Keras
Notes
- What if my loss doesn't match the signature loss?
- Could use JAX to write Keras layers? It's as simple as a call to
jax2tf.convert()
. XLA is the shared language that makes this works.
Links
- Keras Attention Mechanism - Attention mechanism Implementation for Keras.
- Keras Tuner - Hyperparameter tuning for humans.
- Build a scalable, online recommender with Keras, Docker, GCP, and GKE (2020)
- Everything you need to know to use Keras & TF 2.0 to build real-world machine learning solutions
- Keras debugging tips (Twitter)
- Keras Examples
- Working with preprocessing layers
- Keras for Beginners: Implementing a Recurrent Neural Network (2020)
- ActivationFunctions using Custom Layers in Keras
- Training Keras models with TensorFlow Cloud
- Keras Idiomatic Programmer
- Scikit-Learn Wrapper for Keras
- aXeleRate - Keras-based framework for AI on the Edge.
- Insightface Keras implementation
- Conditional RNN (Tensorflow Keras)
- MobileViT: A mobile-friendly Transformer-based model for image classification (Tweet)
- Keras Visualization Toolkit
- Keras in Rust
- Keras implement of flow-based models
- Keras/TensorFlow attention models
- Keras Hypetune - Friendly python package for Keras Hyperparameters Tuning based only on NumPy and Hyperopt.
- KerasNLP - Industry-strength Natural Language Processing workflows with Keras.
- Keras API Docs (Code)