bb857bb2469689da4d2e19473049e47ccc1be04a,examples/reinforcement_learning/tutorial_prioritized_replay.py,DQN,train,#DQN#,432

Before Change

        self.niter += 1
        if self.niter % target_q_update_freq == 0:
            sync(self.qnet, self.targetqnet)
            path = os.path.join(args.save_path, "{}.npz".format(self.niter))
            tl.files.save_npz(self.qnet.trainable_weights, name=path)

        return td_errors.numpy()

After Change

        self.niter += 1
        if self.niter % target_q_update_freq == 0:
            sync(self.qnet, self.targetqnet)
            self.save(args.save_path)
        return td_errors.numpy()

    def save(self, path):

Instances