[Feat] Add CheckpointManager support with TFRA Dynamic Embedding Horovod Training. #57
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Outdated suggestions cannot be applied.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
Now CheckpointManager is available in Deepray when training with TFRA Dynamic Embedding.
[fix] In deepray/core/base_trainer.py, gpu_affinity didn't take effect when NVML Shared Library Not Found.
[fix] In deepray/core/base_trainer.py line 784, self.loss_container.metrics may empty when 'FLAGS.stop_steps = 0' in tools/testing/horovod_sync_train_test.py.
Also the adding script also support test TF Embedding when use Horovod training.
Type of change
Checklist:
How Has This Been Tested?
or