Possible bug in gradient clipping of deepq_learner (tf2 branch)

[Giullar]

baselines/baselines/deepq/deepq_learner.py

Lines 174 to 181 in b99a73a

	grads = tape.gradient(weighted_error, self.q_network.trainable_variables)
	if self.grad_norm_clipping:
	clipped_grads = []
	for grad in grads:
	clipped_grads.append(tf.clip_by_norm(grad, self.grad_norm_clipping))
	clipped_grads = grads
	grads_and_vars = zip(grads, self.q_network.trainable_variables)
	self.optimizer.apply_gradients(grads_and_vars)

In line 179, shouldn't it be:
grads = clipped_grads
instead of
clipped_grads = grads
?

[susanth-24]

yes, I think you are right. It should be grads = clipped_grads, as it is the list containing the gradients after clipping, which should be used for optimization.