Prevent tf.Session() from eating up all the GPU memory #3219
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters. Learn more about bidirectional Unicode characters
chriselion reviewed Jan 14, 2020
| try: | ||
| with tf.Session() as sess: | ||
| # Prevent GPU memory from being eaten up by this small writing session | ||
| config = tf.ConfigProto() |
There was a problem hiding this comment.
Should we move the ConfigProto setup to a utility function? I'm ok with not doing it now, but if we need it in a 3rd place, I think we should.
There was a problem hiding this comment.
We can put it in tf_utils, that seems like the natural place for it
There was a problem hiding this comment.
I moved it to tf in tf_utils - lmk how it looks.
chriselion approved these changes Jan 14, 2020
I think so - this and the Tensorboard fix should probably be a hotfix for 0.13. What do you think? |
andrewcoh approved these changes Jan 14, 2020
This was referenced Jan 24, 2020
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
3 participants
Add this suggestion to a batch that can be applied as a single commit. This suggestion is invalid because no changes were made to the code. Suggestions cannot be applied while the pull request is closed. Suggestions cannot be applied while viewing a subset of changes. Only one suggestion per line can be applied in a batch. Add this suggestion to a batch that can be applied as a single commit. Applying suggestions on deleted lines is not supported. You must change the existing code in this line in order to create a valid suggestion. Outdated suggestions cannot be applied. This suggestion has been applied or marked resolved. Suggestions cannot be applied from pending reviews. Suggestions cannot be applied on multi-line comments. Suggestions cannot be applied while the pull request is queued to merge. Suggestion cannot be applied right now. Please check back later.
By default, a tf.Session will take up all GPU memory. In the Policy, we combat this by enabling allow_growth = True and enabling soft placement (for multi-GPU).
But we introduced another session to write hyperparameters. In the past this session was created after the policy, and respected those configs. However, now it is created before and needs its own config options.