vad model conversion shows the same error as previous fixed

[tensorbuffer]

Issue Type

Others

OS

Linux

onnx2tf version number

1.28.2

onnx version number

1.17.0

onnxruntime version number

1.18.1

onnxsim (onnx_simplifier) version number

0.4.33

tensorflow version number

2.19.0

Download URL for ONNX

https://github.com/snakers4/silero-vad/blob/master/src/silero_vad/data/silero_vad.onnx

Parameter Replacement JSON

None

Description

1. Purpose: Run VAD model on device, it's for product development.
2. What:

onnx2tf -i silero_vad.onnx --overwrite_input_shape "input:1,512"

Depth of output (258) is not a multiple of the number of groups (576) for '{{node tf.nn.convolution/convolution}} = Conv2D[T=DT_FLOAT, data_format="NHWC", dilations=[1, 1, 1, 1], explicit_paddings=[], padding="VALID", strides=[1, 1, 128, 1], use_cudnn_on_gpu=true](tf.nn.convolution/convolution/ExpandDims, tf.nn.convolution/convolution/ExpandDims_1)' with input shapes: [1,1,1,576], [1,256,1,258].

4. How:
   I see two years ago someone asked this model and you fixed with Pad: [silero_vad] Question about Pad #42
   I checked the Pad.py on my system and I do see it's the newer version with that patch. However I don't know why I am seeing the same issue.

Also, from their code, the input shape should be (1, 576), e.g. input:1,1,576 state:1,576,128 sr:1
The 576 is from here: https://github.com/snakers4/silero-vad/blob/master/examples/cpp/silero-vad-onnx.cpp#L315

[PINTO0309]

silero_vad.onnx.zip

onnxslim silero_vad.onnx silero_vad.onnx
onnx2tf -i silero_vad.onnx -kat input state

INFO: 5 / 5
INFO: onnx_op_type: Unsqueeze onnx_op_name: If_0_then_branch__Inline_0__/decoder/rnn/Unsqueeze_3
INFO:  input_name.1: If_0_then_branch__Inline_0__/decoder/rnn/If_output_0 shape: None dtype: float32
INFO:  input_name.2: If_0_else_branch__Inline_0__/decoder/Constant_4_output_0 shape: (1,) dtype: int64
INFO:  output_name.1: If_0_then_branch__Inline_0__/decoder/rnn/Unsqueeze_3_output_0 shape: None dtype: float32
ERROR: The trace log is below.
Traceback (most recent call last):
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 314, in print_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 388, in inverted_operation_enable_disable_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 57, in get_replacement_parameter_wrapper_func
    func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/ops/Unsqueeze.py", line 131, in make_node
    new_shape = copy.deepcopy(input_tensor_shape)
UnboundLocalError: local variable 'input_tensor_shape' referenced before assignment

ERROR: input_onnx_file_path: silero_vad.onnx
ERROR: onnx_op_name: If_0_then_branch__Inline_0__/decoder/rnn/Unsqueeze_3
ERROR: Read this and deal with it. https://github.com/PINTO0309/onnx2tf#parameter-replacement
ERROR: Alternatively, if the input OP has a dynamic dimension, use the -b or -ois option to rewrite it to a static shape and try again.
ERROR: If the input OP of ONNX before conversion is NHWC or an irregular channel arrangement other than NCHW, use the -kt or -kat option.
ERROR: Also, for models that include NonMaxSuppression in the post-processing, try the -onwdt option.
ERROR: The trace log is below.
Traceback (most recent call last):
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 314, in print_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 388, in inverted_operation_enable_disable_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 57, in get_replacement_parameter_wrapper_func
    func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/ops/If.py", line 73, in make_node
    op.make_node(
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 314, in print_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 388, in inverted_operation_enable_disable_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 57, in get_replacement_parameter_wrapper_func
    func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/ops/Unsqueeze.py", line 131, in make_node
    new_shape = copy.deepcopy(input_tensor_shape)
UnboundLocalError: local variable 'input_tensor_shape' referenced before assignment

ERROR: input_onnx_file_path: silero_vad.onnx
ERROR: onnx_op_name: If_0_then_branch__Inline_0__/decoder/If_1
ERROR: Read this and deal with it. https://github.com/PINTO0309/onnx2tf#parameter-replacement
ERROR: Alternatively, if the input OP has a dynamic dimension, use the -b or -ois option to rewrite it to a static shape and try again.
ERROR: If the input OP of ONNX before conversion is NHWC or an irregular channel arrangement other than NCHW, use the -kt or -kat option.
ERROR: Also, for models that include NonMaxSuppression in the post-processing, try the -onwdt option.
ERROR: The trace log is below.
Traceback (most recent call last):
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 314, in print_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 388, in inverted_operation_enable_disable_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 57, in get_replacement_parameter_wrapper_func
    func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/ops/If.py", line 73, in make_node
    op.make_node(
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 314, in print_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 388, in inverted_operation_enable_disable_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 57, in get_replacement_parameter_wrapper_func
    func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/ops/If.py", line 73, in make_node
    op.make_node(
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 314, in print_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 388, in inverted_operation_enable_disable_wrapper_func
    result = func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/utils/common_functions.py", line 57, in get_replacement_parameter_wrapper_func
    func(*args, **kwargs)
  File "/home/b920405/.local/lib/python3.10/site-packages/onnx2tf/ops/Unsqueeze.py", line 131, in make_node
    new_shape = copy.deepcopy(input_tensor_shape)
UnboundLocalError: local variable 'input_tensor_shape' referenced before assignment

ERROR: input_onnx_file_path: silero_vad.onnx
ERROR: onnx_op_name: If_0
ERROR: Read this and deal with it. https://github.com/PINTO0309/onnx2tf#parameter-replacement
ERROR: Alternatively, if the input OP has a dynamic dimension, use the -b or -ois option to rewrite it to a static shape and try again.
ERROR: If the input OP of ONNX before conversion is NHWC or an irregular channel arrangement other than NCHW, use the -kt or -kat option.
ERROR: Also, for models that include NonMaxSuppression in the post-processing, try the -onwdt option.

[tensorbuffer]

Yes I can repro with your steps. I see you added TODO label, so I guess this will be fixed?