Skip to content

fix FSDP2 test case failure on XPU #3771

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

Merged
S1ro1 merged 2 commits into from
Sep 12, 2025
Merged

fix FSDP2 test case failure on XPU #3771

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
7f0cd88
fix FSDP2 test case failure on XPU
yao-matrix Sep 10, 2025
2445836
fix style
yao-matrix Sep 10, 2025
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
19 changes: 6 additions & 13 deletions src/accelerate/accelerator.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -568,25 +568,18 @@ def __init__(
and self.distributed_type not in (DistributedType.DEEPSPEED, DistributedType.MEGATRON_LM)
):
self.native_amp = True
if self.device.type not in (
"xpu",
"cuda",
"npu",
"xla",
"mlu",
"musa",
"hpu",
"sdaa",
"mps",
) or is_torch_xla_available(check_is_tpu=True):
raise ValueError(f"fp16 mixed precision requires a GPU or MPS device (not {self.device.type!r}).")
supported_device = ("xpu", "cuda", "npu", "xla", "mlu", "musa", "hpu", "sdaa", "mps")
if self.device.type not in supported_device or is_torch_xla_available(check_is_tpu=True):
raise ValueError(
f"fp16 mixed precision requires a device in {supported_device} (not {self.device.type!r})."
)
Copy link
Contributor Author

@yao-matrix yao-matrix Sep 10, 2025

Choose a reason for hiding this comment

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

prior error message is misleading.

if self.device.type == "mps" and not is_torch_version(">=", "2.5.0"):
raise ValueError("fp16 mixed precision with MPS device requires a Pytorch >= 2.5.0")
kwargs = self.scaler_handler.to_kwargs() if self.scaler_handler is not None else {}

# FSDP2 doesn't use ShardedGradScaler, don't want to modify `get_grad_scaler`, rather create a simple utility
if self.is_fsdp2:
self.scaler = get_fsdp2_grad_scaler(**kwargs)
self.scaler = get_fsdp2_grad_scaler(device=self.device.type, **kwargs)
else:
self.scaler = get_grad_scaler(self.distributed_type, **kwargs)

Expand Down