wassname/curl
1
0
Fork 0
mirror of https://github.com/wassname/curl.git synced 2026-09-09 11:20:39 +08:00
Code Issues Packages Projects Releases Wiki Activity

text change

Browse Source
This commit is contained in:
MishaLaskin
2020-02-08 18:18:51 -08:00
parent ccfa50880f
commit ca78c256d1
1 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files
train.py
+4 -4
View File
@@ -82,7 +82,7 @@ def evaluate(env, agent, video, num_episodes, L, step, args):
all_ep_rewards = []
def run_eval_loop(sample_stochastically=True):
prefix = 'stochastic' if sample_stochastically else ''
prefix = 'stochastic_' if sample_stochastically else ''
for i in range(num_episodes):
obs = env.reset()
video.init(enabled=(i == 0))
@@ -101,13 +101,13 @@ def evaluate(env, agent, video, num_episodes, L, step, args):
episode_reward += reward
video.save('%d.mp4' % step)
L.log('eval/' + prefix + '_episode_reward', episode_reward, step)
L.log('eval/' + prefix + 'episode_reward', episode_reward, step)
all_ep_rewards.append(episode_reward)
mean_ep_reward = np.mean(all_ep_rewards)
best_ep_reward = np.max(all_ep_rewards)
L.log('eval/' + prefix + '_mean_episode_reward', mean_ep_reward, step)
L.log('eval/' + prefix + '_best_episode_reward', best_ep_reward, step)
L.log('eval/' + prefix + 'mean_episode_reward', mean_ep_reward, step)
L.log('eval/' + prefix + 'best_episode_reward', best_ep_reward, step)
run_eval_loop(sample_stochastically=True)
run_eval_loop(sample_stochastically=False)