RUNTIME_EXCEPTION : Non-zero status code returned while running Mul node. Name:'Mul_5'

Describe the bug A clear and concise description of what the bug is. To avoid repetition please make sure this is not one of the known issues mentioned on the respective release page.

System information

OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Windows 10 19044.1889
ONNX Runtime installed from (binary):
ONNX Runtime version: 1.12.1
Python version: 3.9
Visual Studio version (if applicable):
GCC/Compiler version (if compiling from source):
CUDA/cuDNN version: 11.5
GPU model and memory: Nvidia 2070 8G

To Reproduce

Describe steps/code to reproduce the behavior.
Attach the ONNX model to the issue (where applicable) to expedite investigation.

Export options

# infer_ins is a derived class from torch.nn.Module
infer_ins.pe.to('cpu')
with torch.no_grad():
mel_input = torch.rand(1, 968, 80)

torch.onnx.export(
    infer_ins.pe,
    (
        mel_input
    ),
    "xiaoma_pe.onnx",
    verbose=True,
    input_names=["mel_input"],
    output_names=["f0"],
    dynamic_axes={
        "mel_input": {
            0: "batch_size",
            1: "frames",
            2: "num_mel_bin",
        },
        "f0": {
            0: "batch_size",
            1: "frames"
        }
    },
    opset_version=11,
)

Forward function

def forward(self, mel_input=None):
ret = {}
mel_hidden = self.mel_prenet(mel_input)[1]
if self.conv_layers > 0:
    mel_hidden = self.mel_encoder(mel_hidden)

ret['pitch_pred'] = pitch_pred = self.pitch_predictor(mel_hidden)

pitch_padding = mel_input.abs().sum(-1) == 0
use_uv = hparams['pitch_type'] == 'frame' and hparams['use_uv']

ret['f0_denorm_pred'] = denorm_f0(
    pitch_pred[:, :, 0], (pitch_pred[:, :, 1] > 0) if use_uv else None,
    hparams, pitch_padding=pitch_padding)
return ret

Use case

mel_input = torch.rand(1, 967, 80)
pe2 = ort.InferenceSession("xiaoma_pe.onnx")
f0 = self.pe2.run(None, 
{
    'mel_input': to_numpy(mel_out)
}
)

Expected behavior A clear and concise description of what you expected to happen. No Error.

Screenshots If applicable, add screenshots to help explain your problem.

2022-08-16 00:07:27.3322084 [E:onnxruntime:, sequential_executor.cc:368 onnxruntime::SequentialExecutor::Execute] Non-zero status code returned while running Mul node. Name:'Mul_5' Status Message: D:\a\_work\1\s\onnxruntime\core/providers/cpu/math/element_wise_ops.h:503 onnxruntime::BroadcastIterator::Init axis == 1 || axis == largest was false. Attempting to broadcast an axis by a dimension other than 1. 967 by 968

Traceback (most recent call last):
  File "C:\Users\truef\Documents\GitHub\DiffSinger\onnx_test_hifigan_pe.py", line 99, in <module>
    out = infer_ins.infer_once(c)
  File "C:\Users\truef\Documents\GitHub\DiffSinger\inference\svs\base_svs_infer.py", line 235, in infer_once
    output = self.forward_model(inp)
  File "C:\Users\truef\Documents\GitHub\DiffSinger\onnx_test_hifigan_pe.py", line 68, in forward_model
    f0 = self.pe2.run(None,
  File "C:\CodeEnv\miniconda3\envs\dfs\lib\site-packages\onnxruntime\capi\onnxruntime_inference_collection.py", line 200, in run
    return self._sess.run(output_names, input_feed, run_options)
onnxruntime.capi.onnxruntime_pybind11_state.RuntimeException: [ONNXRuntimeError] : 6 : RUNTIME_EXCEPTION : Non-zero status code returned while running Mul node. Name:'Mul_5' Status Message: D:\a\_work\1\s\onnxruntime\core/providers/cpu/math/element_wise_ops.h:503 onnxruntime::BroadcastIterator::Init axis == 1 || axis == largest was false. Attempting to broadcast an axis by a dimension other than 1. 967 by 968

Additional context Add any other context about the problem here. If the issue is about a particular model, please share the model details as well to facilitate debugging.

microsoft / onnxruntime

RUNTIME_EXCEPTION : Non-zero status code returned while running Mul node. Name:'Mul_5' #12595