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

Update tokenization_code_llama_fast.py #26576

Merged
merged 3 commits into from
Oct 6, 2023
Merged

Update tokenization_code_llama_fast.py #26576

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
963200b
Update tokenization_code_llama_fast.py
andyl98 Oct 3, 2023
1596d9a
Update test_tokenization_code_llama.py
andyl98 Oct 4, 2023
3048f3e
Update test_tokenization_code_llama.py
andyl98 Oct 4, 2023
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
2 changes: 1 addition & 1 deletion src/transformers/models/code_llama/tokenization_code_llama_fast.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -278,7 +278,7 @@ def set_infilling_processor(self, reset, suffix_first=False, add_special_tokens=
special_tokens = [(self.bos_token, self.bos_token_id)] if self.add_bos_token and add_special_tokens else []
if suffix_first:
# format as " <PRE> <SUF>{suf} <MID> {pre}"
pair += [self.prefix_token, self.suffix_token, "$A", self.middle_token, "$B"]
pair += [self.prefix_token, self.suffix_token, "$B", self.middle_token, "$A"]
special_tokens += [
(self.prefix_token, self.prefix_id),
(self.suffix_token, self.suffix_id),
Expand Down
12 changes: 12 additions & 0 deletions tests/models/code_llama/test_tokenization_code_llama.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -643,3 +643,15 @@ def main():
input_ids = tokenizer.encode(PROMPTS[0])
self.assertEqual(input_ids, tokenizer.encode(prefix, suffix=suffix))
self.assertEqual(tokenizer.encode(prefix, suffix=suffix), tokenizer_fast.encode(prefix, suffix=suffix))

# Adding suffix_first check for infilling tasks
suffix_first_formatted_prompt = tokenizer.tokenize(PROMPTS[0], suffix_first=True)
self.assertEqual(suffix_first_formatted_prompt, tokenizer_fast.tokenize(PROMPTS[0], suffix_first=True))
prefix, suffix = PROMPTS[0].split("<FILL_ME>")
self.assertEqual(suffix_first_formatted_prompt, tokenizer.tokenize(prefix, suffix, suffix_first=True))
self.assertEqual(suffix_first_formatted_prompt, tokenizer_fast.tokenize(prefix, suffix, suffix_first=True))

prefix, suffix = PROMPTS[0].split("<FILL_ME>")
suffix_first_input_ids = tokenizer.encode(PROMPTS[0], suffix_first=True)
self.assertEqual(suffix_first_input_ids, tokenizer.encode(prefix, suffix=suffix, suffix_first=True))
self.assertEqual(suffix_first_input_ids, tokenizer_fast.encode(prefix, suffix=suffix, suffix_first=True))
Loading