mirror of
https://github.com/saymrwulf/pytorch.git
synced 2026-05-14 20:57:59 +00:00
cf3f4285a8
Ensures that nested tensors subclasses are guarded properly. It turns out this case is already handled [here](d77af49380/torch/_dynamo/variables/builder.py (L1496)) which will recursively wrap inner tensors adding metadata guards for them.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/131002
Approved by: https://github.com/bdhirsh
86 lines
3 KiB
Python
86 lines
3 KiB
Python
# mypy: ignore-errors
|
|
|
|
import torch
|
|
import torch.utils._pytree as pytree
|
|
from torch.utils._python_dispatch import return_and_correct_aliasing
|
|
|
|
|
|
# A simple tensor subclass that holds two tensors internally, and runs every op on both tensors.
|
|
class TwoTensor(torch.Tensor):
|
|
@staticmethod
|
|
def __new__(cls, a, b):
|
|
assert (
|
|
a.device == b.device
|
|
and a.layout == b.layout
|
|
and a.requires_grad == b.requires_grad
|
|
and a.dtype == b.dtype
|
|
)
|
|
# I guess it would be more accurate to represent the shape as torch.cat(a, b).shape
|
|
shape = a.shape
|
|
kwargs = {}
|
|
kwargs["strides"] = a.stride()
|
|
kwargs["storage_offset"] = a.storage_offset()
|
|
kwargs["device"] = a.device
|
|
kwargs["layout"] = a.layout
|
|
kwargs["requires_grad"] = a.requires_grad
|
|
kwargs["dtype"] = a.dtype
|
|
out = torch.Tensor._make_wrapper_subclass(cls, shape, **kwargs)
|
|
|
|
assert a.shape == b.shape
|
|
assert a.stride() == b.stride()
|
|
assert a.storage_offset() == b.storage_offset()
|
|
return out
|
|
|
|
def __init__(self, a, b):
|
|
self.a = a
|
|
self.b = b
|
|
|
|
def __repr__(self):
|
|
a_repr = repr(self.a)
|
|
b_repr = repr(self.b)
|
|
return f"TwoTensor({a_repr}, {b_repr})"
|
|
|
|
def __tensor_flatten__(self):
|
|
return ["a", "b"], None
|
|
|
|
@staticmethod
|
|
def __tensor_unflatten__(inner_tensors, meta, outer_size, outer_stride):
|
|
assert meta is None
|
|
a, b = inner_tensors["a"], inner_tensors["b"]
|
|
return TwoTensor(a, b)
|
|
|
|
@classmethod
|
|
def __torch_dispatch__(cls, func, types, args, kwargs):
|
|
if kwargs is None:
|
|
kwargs = {}
|
|
args_a = pytree.tree_map_only(TwoTensor, lambda x: x.a, args)
|
|
args_b = pytree.tree_map_only(TwoTensor, lambda x: x.b, args)
|
|
|
|
kwargs_a = pytree.tree_map_only(TwoTensor, lambda x: x.a, kwargs)
|
|
kwargs_b = pytree.tree_map_only(TwoTensor, lambda x: x.b, kwargs)
|
|
|
|
out_a = func(*args_a, **kwargs_a)
|
|
out_b = func(*args_b, **kwargs_b)
|
|
out_a_flat, spec = pytree.tree_flatten(out_a)
|
|
out_b_flat = pytree.tree_leaves(out_b)
|
|
# for aten ops that return non-tensors, just assume that
|
|
# our two inner tensors return the same value
|
|
out_flat = [
|
|
TwoTensor(o_a, o_b) if isinstance(o_a, torch.Tensor) else o_a
|
|
for o_a, o_b in zip(out_a_flat, out_b_flat)
|
|
]
|
|
out = pytree.tree_unflatten(out_flat, spec)
|
|
from torch._higher_order_ops.cond import cond_op
|
|
|
|
if func is cond_op:
|
|
return out
|
|
else:
|
|
return return_and_correct_aliasing(func, args, kwargs, out)
|
|
|
|
|
|
class TwoTensorMode(torch.utils._python_dispatch.TorchDispatchMode):
|
|
def __torch_dispatch__(self, func, types, args=(), kwargs=None):
|
|
out = func(*args, **kwargs)
|
|
if torch._subclasses.fake_tensor._is_tensor_constructor(func):
|
|
out = TwoTensor(out, out.clone())
|
|
return out
|