Skip to:
Next phase of GNN:
Multiple concurrent models (2.0)
Support for non-sequential Keras models (2.0)
Support use of GPUs (2.0)
Support Learning Rate Decay during training (2.0)
Support Batch Size reduction during training (2.0)
Support Tensorflow Version 2.x as well as 1.x (2.0)
Implement 2-stage distribution (Distributed Gradient Descent / Results averaging)
Encapsulation of Fitted Model as a single dataset
Support additional tensor types (e.g. Real8)
Next phase of GNN:
Multiple concurrent models (2.0)
Support for non-sequential Keras models (2.0)
Support use of GPUs (2.0)
Support Learning Rate Decay during training (2.0)
Support Batch Size reduction during training (2.0)
Support Tensorflow Version 2.x as well as 1.x (2.0)
Implement 2-stage distribution (Distributed Gradient Descent / Results averaging)
Encapsulation of Fitted Model as a single dataset
Support additional tensor types (e.g. Real8)