Add vram flushing support #256
Merged
Conversation
This file contains 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
Implement automatic VRAM clearing after a specified period of idleness. * Add a mechanism to track the last activity time and implement a background thread to monitor idleness and clear VRAM after five minutes of inactivity in `app/faster_whisper/core.py` and `app/openai_whisper/core.py`. * Update the `transcribe` and `language_detection` functions in both core files to reset the last activity time upon invocation. * Add a function to fully release the model from memory using `del`, `torch.cuda.empty_cache()`, and `gc.collect()` in both core files. * Add configuration options for the idleness timeout period and enabled/disabled state in the environment variables in `app/webservice.py`.
AngelOnFira
reviewed
Dec 9, 2024
Co-authored-by: Forest Anderson <forestkzanderson@gmail.com>
ahmetoner
approved these changes
Dec 16, 2024
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Outdated suggestions cannot be applied.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 change implements a monitoring thread to unload the whisper model after a user set timeout period. Timeout defaults to being disabled, and can be set with the IDLE_TIMEOUT environment variable.
Once unloaded some residual VRAM allocation appears to remain (~0.25GB), but memory usage remains consistent across reloads, leading me to believe that this is a limitation of Docker.
This closes #216 and closes #196