[rllib] Add more regression tests and autogenerate (#2324)

python/ray/rllib/tuned_examples/generate_regression_tests.py

@@ -0,0 +1,29 @@
+#!/usr/bin/env python
+# This script generates all the regression tests for RLlib.
+
+import glob
+import re
+import os
+import os.path as osp
+
+
+CONFIG_DIR = osp.join(osp.dirname(osp.abspath(__file__)), "regression_tests")
+
+
+TEMPLATE = """
+class Test{name}(Regression):
+    _file = "{filename}"
+
+    def setup_cache(self):
+        return _evaulate_config(self._file)
+
+"""
+
+if __name__ == '__main__':
+    os.chdir(CONFIG_DIR)
+
+    with open("regression_test.py", "a") as f:
+        for filename in sorted(glob.glob("*.yaml")):
+            splits = re.findall(r"\w+", osp.splitext(filename)[0])
+            test_name = "".join([s.capitalize() for s in splits])
+            f.write(TEMPLATE.format(name=test_name, filename=filename))

python/ray/rllib/tuned_examples/regression_tests/cartpole-a3c-pytorch.yaml

@@ -1,6 +1,7 @@
 cartpole-a3c:
     env: CartPole-v0
     run: A3C
+    repeat: 3
     stop:
         episode_reward_mean: 200
         time_total_s: 600

python/ray/rllib/tuned_examples/regression_tests/cartpole-a3c.yaml

@@ -1,6 +1,7 @@
 cartpole-a3c:
     env: CartPole-v0
     run: A3C
+    repeat: 3
     stop:
         episode_reward_mean: 200
         time_total_s: 600

python/ray/rllib/tuned_examples/regression_tests/cartpole-dqn.yaml

@@ -1,6 +1,7 @@
 cartpole-dqn:
     env: CartPole-v0
     run: DQN
+    repeat: 3
     stop:
         episode_reward_mean: 200
         time_total_s: 600

python/ray/rllib/tuned_examples/regression_tests/cartpole-pg.yaml

@@ -1,6 +1,7 @@
 cartpole-pg:
     env: CartPole-v0
     run: PG
+    repeat: 3
     stop:
         episode_reward_mean: 200
         time_total_s: 300

python/ray/rllib/tuned_examples/regression_tests/cartpole-ppo.yaml

@@ -1,6 +1,7 @@
 cartpole-ppo:
     env: CartPole-v0
     run: PPO
+    repeat: 3
     stop:
         episode_reward_mean: 200
         time_total_s: 300

python/ray/rllib/tuned_examples/regression_tests/pendulum-ddpg.yaml

@@ -1,6 +1,7 @@
 pendulum-ddpg:
     env: Pendulum-v0
     run: DDPG
+    repeat: 3
     stop:
         episode_reward_mean: -160
         time_total_s: 900

python/ray/rllib/tuned_examples/regression_tests/pendulum-ppo.yaml

@@ -0,0 +1,19 @@
+pendulum-ppo:
+    env: Pendulum-v0
+    run: PPO
+    repeat: 3
+    stop:
+        episode_reward_mean: -160
+        # expect -140 within 300-500k steps
+        timesteps_total: 600000
+    config:
+        timesteps_per_batch: 2048
+        num_workers: 4
+        lambda: 0.1
+        gamma: 0.95
+        sgd_stepsize: 0.0003
+        sgd_batchsize: 64
+        num_sgd_iter: 10
+        model:
+            fcnet_hiddens: [64, 64]
+            squash_to_range: True

python/ray/rllib/tuned_examples/regression_tests/regression_test.py

@@ -51,52 +51,3 @@ class Regression():
 
     def track_iterations(self, result):
         return result["training_iteration"]
-
-
-class TestCartPolePPO(Regression):
-    _file = "cartpole-ppo.yaml"
-
-    def setup_cache(self):
-        return _evaulate_config(self._file)
-
-
-class TestCartPolePG(Regression):
-    _file = "cartpole-pg.yaml"
-
-    def setup_cache(self):
-        return _evaulate_config(self._file)
-
-
-class TestPendulumDDPG(Regression):
-    _file = "pendulum-ddpg.yaml"
-
-    def setup_cache(self):
-        return _evaulate_config(self._file)
-
-
-class TestCartPoleES(Regression):
-    _file = "cartpole-es.yaml"
-
-    def setup_cache(self):
-        return _evaulate_config(self._file)
-
-
-class TestCartPoleDQN(Regression):
-    _file = "cartpole-dqn.yaml"
-
-    def setup_cache(self):
-        return _evaulate_config(self._file)
-
-
-class TestCartPoleA3C(Regression):
-    _file = "cartpole-a3c.yaml"
-
-    def setup_cache(self):
-        return _evaulate_config(self._file)
-
-
-class TestCartPoleA3CPyTorch(Regression):
-    _file = "cartpole-a3c-pytorch.yaml"
-
-    def setup_cache(self):
-        return _evaulate_config(self._file)