modify train_ray.py for compatibility

Author: yijiezh

src/sagemaker/rl/estimator.py

@@ -508,8 +508,11 @@ def default_metric_definitions(cls, toolkit):
             return [
                 {
                     "Name": "episode_reward_mean",
-                    "Regex": "episode_reward_mean: (%s)" % float_regex,
+                    "Regex": "episode_reward_mean: {}".format(float_regex),
+                },
+                {
+                    "Name": "episode_reward_max",
+                    "Regex": "episode_reward_max: {}".format(float_regex),
                 },
-                {"Name": "episode_reward_max", "Regex": "episode_reward_max: (%s)" % float_regex, },
             ]
         raise ValueError("An unknown RLToolkit enum was passed in. toolkit: {}".format(toolkit))

tests/data/ray_cartpole/train_ray.py

@@ -10,7 +10,7 @@
 config["num_gpus"] = int(os.environ.get("SM_NUM_GPUS", 0))
 checkpoint_dir = os.environ.get("SM_MODEL_DIR", "/Users/nadzeya/gym")
 config["num_workers"] = 1
-agent = ppo.PPOAgent(config=config, env="CartPole-v0")
+agent = ppo.PPOTrainer(config=config, env="CartPole-v0")
 
 # Can optionally call agent.restore(path) to load a checkpoint.