Skip to content

(TBV) fix: memory limit #203

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
YuriCat wants to merge 1 commit into from
Closed

(TBV) fix: memory limit #203

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 7 additions & 10 deletions handyrl/train.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -489,17 +489,14 @@ def feed_episodes(self, episodes):
self.generation_results[model_id] = n + 1, r + outcome, r2 + outcome ** 2

# store generated episodes
mem = psutil.virtual_memory()
mem_used_ratio = mem.used / mem.total
mem_ok = mem_used_ratio <= 0.95
maximum_episodes = self.args['maximum_episodes'] if mem_ok else len(self.trainer.episodes)

if not mem_ok and 'memory_over' not in self.flags:
warnings.warn("memory usage %.1f%% with buffer size %d" % (mem_used_ratio * 100, len(self.trainer.episodes)))
self.flags.add('memory_over')

self.trainer.episodes.extend([e for e in episodes if e is not None])
while len(self.trainer.episodes) > maximum_episodes:
while len(self.trainer.episodes) > self.args['maximum_episodes']:
self.trainer.episodes.popleft()

while psutil.virtual_memory().percent >= 95:
if 'memory_over' not in self.flags:
warnings.warn("memory usage %.1f%% with buffer size %d" % (mem_used_ratio * 100, len(self.trainer.episodes)))
Copy link
Member

@ikki407 ikki407 Jun 1, 2021 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Is mem_used_ratio defined anywhere?

Does there exist a way not to use self.flags? ie. "once" (I'm thinking...)

Copy link
Member

@ikki407 ikki407 Jun 1, 2021 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

related?
https://stackoverflow.com/a/22661838/10809665

Repetitions of a particular warning for the same source location are typically suppressed.

https://docs.python.org/3/library/warnings.html

Copy link
Contributor Author

@YuriCat YuriCat Jun 1, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oops

self.flags.add('memory_over')
self.trainer.episodes.popleft()

def feed_results(self, results):
Expand Down