linting and debugs
This commit is contained in:
parent
42e6df723c
commit
c2c05fcca8
|
@ -140,20 +140,20 @@ def manual_cast_forward(target_dtype):
|
||||||
):
|
):
|
||||||
args = [arg.to(target_dtype) if isinstance(arg, torch.Tensor) else arg for arg in args]
|
args = [arg.to(target_dtype) if isinstance(arg, torch.Tensor) else arg for arg in args]
|
||||||
kwargs = {k: v.to(target_dtype) if isinstance(v, torch.Tensor) else v for k, v in kwargs.items()}
|
kwargs = {k: v.to(target_dtype) if isinstance(v, torch.Tensor) else v for k, v in kwargs.items()}
|
||||||
|
|
||||||
org_dtype = torch_utils.get_param(self).dtype
|
org_dtype = torch_utils.get_param(self).dtype
|
||||||
if org_dtype != target_dtype:
|
if org_dtype != target_dtype:
|
||||||
self.to(target_dtype)
|
self.to(target_dtype)
|
||||||
result = self.org_forward(*args, **kwargs)
|
result = self.org_forward(*args, **kwargs)
|
||||||
if org_dtype != target_dtype:
|
if org_dtype != target_dtype:
|
||||||
self.to(org_dtype)
|
self.to(org_dtype)
|
||||||
|
|
||||||
if target_dtype != dtype_inference:
|
if target_dtype != dtype_inference:
|
||||||
if isinstance(result, tuple):
|
if isinstance(result, tuple):
|
||||||
result = tuple(
|
result = tuple(
|
||||||
i.to(dtype_inference)
|
i.to(dtype_inference)
|
||||||
if isinstance(i, torch.Tensor)
|
if isinstance(i, torch.Tensor)
|
||||||
else i
|
else i
|
||||||
for i in result
|
for i in result
|
||||||
)
|
)
|
||||||
elif isinstance(result, torch.Tensor):
|
elif isinstance(result, torch.Tensor):
|
||||||
|
@ -185,7 +185,7 @@ def autocast(disable=False):
|
||||||
if fp8 and device==cpu:
|
if fp8 and device==cpu:
|
||||||
return torch.autocast("cpu", dtype=torch.bfloat16, enabled=True)
|
return torch.autocast("cpu", dtype=torch.bfloat16, enabled=True)
|
||||||
|
|
||||||
if dtype == torch.float32 and shared.cmd_opts.precision == "full":
|
if dtype == torch.float32:
|
||||||
return contextlib.nullcontext()
|
return contextlib.nullcontext()
|
||||||
|
|
||||||
if has_xpu() or has_mps() or cuda_no_autocast():
|
if has_xpu() or has_mps() or cuda_no_autocast():
|
||||||
|
|
Loading…
Reference in New Issue