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

Jump to bottom

Improve LayoutLM #9476

Merged
merged 7 commits into from
Jan 12, 2021
Merged

Improve LayoutLM #9476

Changes from 1 commit
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
70ec523
Add LayoutLMForSequenceClassification and integration tests
NielsRogge Jan 8, 2021
2dc8a33
Make style & quality
NielsRogge Jan 8, 2021
2c132e0
Address comments by @sgugger, @patrickvonplaten and @LysandreJik
NielsRogge Jan 8, 2021
fffc19a
Fix rebase with master
NielsRogge Jan 8, 2021
50bab59
Reformat in one line
LysandreJik Jan 11, 2021
9616e45
Improve code examples as requested by @patrickvonplaten
NielsRogge Jan 12, 2021
6844797
Merge branch 'master' into Improving_layoutlm_v2
LysandreJik Jan 12, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
61 changes: 24 additions & 37 deletions src/transformers/models/layoutlm/modeling_layoutlm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -738,21 +738,18 @@ def forward(
>>> words = ["Hello", "world"]
>>> normalized_word_boxes = [637, 773, 693, 782], [698, 773, 733, 782]

>>> tokens = []
>>> token_boxes = []
>>> for word, box in zip(words, normalized_word_boxes):
... word_tokens = tokenizer.tokenize(word)
... tokens.extend(word_tokens)
... token_boxes.extend([box] * len(word_tokens))

>>> # add cls + sep tokens
>>> tokens = [tokenizer.cls_token] + tokens + [tokenizer.sep_token]
>>> # add bounding boxes of cls + sep tokens
>>> token_boxes = [[0, 0, 0, 0]] + token_boxes + [[1000, 1000, 1000, 1000]]

>>> input_ids = torch.tensor(tokenizer.convert_tokens_to_ids(tokens)).unsqueeze(0) # batch size of 1
>>> bbox = torch.tensor(token_boxes).unsqueeze(0) # batch size of 1
>>> attention_mask = torch.tensor([1] * len(tokens)).unsqueeze(0) # batch size of 1
>>> token_type_ids = torch.tensor([0] * len(tokens)).unsqueeze(0) # batch size of 1
>>> encoding = tokenizer(' '.join(words), return_tensors="pt")
NielsRogge marked this conversation as resolved.
Show resolved Hide resolved
>>> input_ids = encoding["input_ids"]
>>> attention_mask = encoding["attention_mask"]
>>> token_type_ids = encoding["token_type_ids"]
>>> bbox = torch.tensor([token_boxes])

>>> outputs = model(input_ids=input_ids, bbox=bbox, attention_mask=attention_mask, token_type_ids=token_type_ids)

Expand Down Expand Up @@ -884,21 +881,18 @@ def forward(
>>> words = ["Hello", "[MASK]"]
>>> normalized_word_boxes = [637, 773, 693, 782], [698, 773, 733, 782]

>>> tokens = []
>>> token_boxes = []
>>> for word, box in zip(words, normalized_word_boxes):
... word_tokens = tokenizer.tokenize(word)
... tokens.extend(word_tokens)
... token_boxes.extend([box] * len(word_tokens))

>>> # add cls + sep tokens
>>> tokens = [tokenizer.cls_token] + tokens + [tokenizer.sep_token]
>>> # add bounding boxes of cls + sep tokens
>>> token_boxes = [[0, 0, 0, 0]] + token_boxes + [[1000, 1000, 1000, 1000]]

>>> input_ids = torch.tensor(tokenizer.convert_tokens_to_ids(tokens)).unsqueeze(0) # batch size of 1
>>> bbox = torch.tensor(token_boxes).unsqueeze(0) # batch size of 1
>>> attention_mask = torch.tensor([1] * len(tokens)).unsqueeze(0) # batch size of 1
>>> token_type_ids = torch.tensor([0] * len(tokens)).unsqueeze(0) # batch size of 1
>>> encoding = tokenizer(' '.join(words), return_tensors="pt")
>>> input_ids = encoding["input_ids"]
>>> attention_mask = encoding["attention_mask"]
>>> token_type_ids = encoding["token_type_ids"]
>>> bbox = torch.tensor([token_boxes])

