Skip to content

Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.

fix bug

游雁310006dd60bf
diff --git a/funasr/datasets/openai_datasets/datasets.py b/funasr/datasets/openai_datasets/datasets.py
index 0d12a1c4f..3bc6aa011 100644
--- a/funasr/datasets/openai_datasets/datasets.py
+++ b/funasr/datasets/openai_datasets/datasets.py
@@ -162,7 +162,7 @@ class OpenAIDataset(torch.utils.data.Dataset):
if badcase_flag:
continue
input_ids = torch.tensor(input_ids, dtype=torch.int64) # [: self.max_token_length]
- attention_mask = torch.tensor([len(input_ids)], dtype=torch.int32)
+ attention_mask = torch.tensor([1] * len(input_ids), dtype=torch.int32)
labels = torch.tensor(labels, dtype=torch.int64) # [: self.max_token_length]
fbank = speech[0, :, :]