Skip to content
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.

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

Make FLAGS_force_align_vpp_grad_sum_order default to false #54937

Merged
merged 2 commits into from
Jun 28, 2023
Merged
Changes from all commits
Commits
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
Original file line number Diff line number Diff line change
Expand Up @@ -49,7 +49,7 @@ def __init__(self, clip, hcg):
self.not_sharding_stage1 = True
self._vpp_chunk_num = None
self._force_align_vpp_grad_sum_order = distutils.util.strtobool(
os.getenv('FLAGS_force_align_vpp_grad_sum_order', '1')
os.getenv('FLAGS_force_align_vpp_grad_sum_order', '0')
)

def _get_vpp_chunk_num(self, params_grads):
Expand Down Expand Up @@ -168,9 +168,10 @@ def _add_sum_squares(self, sum_squares):

@no_grad()
def _dygraph_clip(self, params_grads):
chunk_num = self._get_vpp_chunk_num(params_grads)
if chunk_num > 0 and self._force_align_vpp_grad_sum_order:
return self._vpp_dygraph_clip(params_grads, chunk_num)
if self._force_align_vpp_grad_sum_order:
chunk_num = self._get_vpp_chunk_num(params_grads)
if chunk_num > 0:
return self._vpp_dygraph_clip(params_grads, chunk_num)

sum_square_dist_fp16 = []
sum_square_dist_bf16 = []
Expand Down