Skip to content

Fix(fun-asr-dataset): fix dynamic masking error in input_ids #2801

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

Merged
LauraGPT merged 1 commit into from
Feb 4, 2026
Merged

Fix(fun-asr-dataset): fix dynamic masking error in input_ids #2801

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
7 changes: 4 additions & 3 deletions funasr/datasets/fun_asr_datasets/datasets.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -258,8 +258,9 @@ def __getitem__(self, index):
f"text_length: {len(target_ids)} > {self.max_target_length}, drop it: {item}"
)
# simulate prev-token fixed output
target_labels = target_ids.copy()
if np.random.rand() < self.use_dynamic_output_ratio:
max_len = len(target_ids)
max_len = len(target_labels)
min_output_mask_token_len = min(self.min_output_mask_token_len, max_len)
min_output_non_mask_token_len = min(self.min_output_non_mask_token_len, max_len)
if max_len - min_output_non_mask_token_len > min_output_mask_token_len:
Expand All @@ -268,10 +269,10 @@ def __getitem__(self, index):
else:
end_index = max_len - min_output_non_mask_token_len
if end_index > 0:
target_ids[:end_index] = [-100] * end_index
target_labels[:end_index] = [-100] * end_index

input_ids += source_ids + target_ids
labels += source_mask + target_ids
labels += source_mask + target_labels
if len(speech) > 0:
fbank.append(speech[0, :, :])
fbank_lens.append(speech_lengths)
Expand Down