gaspersic/baselines
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

B -> nenvs for consistency with other algos, small cleanups

Browse Source
This commit is contained in:
Peter Zhokhov
2018-10-15 18:29:48 -07:00
parent 4bca9158a1
commit 7b33af0395
1 changed files with 9 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
baselines/ddpg/ddpg.py
View File

@@ -99,10 +99,10 @@ def learn(network, env,
obs = env.reset()
if eval_env is not None:
eval_obs = eval_env.reset()
B = obs.shape[0]
nenvs = obs.shape[0]
episode_reward = np.zeros(B, dtype = np.float32) #vector
episode_step = np.zeros(B, dtype = int) # vector
episode_reward = np.zeros(nenvs, dtype = np.float32) #vector
episode_step = np.zeros(nenvs, dtype = int) # vector
episodes = 0 #scalar
t = 0 # scalar
@@ -120,7 +120,7 @@ def learn(network, env,
for epoch in range(nb_epochs):
for cycle in range(nb_epoch_cycles):
# Perform rollouts.
if B > 1:
if nenvs > 1:
# if simulating multiple envs in parallel, impossible to reset agent at the end of the episode in each
# of the environments, so resetting here instead
agent.reset()
@@ -132,7 +132,7 @@ def learn(network, env,
if rank == 0 and render:
env.render()
# max_action is of dimension A, whereas action is dimension (B,A) - the multiplication gets broadcasted to the batch
# max_action is of dimension A, whereas action is dimension (nenvs, A) - the multiplication gets broadcasted to the batch
new_obs, r, done, info = env.step(max_action * action) # scale for execution in env (as far as DDPG is concerned, every action is in [-1, 1])
# note these outputs are batched from vecenv
@@ -150,7 +150,7 @@ def learn(network, env,
obs = new_obs
for d in range(len(done)):
if done[d] == True:
if done[d]:
# Episode done.
epoch_episode_rewards.append(episode_reward[d])
episode_rewards_history.append(episode_reward[d])
@@ -159,7 +159,7 @@ def learn(network, env,
episode_step[d] = 0
epoch_episodes += 1
episodes += 1
if B == 1:
if nenvs == 1:
agent.reset()
@@ -183,8 +183,8 @@ def learn(network, env,
eval_episode_rewards = []
eval_qs = []
if eval_env is not None:
B = eval_obs.shape[0]
eval_episode_reward = np.zeros(B, dtype = np.float32)
nenvsa_eval = eval_obs.shape[0]
eval_episode_reward = np.zeros(nenvs_eval, dtype = np.float32)
for t_rollout in range(nb_eval_steps):
eval_action, eval_q, _, _ = agent.step(eval_obs, apply_noise=False, compute_Q=True)
eval_obs, eval_r, eval_done, eval_info = eval_env.step(max_action * eval_action) # scale for execution in env (as far as DDPG is concerned, every action is in [-1, 1])