wassname/DeepRL
1
0
Fork 0
mirror of https://github.com/wassname/DeepRL.git synced 2026-09-10 11:40:58 +08:00
Code Issues Packages Projects Releases Wiki Activity

Major update

Browse Source
This commit is contained in:
Shangtong Zhang
2017-11-11 20:52:57 -07:00
parent 8c197e183c
commit c1fdb1bd2f
17 changed files with 48 additions and 199 deletions
Download Patch File Download Diff File Expand all files Collapse all files
agent/async_agent.py
+1 -2
View File
@@ -51,7 +51,7 @@ def evaluate(config, task, learning_network, extra):
with open('data/%s-%s-statistics-%s.bin' % (
config.tag, config.worker.__name__, task.name), 'wb') as f:
pickle.dump([test_rewards, test_points, test_wall_times], f)
if np.mean(rewards) > task.success_threshold or (config.max_steps and steps >= config.max_steps):
if np.mean(rewards) >= config.success_threshold or (config.max_steps and steps >= config.max_steps):
config.stop_signal.value = True
break
@@ -87,7 +87,6 @@ class AsyncAgent:
extra = None
args = [(i, config, learning_network, extra) for i in range(config.num_workers)]
args.append((config, task, learning_network, extra))
# procs = []
procs = [mp.Process(target=evaluate, args=args[-1])]
procs.extend([mp.Process(target=train, args=args[i]) for i in range(config.num_workers)])
for p in procs: p.start()