You Must Know
1. Responsive Construct 2. Flexible 3. Easily Trainable 4. Parallel Neural Network Training 5. Large Community 6. Open Source
1. Cross-validation 2. Unsupervised learning algorithms 3. Feature extraction:
1. Interactive 2. Mathematics 3. Intuitive 4. Lot of Interaction
1. Runs smoothly on both CPU and GPU 2. Supports almost all the models of a neural network 3. Incredibly expressive, flexible, and apt for innovative research 4. Easy to debug and explore
1. Hybrid Front-End 2. Distributed Training 3. Python First 4. Libraries And Tools
1. Very fast computation 2. Intuitive 3. Faster training 4. Will not produce errors
Eli5 supports wother libraries XGBoost, lightning, scikit-learn, and sklearn-crfsuite libraries.
1. Developed using NumPy 2. Provides all the efficient numerical routines
1. Tight integration with NumPy 2. Transparent use of a GPU 3. Efficient symbolic differentiation 4. Speed and stability optimizations 5. Dynamic C code generation
1. Easy Data Manipulation 2. Support for operations