Will now print summaries even when not training or when training is over (#1020)

* [Initial Commit]

* [Addressed comments]

* [Now using global step to write the summaries]

python/unitytrainers/trainer.py

     @property
     def get_step(self):
         """
-        Returns the number of steps the trainer has performed
+        Returns the number of training steps the trainer has performed
         :return: the step count of the trainer
         """
         raise UnityTrainerException("The get_step property was not implemented.")
         """
         raise UnityTrainerException("The update_model method was not implemented.")
 
-    def write_summary(self, lesson_num=0):
+    def write_summary(self, global_step, lesson_num=0):
-        :param lesson_num: The lesson the trainer is at.
+        :param global_step: The number of steps the simulation has been going for
+        :param lesson_number: The lesson the trainer is at.
-        if (self.get_step % self.trainer_parameters['summary_freq'] == 0 and self.get_step != 0 and
-                self.is_training and self.get_step <= self.get_max_steps):
+        if global_step % self.trainer_parameters['summary_freq'] == 0 and global_step != 0:
+            is_training = "Training." if self.is_training and self.get_step <= self.get_max_steps else "Not Training."
-                logger.info("{}: {}: Step: {}. Mean Reward: {:0.3f}. Std of Reward: {:0.3f}."
-                            .format(self.run_id, self.brain_name, self.get_step,
-                                    mean_reward, np.std(self.stats['cumulative_reward'])))
+                logger.info(" {}: {}: Step: {}. Mean Reward: {:0.3f}. Std of Reward: {:0.3f}. {}"
+                            .format(self.run_id, self.brain_name, min(self.get_step, self.get_max_steps),
+                                    mean_reward, np.std(self.stats['cumulative_reward']), is_training))
-                logger.info("{}: {}: Step: {}. No episode was completed since last summary."
-                            .format(self.run_id, self.brain_name, self.get_step))
+                logger.info(" {}: {}: Step: {}. No episode was completed since last summary. {}"
+                            .format(self.run_id, self.brain_name, self.get_step, is_training))
             summary = tf.Summary()
             for key in self.stats:
                 if len(self.stats[key]) > 0:

python/unitytrainers/trainer_controller.py

                             trainer.update_model()
                         # Write training statistics to Tensorboard.
                         if self.meta_curriculum is not None:
-                            trainer.write_summary(lesson=self.meta_curriculum.brains_to_curriculums[brain_name].lesson_num)
+                            trainer.write_summary(
+                                global_step,
+                                lesson=self.meta_curriculum.brains_to_curriculums[brain_name].lesson_num)
-                            trainer.write_summary()
+                            trainer.write_summary(global_step)
-                    if self.train_model:
-                        global_step += 1
+                    global_step += 1
                     if global_step % self.save_freq == 0 and global_step != 0 and self.train_model:
                         # Save Tensorflow model
                         self._save_model(sess, steps=global_step, saver=saver)