
👑 Embeddings
📗 Recurrent Neural Networks


  • 📚 Illustrative ML notebooks available in both TensorFlow 2.0 + Keras and PyTorch.
    • Should I pick TensorFlow or PyTorch? Choice of framework doesn’t matter! We see a lot of great projects that use either TensorFlow + Keras or PyTorch and there’s tremendous value is knowing how to at least read both. If you have to work with a specific framework because of work/team constraints, you absolutely need to be literate in both so you can reimplement what you need. Don’t dismiss a project because it's not in your framework, especially now when they all share so many similarities. Check out the basic lessons and choose what you find more intuitive/suitable but the most important thing is to work on projects and share them with the community.
    • Do I need to know both TensorFlow or PyTorch? It is very important to at least know how to read both frameworks because cutting edge research continues to use both frameworks. Luckily, they're both very easy to learn and very easy to rewrite in the other framework.
  • 💻 These are not just a set of tutorials where we just load a bunch of packages and apply it on preloaded datasets. We explain every concept in the notebooks with clean code, simple math and visualizations to make them as intuitive as possible.
  • 📓 If you prefer Jupyter Notebooks or want to add/fix content, check out the notebooks directory.

Next Steps

As you learn ML, it's important to work on projects, so check out Made With ML for inspiration and to create a profile to showcase your own projects! Showcase your projects because everyone has Coursera, Kaggle, and fastai on their resumes so you need to differentiate yourself by showing what you can do using those fantastic resources.

Sign up for your free account →

Don't forget to tag @GokuMohandas in your comment, otherwise they may not be notified.

AI Research @apple. Author @oreillymedia. ML Lead @Ciitizen. Alum @hopkinsmedicine and @gatech
Share this project
Similar projects
Interactive Analysis of Sentence Embeddings
Learn how to interactively explore sentence embedding and labels in Tensorflow Embedding Projector.
Train ALBERT for NLP with TensorFlow on Amazon SageMaker
To train BERT in 1 hour, we efficiently scaled out to 2,048 NVIDIA V100 GPUs by improving the underlying infrastructure, network, and ML framework.
How to Benchmark Models with Transformers
HuggingFace's Transformer library allows users to benchmark models for both TensorFlow 2 and PyTorch using the PyTorchBenchmark and TensorFlowBenchmark ...
NER model for 40 languages trained with the new TFTrainer
This model is a fine-tuned XLM-Roberta-base over the 40 languages proposed in XTREME from Wikiann.
Top collections