diff --git a/scratch/etienne/trpo/experiments/sgail-ppo-options-setobs2.py b/scratch/etienne/trpo/experiments/sgail-ppo-options-setobs2.py
index 52a05a0..efbe7dc 100644
--- a/scratch/etienne/trpo/experiments/sgail-ppo-options-setobs2.py
+++ b/scratch/etienne/trpo/experiments/sgail-ppo-options-setobs2.py
@@ -17,38 +17,38 @@ from src.safe_options.options import SafeOptionsEnv
 from torch.utils.tensorboard import SummaryWriter
 from ray import tune
 
-obs_min = np.array([
-    [-1000, -1000, 0, -np.pi, -1e-1, 0.],
-    [0, -np.pi, -20, -20, -np.pi, -1e-1],
-    [0, -np.pi, -20, -20, -np.pi, -1e-1],
-    [0, -np.pi, -20, -20, -np.pi, -1e-1],
-    [0, -np.pi, -20, -20, -np.pi, -1e-1],
-    [0, -np.pi, -20, -20, -np.pi, -1e-1],
-]).reshape(-1)
-
-obs_max = np.array([
-    [1000, 1000, 20, np.pi, 1e-1, 0.],
-    [50, np.pi, 20, 20, np.pi, 1e-1],
-    [50, np.pi, 20, 20, np.pi, 1e-1],
-    [50, np.pi, 20, 20, np.pi, 1e-1],
-    [50, np.pi, 20, 20, np.pi, 1e-1],
-    [50, np.pi, 20, 20, np.pi, 1e-1],
-]).reshape(-1)
-
-envs = [SafeOptionsEnv(Setobs(
-    TransformObservation(CollisionPenaltyWrapper(IntersimpleLidarFlatRandom(
-        n_rays=5,
-        reward=functools.partial(
-            speed_reward,
-            collision_penalty=0
-        ),
-        stop_on_collision=True,
-    ), collision_distance=6, collision_penalty=100), lambda obs: (obs - obs_min) / (obs_max - obs_min + 1e-10))
-), options=[(0, 5), (1, 5), (2, 5), (4, 5), (6, 5), (8, 5), (10, 5)], safe_actions_collision_method='circle', abort_unsafe_collision_method='circle') for _ in range(60)]
-
-env_fn = lambda i: envs[i]
-
 def training_function(config):
+    obs_min = np.array([
+        [-1000, -1000, 0, -np.pi, -1e-1, 0.],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+    ]).reshape(-1)
+
+    obs_max = np.array([
+        [1000, 1000, 20, np.pi, 1e-1, 0.],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+    ]).reshape(-1)
+
+    envs = [SafeOptionsEnv(Setobs(
+        TransformObservation(CollisionPenaltyWrapper(IntersimpleLidarFlatRandom(
+            n_rays=5,
+            reward=functools.partial(
+                speed_reward,
+                collision_penalty=0
+            ),
+            stop_on_collision=True,
+        ), collision_distance=6, collision_penalty=100), lambda obs: (obs - obs_min) / (obs_max - obs_min + 1e-10))
+    ), options=[(0, 5), (1, 5), (2, 5), (4, 5), (6, 5), (8, 5), (10, 5)], safe_actions_collision_method='circle', abort_unsafe_collision_method='circle') for _ in range(60)]
+
+    env_fn = lambda i: envs[i]
+
     policy = SetMaskedDiscretePolicy(env_fn(0).action_space.n) # config net architecture
     pi_opt = torch.optim.Adam(policy.parameters(), lr=3e-5)  # config learning rate
     pi_lr_scheduler = torch.optim.lr_scheduler.ExponentialLR(pi_opt, gamma=0.98) # config lr decay
diff --git a/sgail-ppo-options-setobs2.py b/sgail-ppo-options-setobs2.py
new file mode 100644
index 0000000..04049ee
--- /dev/null
+++ b/sgail-ppo-options-setobs2.py
@@ -0,0 +1,117 @@
+# %%
+import os
+
+import gym
+from src.safe_options.options import gail_ppo, Buffer
+from src.core.value import SetValue
+from src.safe_options.policy import SetMaskedDiscretePolicy
+from src.core.discriminator import DeepsetDiscriminator
+import torch.optim
+from intersim.envs import IntersimpleLidarFlatRandom
+from intersim.envs.intersimple import speed_reward
+import functools
+from src.util.wrappers import CollisionPenaltyWrapper, TransformObservation, Setobs
+import numpy as np
+from src.safe_options.options import SafeOptionsEnv
