- OS Machine learning/ Deep learning platform
- Lazy evaluate during build, full evaluate during execution.
- TensorFlow Estimator API
- High level object oriented API
- Makes it easy to build models.
- Specifies predefined architectures, such as linear regressors or neural networks.
- Tf.layers, tf.losses, tf.metrics
- Reusable libraries for common model components.
- Python TensorFlow
- Provides Ops, which wrap C++ Kernels
- Can run on CPU, GPU, or TPU
- Kernels work on more than one platform.
- Feature Engineering
- Often means converting raw log file entries to tf.Example protocol buffers. See also tf.Transform