Experience training, evaluating, and deploying models end-to-end (data preprocessing → training → validation → inference → monitoring)
Open-source Big data processing and analytics
Open-source
For large-scale production Static computation graphs
Declarative
Works with multidimensional immutable arrays (which they call tensors)
Layers are objects with the following methods
Some standard layers are
If you're working with images, normalize the data, so that the pixels are all given values between 0 and 1
Loss Function
Optimization
Metrics
Feed in the data
model.fit(train_images, train_labels, epochs = 20)
Evaluate performance
test_loss, test_acc = model.evaluate(test_images, test_labels, verbose = 2)
print('\nTest accuracy:', test_acc)
Softmax
Converts the logits to probabilities
Then use
probability_model.predict(test_images)
To see how confident the model is
Open-source ML library
Dynamic computation graphs
Imperative
Open-source numerical computation libraryn
Just-in-time compilation