Spaces:
Runtime error
Runtime error
JustinLin610
commited on
Commit
•
6cf5e8c
1
Parent(s):
3006ddf
debug
Browse files
app.py
CHANGED
@@ -136,8 +136,8 @@ def patch_resize_transform(patch_image_size=480, is_document=False):
|
|
136 |
|
137 |
|
138 |
# Construct input for caption task
|
139 |
-
def construct_sample(image: Image, patch_image_size=480):
|
140 |
-
patch_image = patch_resize_transform(patch_image_size)(image).unsqueeze(0)
|
141 |
patch_mask = torch.tensor([True])
|
142 |
src_text = encode_text("图片上的文字是什么?", append_bos=True, append_eos=True).unsqueeze(0)
|
143 |
src_length = torch.LongTensor([s.ne(pad_idx).long().sum() for s in src_text])
|
@@ -177,7 +177,7 @@ def ocr(img, task_type):
|
|
177 |
ocr_result = []
|
178 |
for i, (box, image) in enumerate(zip(box_list, image_list)):
|
179 |
image = Image.fromarray(image)
|
180 |
-
sample = construct_sample(
|
181 |
sample = utils.move_to_cuda(sample) if use_cuda else sample
|
182 |
sample = utils.apply_to_sample(apply_half, sample) if use_fp16 else sample
|
183 |
|
|
|
136 |
|
137 |
|
138 |
# Construct input for caption task
|
139 |
+
def construct_sample(image: Image, patch_image_size=480, is_document=False):
|
140 |
+
patch_image = patch_resize_transform(patch_image_size, is_document=is_document)(image).unsqueeze(0)
|
141 |
patch_mask = torch.tensor([True])
|
142 |
src_text = encode_text("图片上的文字是什么?", append_bos=True, append_eos=True).unsqueeze(0)
|
143 |
src_length = torch.LongTensor([s.ne(pad_idx).long().sum() for s in src_text])
|
|
|
177 |
ocr_result = []
|
178 |
for i, (box, image) in enumerate(zip(box_list, image_list)):
|
179 |
image = Image.fromarray(image)
|
180 |
+
sample = construct_sample(image, cfg.task.patch_image_size, is_document=(task_type=='Document'))
|
181 |
sample = utils.move_to_cuda(sample) if use_cuda else sample
|
182 |
sample = utils.apply_to_sample(apply_half, sample) if use_fp16 else sample
|
183 |
|