Open-source speech recognition toolkit for training, inference, streaming ASR, VAD, punctuation, speaker diarization pipelines, and OpenAI-compatible/MCP serving.
Python asraudiochineseemotion-recognitionfunasrmcp-servermultilingual-asropenai-compatible-apiparaformerpunctuationpytorchreal-time-asrspeaker-diarizationspeech-recognitionspeech-to-textstreaming-asrtranscriptionvllmvoice-activity-detectionwhisper-alternative
fix bug
游雁310006dd60bf
diff --git a/funasr/datasets/openai_datasets/datasets.py b/funasr/datasets/openai_datasets/datasets.pyindex 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:continueinput_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, :, :]