Hi, I don’t understand why Model Training Tensorflow Graph is used when serving - don’t we already have trained model?
Doesn’t that graph represent the trained model itself…its defined layers and learned parameters?
1 Like