Skip to content

Torch tensors as inference input #2014

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

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Torch tensors as inference input #2014

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
39 changes: 27 additions & 12 deletions doctr/models/preprocessor/pytorch.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -60,12 +60,21 @@

return batches

def sample_transforms(self, x: np.ndarray) -> torch.Tensor:
if x.ndim != 3:
raise AssertionError("expected list of 3D Tensors")
if x.dtype not in (np.uint8, np.float32, np.float16):
raise TypeError("unsupported data type for numpy.ndarray")
tensor = torch.from_numpy(x.copy()).permute(2, 0, 1)
def sample_transforms(self, x: np.ndarray | torch.Tensor) -> torch.Tensor:
if isinstance(x, np.ndarray):
if x.ndim != 3:
raise AssertionError("expected list of 3D Tensors")
if x.dtype not in (np.uint8, np.float32, np.float16):
raise TypeError("unsupported data type for numpy.ndarray")
x = torch.from_numpy(x.copy())
elif isinstance(x, torch.Tensor):
if x.ndim != 3:
raise AssertionError("expected 3D Tensor")
else:
raise TypeError(f"invalid input type: {type(x)}")

Check notice on line 75 in doctr/models/preprocessor/pytorch.py

View check run for this annotation

Codacy Production / Codacy Static Code Analysis

doctr/models/preprocessor/pytorch.py#L75 

Trailing whitespace
tensor = x.permute(2, 0, 1)

# Resizing
tensor = self.resize(tensor)
# Data type
Expand All @@ -76,22 +85,28 @@

return tensor

def __call__(self, x: np.ndarray | list[np.ndarray]) -> list[torch.Tensor]:
def __call__(self, x: np.ndarray | torch.Tensor | list[np.ndarray] | list[torch.Tensor]) -> list[torch.Tensor]:
"""Prepare document data for model forwarding

Args:
x: list of images (np.array) or a single image (np.array) of shape (H, W, C)
x: list of images (np.array | torch.Tensor) or a single image (np.array | torch.Tensor) of shape (H, W, C)

Returns:
list of page batches (*, C, H, W) ready for model inference
"""
# Input type check
if isinstance(x, np.ndarray):
if isinstance(x, np.ndarray) or isinstance(x, torch.Tensor):

if x.ndim != 4:
raise AssertionError("expected 4D Tensor")
if x.dtype not in (np.uint8, np.float32, np.float16):
if x.dtype not in (np.uint8, np.float32, np.float16) and isinstance(x, np.ndarray):
raise TypeError("unsupported data type for numpy.ndarray")
tensor = torch.from_numpy(x.copy()).permute(0, 3, 1, 2)
if x.dtype not in (torch.uint8, torch.float32, torch.float16) and isinstance(x, torch.Tensor):
raise TypeError("unsupported data type for torch.Tensor")
if isinstance(x, np.ndarray):
x = torch.from_numpy(x.copy())

tensor = x.permute(0, 3, 1, 2)

# Resizing
if tensor.shape[-2] != self.resize.size[0] or tensor.shape[-1] != self.resize.size[1]:
Expand All @@ -105,7 +120,7 @@
tensor = tensor.to(dtype=torch.float32)
batches = [tensor]

elif isinstance(x, list) and all(isinstance(sample, np.ndarray) for sample in x):
elif isinstance(x, list) and all(isinstance(sample, np.ndarray) or isinstance(sample, torch.Tensor) for sample in x):
# Sample transform (to tensor, resize)
samples = list(multithread_exec(self.sample_transforms, x))
# Batching
Expand Down