>>> labels = tokenizer("Hello world", return_tensors="pt")["input_ids"]

Expand Down Expand Up @@ -1002,22 +996,18 @@ def forward(
>>> words = ["Hello", "world"]
>>> normalized_word_boxes = [637, 773, 693, 782], [698, 773, 733, 782]

>>> tokens = []
>>> token_boxes = []
>>> for word, box in zip(words, normalized_word_boxes):
... word_tokens = tokenizer.tokenize(word)
... tokens.extend(word_tokens)
... token_boxes.extend([box] * len(word_tokens))

>>> # add cls + sep tokens
>>> tokens = [tokenizer.cls_token] + tokens + [tokenizer.sep_token]
>>> # add bounding boxes of cls + sep tokens
>>> token_boxes = [[0, 0, 0, 0]] + token_boxes + [[1000, 1000, 1000, 1000]]

>>> input_ids = torch.tensor(tokenizer.convert_tokens_to_ids(tokens)).unsqueeze(0) # batch size of 1
>>> bbox = torch.tensor(token_boxes).unsqueeze(0) # batch size of 1
>>> attention_mask = torch.tensor([1] * len(tokens)).unsqueeze(0) # batch size of 1
>>> token_type_ids = torch.tensor([0] * len(tokens)).unsqueeze(0) # batch size of 1

>>> encoding = tokenizer(' '.join(words), return_tensors="pt")
>>> input_ids = encoding["input_ids"]
>>> attention_mask = encoding["attention_mask"]
>>> token_type_ids = encoding["token_type_ids"]
>>> bbox = torch.tensor([token_boxes])
>>> sequence_label = torch.tensor([1])

>>> outputs = model(input_ids=input_ids, bbox=bbox, attention_mask=attention_mask, token_type_ids=token_type_ids,
Expand Down Expand Up @@ -1123,21 +1113,18 @@ def forward(
>>> words = ["Hello", "world"]
>>> normalized_word_boxes = [637, 773, 693, 782], [698, 773, 733, 782]

>>> tokens = []
>>> token_boxes = []
>>> for word, box in zip(words, normalized_word_boxes):
NielsRogge marked this conversation as resolved.
Show resolved Hide resolved
... word_tokens = tokenizer.tokenize(word)
... tokens.extend(word_tokens)
... token_boxes.extend([box] * len(word_tokens))

>>> # add cls + sep tokens
>>> tokens = [tokenizer.cls_token] + tokens + [tokenizer.sep_token]
>>> # add bounding boxes of cls + sep tokens
>>> token_boxes = [[0, 0, 0, 0]] + token_boxes + [[1000, 1000, 1000, 1000]]

>>> input_ids = torch.tensor(tokenizer.convert_tokens_to_ids(tokens)).unsqueeze(0) # batch size of 1
>>> bbox = torch.tensor(token_boxes).unsqueeze(0) # batch size of 1
>>> attention_mask = torch.tensor([1] * len(tokens)).unsqueeze(0) # batch size of 1
>>> token_type_ids = torch.tensor([0] * len(tokens)).unsqueeze(0) # batch size of 1
>>> encoding = tokenizer(' '.join(words), return_tensors="pt")
>>> input_ids = encoding["input_ids"]
>>> attention_mask = encoding["attention_mask"]
>>> token_type_ids = encoding["token_type_ids"]
>>> bbox = torch.tensor([token_boxes])
>>> token_labels = torch.tensor([1,1,0,0]).unsqueeze(0) # batch size of 1

>>> outputs = model(input_ids=input_ids, bbox=bbox, attention_mask=attention_mask, token_type_ids=token_type_ids,
Expand Down