RNN tutorial

RNN tutorials for the ErUM-Data-Hub Deep Learning School https://indico.scc.kit.edu/event/2851/

We will use Exercises from Chapter 9 of http://deeplearningphysics.org and additional exercises from this repository:

1. Understand the keras RNN implementations: understand_rnns.ipynb
2. Predict a sine curve - Exercise_09_1.ipynb from Deep Learning for Physics Research

# Technical hint: Exercise_09_1 runs sufficiently fast on CPU
# but it may be beneficial to restrict the number of cores in the first notebook cell via
import tensorflow as tf
tf.config.threading.set_intra_op_parallelism_threads(1)
tf.config.threading.set_inter_op_parallelism_threads(1)

3. Identify cosmic ray signals - Exercise_09_2.ipynb from Deep Learning for Physics Research - recommended to run on GPU
4. Work with variable length sequences - variable_length_masking.ipynb - recommended to run on GPU