Suggest a value for frames_per_block if the memory is insufficient #582
Open
Suggest a value for frames_per_block if the memory is insufficient #582
Conversation
… frames_per_block
34bf3a8 to
f7141e8
Compare
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR aims to provide a conservative value for the parameter
frames_per_blockif the block cannot be fitted into the current GPU memory.It is more helpful for PtyPy to provide a value for
frames_per_blockas most of the users do not have any idea what a sensible value should be. The ideal way is to provide an option such asframes_per_block='auto'so the user can forget about it but the memory estimation happens after the data is loaded alongside with other stuff and it also depends on the algorithm. As a result it is not possible to achieve this as it is wished here.When the block is too big, it prints out a message similar to this:
This is only relevant for any
_Blockmodel.This also fixes a small bug by removing
self.contextin the CuPy's EPIE engine.