Fixed mistypes in hidden layers sizes in TRPO examples (rll#149)

ViktorM · dementrock · commit 393be079b830 · 2017-06-07T15:31:12.000-07:00
diff --git a/examples/trpo_gym_cartpole.py b/examples/trpo_gym_cartpole.py
@@ -16,7 +16,7 @@ def run_task(*_):
     policy = CategoricalMLPPolicy(
         env_spec=env.spec,
         # The neural network policy should have two hidden layers, each with 32 hidden units.
-        hidden_sizes=(8, 8)
+        hidden_sizes=(32, 32)
     )
 
     baseline = LinearFeatureBaseline(env_spec=env.spec)
diff --git a/examples/trpo_gym_pendulum.py b/examples/trpo_gym_pendulum.py
@@ -15,7 +15,7 @@ def run_task(*_):
     policy = GaussianMLPPolicy(
         env_spec=env.spec,
         # The neural network policy should have two hidden layers, each with 32 hidden units.
-        hidden_sizes=(8, 8)
+        hidden_sizes=(32, 32)
     )
 
     baseline = LinearFeatureBaseline(env_spec=env.spec)

Original file line number	Diff line number	Diff line change
`@@ -16,7 +16,7 @@ def run_task(*_):`
`16`	`16`	`policy = CategoricalMLPPolicy(`
`17`	`17`	`env_spec=env.spec,`
`18`	`18`	`# The neural network policy should have two hidden layers, each with 32 hidden units.`
`19`		`- hidden_sizes=(8, 8)`
	`19`	`+ hidden_sizes=(32, 32)`
`20`	`20`	`)`
`21`	`21`
`22`	`22`	`baseline = LinearFeatureBaseline(env_spec=env.spec)`
Original file line number	Diff line number	Diff line change
`@@ -15,7 +15,7 @@ def run_task(*_):`
`15`	`15`	`policy = GaussianMLPPolicy(`
`16`	`16`	`env_spec=env.spec,`
`17`	`17`	`# The neural network policy should have two hidden layers, each with 32 hidden units.`
`18`		`- hidden_sizes=(8, 8)`
	`18`	`+ hidden_sizes=(32, 32)`
`19`	`19`	`)`
`20`	`20`
`21`	`21`	`baseline = LinearFeatureBaseline(env_spec=env.spec)`