[RL-baseline] Model v4, experiment #3

[ziritrion]

The policy network for model v4 for REINFORCE with Baseline is essentially the same network as in v2, but the actor and critic heads have an additional fully connected layer each similar to v3. This tweak was added with the hopes of seeing the initial gains in reward that we observed with model v3 but with the elevated sustained reward value that we observed in v2.

The action sets are the same as in Model v3. For this experiment, action set #2 is chosen:
[0.0, 0.0, 0.0], # no action
[0.0, 0.8, 0.0], # throttle
[0.0, 0.0, 0.6], # break
[-0.9, 0.0, 0.0], # left
[0.9, 0.0, 0.0], # right

Results were disappointing. Running Reward was negative for most of the experiment, with a maximum peak of 321 right before the 2k episode mark but quickly dropping afterwards. Entropy and Loss function both collapsed before the 12k episode mark.

Results are below:

Sample video below:
https://user-images.githubusercontent.com/1465235/113128553-34593a80-921a-11eb-8e61-6e3150f119bd.mp4