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.

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

Fix nn.init.trunc_normal_ call on torch.float16 data #21789

Merged
merged 1 commit into from
Feb 27, 2023
Merged

Fix nn.init.trunc_normal_ call on torch.float16 data #21789

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
12 changes: 6 additions & 6 deletions src/transformers/models/vit/modeling_vit.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -449,17 +449,17 @@ def _init_weights(self, module: Union[nn.Linear, nn.Conv2d, nn.LayerNorm]) -> No
module.bias.data.zero_()
module.weight.data.fill_(1.0)
elif isinstance(module, ViTEmbeddings):
nn.init.trunc_normal_(
module.position_embeddings,
module.position_embeddings.data = nn.init.trunc_normal_(
module.position_embeddings.data.to(torch.float32),
mean=0.0,
std=self.config.initializer_range,
)
).to(module.position_embeddings.dtype)

nn.init.trunc_normal_(
module.cls_token,
module.cls_token.data = nn.init.trunc_normal_(
module.cls_token.data.to(torch.float32),
mean=0.0,
std=self.config.initializer_range,
)
).to(module.cls_token.dtype)
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just a few lines above this change, we have already the same change shown below, so the changes in this PR is the way to go! Thank you.

        if isinstance(module, (nn.Linear, nn.Conv2d)):
            # Upcast the input in `fp32` and cast it back to desired `dtype` to avoid
            # `trunc_normal_cpu` not implemented in `half` issues
            module.weight.data = nn.init.trunc_normal_(
                module.weight.data.to(torch.float32), mean=0.0, std=self.config.initializer_range
            ).to(module.weight.dtype)


def _set_gradient_checkpointing(self, module: ViTEncoder, value: bool = False) -> None:
if isinstance(module, ViTEncoder):
Expand Down
12 changes: 6 additions & 6 deletions src/transformers/models/vit_hybrid/modeling_vit_hybrid.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -474,17 +474,17 @@ def _init_weights(self, module: Union[nn.Linear, nn.Conv2d, nn.LayerNorm]) -> No
module.bias.data.zero_()
module.weight.data.fill_(1.0)
elif isinstance(module, ViTHybridEmbeddings):
nn.init.trunc_normal_(
module.position_embeddings,
module.position_embeddings.data = nn.init.trunc_normal_(
module.position_embeddings.data.to(torch.float32),
mean=0.0,
std=self.config.initializer_range,
)
).to(module.position_embeddings.dtype)

nn.init.trunc_normal_(
module.cls_token,
module.cls_token.data = nn.init.trunc_normal_(
module.cls_token.data.to(torch.float32),
mean=0.0,
std=self.config.initializer_range,
)
).to(module.cls_token.dtype)

def _set_gradient_checkpointing(self, module: ViTHybridEncoder, value: bool = False) -> None:
if isinstance(module, ViTHybridEncoder):
Expand Down