huangfu/InteractionImitation
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Track collision rate

Browse Source
This commit is contained in:
ebuehrle
2022-02-26 13:50:53 +01:00
parent dc9cbf329b
commit 6876cf9625
2 changed files with 10 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
sgail-ppo-options-setobs2.py
View File

@@ -45,6 +45,7 @@ def training_function(config):
speed_reward,
collision_penalty=0
),
check_collisions=True,
stop_on_collision=config['env']['stop_on_collision'],
), 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)],

12
src/safe_options/options.py
View File

@@ -47,7 +47,7 @@ def gail(env_fn, expert_data, discriminator, disc_opt, disc_iters, policy, value
logger.add_scalar('expert/mean_reward_per_episode', expert_data.rewards[~expert_data.dones].sum() / expert_data.states.shape[0])
for epoch in tqdm(range(epochs)):
hl_data, ll_data = rollout(env_fn, policy, rollout_episodes, rollout_steps)
hl_data, ll_data, collisions = rollout(env_fn, policy, rollout_episodes, rollout_steps)
generator_data = OptionsRollout(HLBuffer(*hl_data), Buffer(*ll_data))
generator_data.ll.actions += 0.1 * torch.randn_like(generator_data.ll.actions)
@@ -56,6 +56,7 @@ def gail(env_fn, expert_data, discriminator, disc_opt, disc_iters, policy, value
gen_mean_reward_per_episode = generator_data.hl.rewards[~generator_data.hl.dones].sum() / generator_data.hl.states.shape[0]
logger.add_scalar('gen/mean_reward_per_episode', gen_mean_reward_per_episode, epoch)
logger.add_scalar('gen/unsafe_probability_mass', policy.unsafe_probability_mass(policy(generator_data.hl.states[~generator_data.hl.dones], generator_data.hl.safe_actions[~generator_data.hl.dones])).mean(), epoch)
logger.add_scalar('gen/collision_rate', (1. * collisions.any(-1)).mean(), epoch)
discriminator, loss = train_discriminator(expert_data, generator_data.ll, discriminator, disc_opt, disc_iters, wasserstein, wasserstein_c)
if wasserstein:
@@ -89,7 +90,7 @@ def gail_ppo(env_fn, expert_data, discriminator, disc_opt, disc_iters, policy, v
logger.add_scalar('expert/mean_reward_per_episode', expert_data.rewards[~expert_data.dones].sum() / expert_data.states.shape[0])
for epoch in range(epochs):
hl_data, ll_data = rollout(env_fn, policy, rollout_episodes, rollout_steps)
hl_data, ll_data, collisions = rollout(env_fn, policy, rollout_episodes, rollout_steps)
generator_data = OptionsRollout(HLBuffer(*hl_data), Buffer(*ll_data))
generator_data.ll.actions += 0.1 * torch.randn_like(generator_data.ll.actions)
@@ -98,6 +99,7 @@ def gail_ppo(env_fn, expert_data, discriminator, disc_opt, disc_iters, policy, v
gen_mean_reward_per_episode = generator_data.hl.rewards[~generator_data.hl.dones].sum() / generator_data.hl.states.shape[0]
logger.add_scalar('gen/mean_reward_per_episode', gen_mean_reward_per_episode, epoch)
logger.add_scalar('gen/unsafe_probability_mass', policy.unsafe_probability_mass(policy(generator_data.hl.states[~generator_data.hl.dones], generator_data.hl.safe_actions[~generator_data.hl.dones])).mean(), epoch)
logger.add_scalar('gen/collision_rate', (1. * collisions.any(-1)).mean(), epoch)
discriminator, loss = train_discriminator(expert_data, generator_data.ll, discriminator, disc_opt, disc_iters, wasserstein, wasserstein_c)
if wasserstein:
@@ -134,6 +136,7 @@ def rollout(env_fn, policy, n_episodes, max_steps_per_episode):
actions = torch.zeros(n_episodes, max_steps_per_episode + 1, *env.action_space.shape)
rewards = torch.zeros(n_episodes, max_steps_per_episode + 1)
dones = torch.ones(n_episodes, max_steps_per_episode + 1, dtype=bool)
collisions = torch.zeros(n_episodes, max_steps_per_episode, dtype=bool)
ll_states = torch.zeros(n_episodes, max_steps_per_episode, env.max_plan_length + 1, *env.observation_space['observation'].shape)
ll_actions = torch.zeros(n_episodes, max_steps_per_episode, env.max_plan_length + 1, *env.ll_action_space.shape)
@@ -159,6 +162,9 @@ def rollout(env_fn, policy, n_episodes, max_steps_per_episode):
safe_actions[:, s + 1] = torch.tensor(o['safe_actions']).clone().detach()
rewards[:, s] = torch.tensor(r).clone().detach()
dones[:, s + 1] = torch.tensor(d).clone().detach()
collisions[:, s] = torch.from_numpy(np.stack([
any(k['collision'] for k in i['ll']['infos']) for i in info
])).detach().clone()
ll_states[:, s] = torch.from_numpy(np.stack([i['ll']['observations'] for i in info])).clone().detach()
ll_actions[:, s] = torch.from_numpy(np.stack([i['ll']['actions'] for i in info])).clone().detach()
@@ -173,7 +179,7 @@ def rollout(env_fn, policy, n_episodes, max_steps_per_episode):
rewards = rewards[:, :max_steps_per_episode]
dones = dones[:, :max_steps_per_episode]
return (states, safe_actions, actions, rewards, dones), (ll_states, ll_actions, ll_rewards, ll_dones)
return (states, safe_actions, actions, rewards, dones), (ll_states, ll_actions, ll_rewards, ll_dones), collisions
class SafeOptionsEnv(OptionsEnv):