SenseVoice-ONNX在昇腾Ascend910B4-1显卡推理报错

一、问题现象（附报错日志上下文）：
xxxx

二、软件版本:
-- CANN 版本：910b_8.2.RC1
--Mindie镜像 版本: mindie:2.1.RC2-800I-A3-py311-openeuler24.03-lts
--Python 版本 (e.g., Python 3.11.6):
--操作系统版本 (e.g., openEuler 24.03 (LTS)):

三、测试步骤：
根据文档的执行步骤到：python3 infer_onnx.py --model_path=SenseVoiceSmall --om_path=SenseVoice_linux_aarch64.om --device=0 --input="./SenseVoiceSmall/example/zh.mp3" --perform=True --loop=20
当前目录下面的文件如下：![输入图片说明](https://foruda.gitee.com/images/1759028988947058573/d6c858a5_5246788.png "屏幕截图")

四、日志信息:
[root@gxai-infer-3 SenseVoice]# python3 infer_onnx.py --model_path=SenseVoiceSmall --om_path=SenseVoice_linux_aarch64.om --device=0 --input="./SenseVoiceSmall/example/zh.mp3" --perform=True --loop=20
/usr/local/lib64/python3.11/site-packages/torch_npu/contrib/transfer_to_npu.py:301: ImportWarning:
    *************************************************************************************************************
    The torch.Tensor.cuda and torch.nn.Module.cuda are replaced with torch.Tensor.npu and torch.nn.Module.npu now..
    The torch.cuda.DoubleTensor is replaced with torch.npu.FloatTensor cause the double type is not supported now..
    The backend in torch.distributed.init_process_group set to hccl now..
    The torch.cuda.* and torch.cuda.amp.* are replaced with torch.npu.* and torch.npu.amp.* now..
    The device parameters have been replaced with npu in the function below:
    torch.logspace, torch.randint, torch.hann_window, torch.rand, torch.full_like, torch.ones_like, torch.rand_like, torch.randperm, torch.arange, torch.frombuffer, torch.normal, torch._empty_per_channel_affine_quantized, torch.empty_strided, torch.empty_like, torch.scalar_tensor, torch.tril_indices, torch.bartlett_window, torch.ones, torch.sparse_coo_tensor, torch.randn, torch.kaiser_window, torch.tensor, torch.triu_indices, torch.as_tensor, torch.zeros, torch.randint_like, torch.full, torch.eye, torch._sparse_csr_tensor_unsafe, torch.empty, torch._sparse_coo_tensor_unsafe, torch.blackman_window, torch.zeros_like, torch.range, torch.sparse_csr_tensor, torch.randn_like, torch.from_file, torch._cudnn_init_dropout_state, torch._empty_affine_quantized, torch.linspace, torch.hamming_window, torch.empty_quantized, torch._pin_memory, torch.autocast, torch.load, torch.Generator, torch.set_default_device, torch.Tensor.new_empty, torch.Tensor.new_empty_strided, torch.Tensor.new_full, torch.Tensor.new_ones, torch.Tensor.new_tensor, torch.Tensor.new_zeros, torch.Tensor.to, torch.nn.Module.to, torch.nn.Module.to_empty
    *************************************************************************************************************

warnings.warn(msg, ImportWarning)
/usr/local/lib64/python3.11/site-packages/torch_npu/contrib/transfer_to_npu.py:260: RuntimeWarning: torch.jit.script and torch.jit.script_method will be disabled by transfer_to_npu, which currently does not support them, if you need to enable them, please do not use transfer_to_npu.
  warnings.warn(msg, RuntimeWarning)
/usr/local/lib/python3.11/site-packages/pydub/utils.py:14: DeprecationWarning: 'audioop' is deprecated and slated for removal in Python 3.13
  import audioop
[INFO] download models from model hub: ms
[WARNING] trust_remote_code: True
Loading remote code successfully: model
[INFO] Loading pretrained params from SenseVoiceSmall/model.pt
[INFO] ckpt: SenseVoiceSmall/model.pt
[INFO] scope_map: ['module.', 'None']
[INFO] excludes: None
[INFO] Loading ckpt: SenseVoiceSmall/model.pt, status: <All keys matched successfully>
[WARN] acl repeat initialize
[INFO] acl init success
[INFO] open device 0 success
[INFO] get current context
[INFO] load model SenseVoice_linux_aarch64.om success
[INFO] create model description success
[ACL ERROR] EZ9999: Inner Error!
EZ9999: [PID: 65596] 2025-09-28-11:05:15.541.984 Parse dynamic kernel config fail.
        TraceBack (most recent call last):
       AclOpKernelInit failed opType
       ConcatD ADD_TO_LAUNCHER_LIST_AICORE failed.
       call aclnnCat failed:[FUNC:ConcatExecuteFuncD][FILE:fallback_concat.cpp][LINE:57]
       api_ret faild:-1[FUNC:ConcatExecuteFuncD][FILE:fallback_concat.cpp][LINE:59]
       Assert ((op_execute_func(op_execute_context)) == ge::SUCCESS) failed[FUNC:ExecuteOpFunc][FILE:aclnn_op_execute_kernel.cc][LINE:47]
       [Exec][Model]Execute model failed, ge result[1343225857], modelId[2147483648][FUNC:ReportCallError][FILE:log_inner.cpp][LINE:161]
       [Exec][Model]modelId[2147483648] execute failed, result[500002][FUNC:ReportInnerError][FILE:log_inner.cpp][LINE:145]

[ERROR] execute model failed, modelId is 2147483648
[ERROR] acl execute failed:1
[ERROR] execute Infer failed ret:-1
Traceback (most recent call last):
  File "/workspace/ModelZoo-PyTorch/ACL_PyTorch/built-in/audio/SenseVoice/SenseVoice/infer_onnx.py", line 115, in <module>
    res, _ = m.infer_onnx(
             ^^^^^^^^^^^^^
  File "/workspace/ModelZoo-PyTorch/ACL_PyTorch/built-in/audio/SenseVoice/SenseVoice/infer_onnx.py", line 72, in infer_onnx
    ctc_logits, encoder_out_lens = om_sess.infer(feed, mode='dymshape', custom_sizes=10000000)
                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/ais_bench/infer/interface.py", line 247, in infer
    return self.run(inputs, out_array)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/ais_bench/infer/interface.py", line 184, in run
    outputs = self.session.run(self.outputs_names, inputs)
              ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
RuntimeError: [-1][ACL: general failure]
[ERROR] 2025-09-28-11:05:16 (PID:65596, Device:0, RankID:-1) ERR99999 UNKNOWN application exception
[INFO] unload model success, model Id is 2147483648
[WARN] acl repeat destroy

Ascend/ModelZoo-PyTorch

Content Risk Flag

Comments (0)

Ascend/ModelZoo-PyTorch .gitee-modal { width: 500px !important; }

Content Risk Flag