Skip to content
This repository has been archived by the owner on Mar 1, 2024. It is now read-only.

Handle AssertionError when mention exceeds max_seq_length #98

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Handle AssertionError when mention exceeds max_seq_length #98

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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: 4 additions & 0 deletions blink/biencoder/data_process.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,6 +36,10 @@ def get_context_representation(
if sample[mention_key] and len(sample[mention_key]) > 0:
mention_tokens = tokenizer.tokenize(sample[mention_key])
mention_tokens = [ent_start_token] + mention_tokens + [ent_end_token]
# Make sure that mention_tokens do not exceed max_seq_length
# And that at least one token from left and right context are used (left_quota >= 1, right_quota >= 1)
assert max_seq_length > 4
mention_tokens = mention_tokens[:max_seq_length - 4]

context_left = sample[context_key + "_left"]
context_right = sample[context_key + "_right"]
Expand Down