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 Qwen2VL processor to handle odd number of frames #35431

Open
wants to merge 3 commits into
base: main
Choose a base branch
from
+13 −2
Open

Fix Qwen2VL processor to handle odd number of frames #35431

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
e6784c9
fix: processing odd number of frames
jla524 Dec 27, 2024
105fa10
feat: add test case
jla524 Dec 27, 2024
a0afcb8
update: test one frame
jla524 Dec 28, 2024
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
4 changes: 2 additions & 2 deletions src/transformers/models/qwen2_vl/image_processing_qwen2_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -291,8 +291,8 @@ def _preprocess(
patches = np.array(processed_images)
if data_format == ChannelDimension.LAST:
patches = patches.transpose(0, 3, 1, 2)
if patches.shape[0] == 1:
patches = np.tile(patches, (self.temporal_patch_size, 1, 1, 1))
if patches.shape[0] % 2 == 1:
patches = np.concatenate([patches, patches[-1][np.newaxis]], axis=0)
channel = patches.shape[1]
grid_t = patches.shape[0] // self.temporal_patch_size
grid_h, grid_w = resized_height // self.patch_size, resized_width // self.patch_size
Expand Down
11 changes: 11 additions & 0 deletions tests/models/qwen2_vl/test_image_processing_qwen2_vl.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -247,3 +247,14 @@ def test_nested_input(self):
# Image processor should return same pixel values, independently of ipnut format
self.assertTrue((encoded_images_nested == encoded_images).all())
self.assertTrue((image_grid_thws_nested == expected_image_grid_thws).all())

def test_odd_number_of_frames(self):
image_processing = self.image_processing_class(**self.image_processor_dict)
expected_dims_by_frames = {1: 324, 3: 648, 5: 972, 7: 1296, 9: 1620}

for num_frames, expected_dims in expected_dims_by_frames.items():
image_inputs = np.random.randint(0, 255, size=(num_frames, 256, 256, 3))
prcocess_out = image_processing(image_inputs, return_tensors="pt")
encoded_images = prcocess_out.pixel_values
expected_output_image_shape = (expected_dims, 1176)
self.assertEqual(tuple(encoded_images.shape), expected_output_image_shape)