Skip to content

Update metadata_utils.py #188

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

Draft
jordiclive wants to merge 2 commits into
base: master
Choose a base branch
from
Draft

Update metadata_utils.py #188

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
1db784e
Update metadata_utils.py
jordiclive Apr 28, 2023
fb6c139
Update metadata_utils.py
jordiclive Apr 28, 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
32 changes: 21 additions & 11 deletions bsmetadata/metadata_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -48,7 +48,7 @@ class BasicMetadata:


def add_metadata_and_chunk_examples(
examples: Dict[str, List], tokenizer: PreTrainedTokenizerFast, cfg: MetadataConfig
examples: Dict[str, List], tokenizer: PreTrainedTokenizerFast, cfg: MetadataConfig, without_metadata: bool = False
) -> Dict[str, List]:
"""Adds metadata to the provided input examples, encodes them and groups them in chunks of size `cfg.max_seq_len`.

Expand Down Expand Up @@ -124,16 +124,26 @@ def is_metadata(idx: int) -> bool:
for text_chunk_encoded, chunk_metadata_mask in chunks(
max_text_len, text_with_local_metadata_encoded.input_ids, token_level_metadata_mask
):
total_len = prefix_len + len(text_chunk_encoded)
padding_len = max_text_len - len(text_chunk_encoded)

input_ids = metadata_prefix_encoded + text_chunk_encoded + [tokenizer.eos_token_id] * padding_len
attention_mask = [1] * total_len + [0] * padding_len
metadata_mask = [1] * prefix_len + [int(x) for x in chunk_metadata_mask] + [0] * padding_len

linearized_examples["input_ids"].append(input_ids)
linearized_examples["attention_mask"].append(attention_mask)
linearized_examples["metadata_mask"].append(metadata_mask)
if without_metadata:
total_len = len(text_chunk_encoded)
padding_len = cfg.max_seq_len - len(text_chunk_encoded)
attention_mask = [1] * total_len + [0] * padding_len
input_ids = text_chunk_encoded + [tokenizer.eos_token_id] * padding_len
metadata_mask = [0] * total_len
linearized_examples["input_ids"].append(input_ids)
linearized_examples["attention_mask"].append(attention_mask)
linearized_examples["metadata_mask"].append(metadata_mask)
else:
total_len = prefix_len + len(text_chunk_encoded)
padding_len = max_text_len - len(text_chunk_encoded)

input_ids = metadata_prefix_encoded + text_chunk_encoded + [tokenizer.eos_token_id] * padding_len
attention_mask = [1] * total_len + [0] * padding_len
metadata_mask = [1] * prefix_len + [int(x) for x in chunk_metadata_mask] + [0] * padding_len

linearized_examples["input_ids"].append(input_ids)
linearized_examples["attention_mask"].append(attention_mask)
linearized_examples["metadata_mask"].append(metadata_mask)

return linearized_examples

Expand Down