Add comment to ghost trainer

ml-agents/mlagents/trainers/torch/networks.py

         if self.use_lstm:
             # Use only the back half of memories for critic
             actor_mem, critic_mem = torch.split(memories, self.memory_size // 2, -1)
-        all_net_inputs = [net_inputs]
+        all_net_inputs = []
         if critic_obs is not None:
             all_net_inputs.extend(critic_obs)
         value_outputs, critic_mem_out = self.critic(
             net_inputs, memories=actor_mem, sequence_length=sequence_length
         )
         log_probs, entropies = self.action_model.evaluate(encoding, masks, actions)
-        all_net_inputs = [net_inputs]
+        all_net_inputs = []
         if critic_obs is not None:
             all_net_inputs.extend(critic_obs)
         value_outputs, critic_mem_outs = self.critic(
             critic_mem = None
             actor_mem = None
 
-        all_net_inputs = [net_inputs]
+        all_net_inputs = []
         if critic_obs is not None:
             all_net_inputs.extend(critic_obs)