#7096 语音识别阶段出错[Huggingface_ASR(内置)] The expanded size of the tensor (3104) must match the existing size (3000) at non-singleto

该错误是 Huggingface_ASR 渠道使用的模型在批量处理音频切片时，音频特征张量长度不一致（3000 vs 3104），导致 RuntimeError。最常见原因是模型文件下载不完整或损坏，也可能是该模型与当前音频切片长度不兼容。

按以下顺序尝试解决：

1. 重新下载并覆盖模型：根据你的模型 kotoba-tech/kotoba-whisper-v2.0，删除本地 软件目录/models/models--kotoba-tech--kotoba-whisper-v2.0 文件夹，然后重新在软件中运行一次，让软件自动下载；若自动下载失败，请手动从 https://huggingface.co/kotoba-tech/kotoba-whisper-v2.0 下载所有文件，放入上述文件夹内。注意：模型内的文件名不能有括号或重命名。
2. 更换语音识别渠道：如果重新下载后仍报错，请改用 faster-whisper(内置) 或 openai-whisper(内置) 渠道，模型选择 large-v3-turbo 或 large-v3，日语效果也很好。
3. 检查 VAD/切片设置：进入 菜单 → 工具 → 高级选项 → 语音识别参数，将“最长语音持续(秒)”调大（例如 60），“最短语音持续(毫秒)”调大（例如 1000），减少短切片导致的形状不一致。

如果以上方法均无法解决，请提供软件安装目录下 logs/ 文件夹中最新日志文件（复制底部约30行报错信息）以便进一步分析。

请查阅相关文档：

• Huggingface_ASR渠道
• 手动下载模型

原始输入文件名: C:/Users/1/Desktop/BOBB-442J.mp4,
输出结果保存到文件夹: D:/win-pyvideotrans-v4.11/output/recogn,
临时文件夹: D:/win-pyvideotrans-v4.11/tmp/21316/2adbc3d2c8
未启用CUDA加速
未选 降噪
语音识别:Huggingface_ASR(内置), kotoba-tech/kotoba-whisper-v2.0 断句方式:默认断句
发音语言:ja
未选 降噪

File "videotrans\process\stt_pipe.py", line 65, in pipe_asr
File "D:\win-pyvideotrans-v4.11\_internal\transformers\pipelines\automatic_speech_recognition.py", line 244, in call

return super().__call__(inputs, **kwargs)

File "D:\win-pyvideotrans-v4.11\_internal\transformers\pipelines\base.py", line 1245, in call

outputs = list(final_iterator)

File "D:\win-pyvideotrans-v4.11\_internal\transformers\pipelines\pt_utils.py", line 126, in next

item = next(self.iterator)

File "D:\win-pyvideotrans-v4.
......
line 35, in _exec
File "videotrans\recognition\_huggingface.py", line 71, in _pipe_asr
File "videotrans\configure\base.py", line 264, in _new_process
videotrans.configure.excepts.VideoTransError: The expanded size of the tensor (3104) must match the existing size (3000) at non-singleton dimension 1. Target sizes: [128, 3104]. Tensor sizes: [128, 3000]:Traceback (most recent call last):
File "videotrans\process\stt_pipe.py", line 65, in pipe_asr
File "D:\win-pyvideotrans-v4.11\_internal\transformers\pipelines\automatic_speech_recognition.py", line 244, in call

return super().__call__(inputs, **kwargs)

File "D:\win-pyvideotrans-v4.11\_internal\transformers\pipelines\base.py", line 1245, in call

outputs = list(final_iterator)

File "D:\win-pyvideotrans-v4.11\_internal\transformers\pipelines\pt_utils.py", line 126, in next

item = next(self.iterator)

File "D:\win-pyvideotrans-v4.11\_internal\transformers\pipelines\pt_utils.py", line 271, in next

processed = self.infer(next(self.iterator), **self.params)

File "D:\win-pyvideotrans-v4.11\_internal\torch\utils\data\dataloader.py", line 733, in next

data = self._next_data()

File "D:\win-pyvideotrans-v4.11\_internal\torch\utils\data\dataloader.py", line 789, in _next_data

data = self._dataset_fetcher.fetch(index)  # may raise StopIteration

File "D:\win-pyvideotrans-v4.11\_internal\torch\utils\data\_utils\fetch.py", line 43, in fetch

return self.collate_fn(data)

File "D:\win-pyvideotrans-v4.11\_internal\transformers\pipelines\base.py", line 173, in inner

padded[key] = _pad(items, key, _padding_value, padding_side)

File "D:\win-pyvideotrans-v4.11\_internal\transformers\pipelines\base.py", line 109, in _pad

tensor[i, : len(item[key][0])] = item[key][0]

RuntimeError: The expanded size of the tensor (3104) must match the existing size (3000) at non-singleton dimension 1. Target sizes: [128, 3104]. Tensor sizes: [128, 3000]

该报错是模型文件不完整或损坏，导致音频特征张量长度不匹配（3000 与 3104 不一致）。按以下顺序处理：

1. 删除模型文件夹 D:/win-pyvideotrans-v4.11/models/models--kotoba-tech--kotoba-whisper-v2.0，回到软件重跑，让软件自动重新下载。
2. 若自动下载仍失败，手动下载该模型所有文件（https://huggingface.co/kotoba-tech/kotoba-whisper-v2.0），放入上述文件夹，确保文件名无括号、无重命名。
3. 若重试后仍报错，请改用 faster-whisper(内置) 渠道，模型选择 large-v3-turbo 或 large-v3，日语识别效果好。

请查阅相关文档：

• Huggingface_ASR渠道
• 手动下载模型