January 2020
Intermediate to advanced
432 pages
10h 18m
English
There are a number of DL frameworks available for use, but only a few have been readily used in RL research or projects. Most of these frameworks share a number of similarities, so transferring knowledge from one to the other is relatively straightforward. The three most popular frameworks for RL in the past few years have been Keras, TensorFlow, and PyTorch.
A summary of the strengths and weaknesses of each framework is shown in the table here:
| Framework | Keras | TensorFlow | PyTorch |
| Accessibility | Easiest to learn and use | Provides a high-level Keras interface and lower-level interface. | Medium to low-level interface |
| Scalability | Scales well for smaller projects | Scales to any size project and supported output network ... |
Read now
Unlock full access