feat(ml): configurable batch size for facial recognition #13689
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
github-actions
bot
added
documentation
jrasm91
approved these changes
Oct 22, 2024
| | `MACHINE_LEARNING_ANN_FP16_TURBO` | Execute operations in FP16 precision: increasing speed, reducing precision (applies only to ARM-NN) | `False` | machine learning | | ||
| | `MACHINE_LEARNING_ANN_TUNING_LEVEL` | ARM-NN GPU tuning level (1: rapid, 2: normal, 3: exhaustive) | `2` | machine learning | | ||
| | `MACHINE_LEARNING_DEVICE_IDS`<sup>\*4</sup> | Device IDs to use in multi-GPU environments | `0` | machine learning | | ||
| | `MACHINE_LEARNING_MAX_BATCH_SIZE__FACIAL_RECOGNITION` | Set the maximum number of faces that will be processed at once by the facial recognition model | None (`1` if using OpenVINO) | machine learning | |
There was a problem hiding this comment.
Why is this a double underscore?
There was a problem hiding this comment.
It indicates a nesting level. Since there'll likely be other things batched in the future, these settings are grouped under max_batch_size.
alextran1502
approved these changes
Oct 23, 2024
yosit
pushed a commit
to yosit/immich
that referenced
this pull request
Oct 24, 2024
…13689) * configurable batch size, default openvino to 1 * update docs * don't add a new dependency for two lines * fix typing
|
Does this commit can solve my problem? |
I tested，It still not working.... |
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Outdated suggestions cannot be applied.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.
Description
The facial recognition model currently receives all faces in an image in one pass. For OpenVINO users, this can lead to surprisingly high RAM usage because of its allocation patterns. This PR sets facial recognition to process faces one by one for OpenVINO without changing the behavior elsewhere. There is a new env to configure the batch size, both for users on e.g. CUDA who also want to set a limit, and for OpenVINO users with beefy systems that want to benefit from batching.
Fixes #13674
How Has This Been Tested?
Tested both with
MACHINE_LEARNING_MAX_BATCH_SIZE__FACIAL_RECOGNITION=2and without setting the env and confirming that face detection works in both cases.