uniform deepspeed overflow check #5424
Open
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
|
Why not using tensor.isnan() and tensor.isinf()? |
tjruwase
reviewed
May 4, 2024
| @@ -181,12 +181,13 @@ def get_norm_with_moe_layers_fast(all_groups_norm, group): | |||
| class CheckOverflow(object): | |||
| '''Checks for overflow in gradient across parallel process''' | |||
|
|
|||
| def __init__(self, param_groups=None, mpu=None, zero_reduce_scatter=False, deepspeed=None): | |||
| def __init__(self, param_groups=None, mpu=None, zero_reduce_scatter=False, deepspeed=None, partition_grads=False): | |||
There was a problem hiding this comment.
Passing deepseed engine into a submodule is not a good design and create all sorts of cyclic reference issues. It is better to pass the specific attributes that are needed, such as enable_backward_allreduce
tjruwase
reviewed
May 4, 2024
| @@ -1473,6 +1473,7 @@ def _configure_bf16_optimizer(self, optimizer): | |||
| timers = self.timers if self.wall_clock_breakdown() else NoopTimer() | |||
| optimizer = BF16_Optimizer(optimizer, | |||
| self.param_names, | |||
| deepspeed=self, | |||
There was a problem hiding this comment.
Don't pass self into submodules.
tjruwase
reviewed
May 4, 2024
| @@ -92,6 +94,10 @@ def __init__(self, | |||
| if self.using_real_optimizer: | |||
| self._setup_for_real_optimizer() | |||
|
|
|||
| # Overflow check init | |||
| self.overflow = False | |||
There was a problem hiding this comment.
Should self.overflow be a class member since it seems to be only used once?
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Outdated suggestions cannot be applied.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.
Before: Overflow check is scattered and duplicated in all places.
This PR:
cc @tjruwase