IntelLabs / coach

Reinforcement Learning Coach by Intel AI Lab enables easy experimentation with state of the art Reinforcement Learning algorithms
https://intellabs.github.io/coach/
Apache License 2.0
2.33k stars 461 forks source link

Using custom tf.keras model #459

Open patrickhegemann opened 4 years ago

patrickhegemann commented 4 years ago

Is it possible to use a custom tf.keras model I made as the network or middleware in e.g. the Rainbow agent? My network takes images as input and has as many output neurons as there are actions. I haven't found how to use a custom network in the documentation or tutorial.