There is no save model option for DDPG

[Kuppharish]

I'm using the DDPG API to run my environment. It trains perfectly fine when I use the command:
python -m baselines.ddpg.main
But I can't store the progress as the API doesn't have a save option. I tried to edit the code to get the desirable results. I added tf.train.Saver.save(sess) at the end of the code and tf.train.Saver.restore after the with U.single_threaded_session() as sess: statement which throws out following error:

TypeError: restore() missing 1 required positional argument: 'save_path'

Please tell me what is going wrong here, Thanks.

[iSaran]

Take a look at #162, it is the same issue.

[brendenpetersen]

@Kuppharish Keep in mind that even if you save/restore the policy as in #162, if normalize_observations=True (the default value), then the mean/std used for normalization will not be recoverable.