+from torch.utils.tensorboard import SummaryWriter
+from ray import tune
+
+def training_function(config):
+    obs_min = np.array([
+        [-1000, -1000, 0, -np.pi, -1e-1, 0.],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+        [0, -np.pi, -20, -20, -np.pi, -1e-1],
+    ]).reshape(-1)
+
+    obs_max = np.array([
+        [1000, 1000, 20, np.pi, 1e-1, 0.],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+        [50, np.pi, 20, 20, np.pi, 1e-1],
+    ]).reshape(-1)
+
+    envs = [SafeOptionsEnv(Setobs(
+        TransformObservation(CollisionPenaltyWrapper(IntersimpleLidarFlatRandom(
+            n_rays=5,
+            reward=functools.partial(
+                speed_reward,
+                collision_penalty=0
+            ),
+            stop_on_collision=True,
+        ), collision_distance=6, collision_penalty=100), lambda obs: (obs - obs_min) / (obs_max - obs_min + 1e-10))
+    ), options=[(0, 5), (1, 5), (2, 5), (4, 5), (6, 5), (8, 5), (10, 5)], safe_actions_collision_method='circle', abort_unsafe_collision_method='circle') for _ in range(60)]
+
+    env_fn = lambda i: envs[i]
+
+    policy = SetMaskedDiscretePolicy(env_fn(0).action_space.n) # config net architecture
+    pi_opt = torch.optim.Adam(policy.parameters(), lr=3e-5)  # config learning rate
+    pi_lr_scheduler = torch.optim.lr_scheduler.ExponentialLR(pi_opt, gamma=0.98) # config lr decay
+
+    value = SetValue() # config net architecture
+    v_opt = torch.optim.Adam(value.parameters(), lr=1e-3) # config lr
+
+    discriminator = DeepsetDiscriminator() # config net architecture
+    disc_opt = torch.optim.Adam(discriminator.parameters(), lr=1e-3, weight_decay=1e-3) # config lr, weight decay
+
+    expert_data = torch.load(os.path.join(os.path.dirname(os.path.abspath(__file__)), 'intersimple-expert-data-setobs2.pt'))
+    expert_data = Buffer(*expert_data)
+
+    def callback(info):
+        tune.report(gen_mean_reward_per_episode=info['gen/mean_reward_per_episode'])
+
+    value, policy = gail_ppo(
+        env_fn=env_fn,
+        expert_data=expert_data,
+        discriminator=discriminator,
+        disc_opt=disc_opt,
+        disc_iters=100, # config
+        policy=policy,
+        value=value,
+        v_opt=v_opt,
+        v_iters=1000, # config
+        epochs=200,
+        rollout_episodes=60,
+        rollout_steps=60,
+        gamma=0.99,
+        gae_lambda=0.9,
+        clip_ratio=0.2, # config
+        pi_opt=pi_opt,
+        pi_iters=100, # config
+        logger=SummaryWriter(comment='sgail-ppo-options-setobs2'),
+        callback=callback,
+        lr_schedulers=[pi_lr_scheduler],
+    )
+
+analysis = tune.run(
+    training_function,
+    config={
+        'dummy': tune.grid_search([0.001, 0.01, 0.1]),
+    }
+)
+
+print('Best config: ', analysis.get_best_config(metric='gen_mean_reward_per_episode', mode='min'))
+
+# %%
+# policy = SetMaskedDiscretePolicy(env_fn(0).action_space.n)
+# policy(torch.zeros(env_fn(0).observation_space['observation'].shape), torch.zeros(env_fn(0).observation_space['safe_actions'].shape))
+# policy.load_state_dict(torch.load('sgail-ppo-options-setobs2.pt'))
+
+# env = env_fn(0)
+# obs = env.reset()
+# env.render(mode='post')
+# for i in range(300):
+#     action = policy.sample(policy(
+#         torch.tensor(obs['observation'], dtype=torch.float32),
+#         torch.tensor(obs['safe_actions'], dtype=torch.float32),
+#     ))
+#     obs, reward, done, _ = env.step(action, render_mode='post')
+#     print('step', i, 'reward', reward, 'safe actions', obs['safe_actions'])
+#     if done:
+#         break
+# env.close()
+# %%