Skip to content

fix(pipeline): scale DSv4 hybrid indexer loss #5942

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
FDecaYed wants to merge 1 commit into from
Closed

fix(pipeline): scale DSv4 hybrid indexer loss #5942

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion megatron/core/pipeline_parallel/schedules.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -277,7 +277,7 @@ def _get_experimental_attention_variant_loss_scale_func(config):
if loss_scale_func is not None:
return loss_scale_func

if getattr(config, 'experimental_attention_variant', None) == 'dsa':
if getattr(config, 'experimental_attention_variant', None) in ('dsa', 'dsv4_hybrid'):
from megatron.core.transformer.experimental_attention_variant.dsa import (
DSAIndexerLossAutoScaler,
)
Expand Down
5 changes: 3 additions & 2 deletions tests/unit_tests/pipeline_parallel/test_schedules.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -377,15 +377,16 @@ def test_dsa_indexer_loss_scale_accepts_dict_output_tensor():
)


def test_dsa_indexer_loss_scale_defaults_from_variant_without_mutating_config():
@pytest.mark.parametrize("variant", ["dsa", "dsv4_hybrid"])
def test_indexer_loss_scale_defaults_from_variant_without_mutating_config(variant):
from megatron.core.transformer.experimental_attention_variant.dsa import (
DSAIndexerLossAutoScaler,
)

config = SimpleNamespace(
calculate_per_token_loss=True,
experimental_attention_variant_loss_scale_func=None,
experimental_attention_variant='dsa',
experimental_attention_variant=variant,
grad_scale_func=lambda tensor: tensor * 7.0,
num_moe_experts=None,
mtp_num_layers=None,
Expand Down
Loading