将VITS模型转换为ONNX

[gyu-bbang]

你好，我是一名来自韩国的学生，正在学习"VITS TTS"。

最近，在进行"将VITS模型转换为ONNX"的过程中，我发现了您的GitHub存储库。

通过您的代码，我遇到了一些困难，想请教一下如何将韩语预训练模型转换为ONNX。

我遇到的问题如下：

如果您觉得回答有困难，也请给予一些建议或鼓励：）

谢谢！

INFO:root:Loaded checkpoint '/home/ubuntu/dev/VITS-Training/models/standard-3/G_220000.pth' (iteration 3438)
35 70
/home/ubuntu/dev/vits-onnx/vits/attentions.py:216: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
assert t_s == t_t, msg
/home/ubuntu/dev/vits-onnx/vits/attentions.py:272: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
pad_length = max(length - (self.window_size + 1), 0)
/home/ubuntu/dev/vits-onnx/vits/attentions.py:273: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
slice_start_position = max((self.window_size + 1) - length, 0)
/home/ubuntu/dev/vits-onnx/vits/attentions.py:275: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
if pad_length > 0:
/home/ubuntu/dev/vits-onnx/vits/transforms.py:106: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
if torch.min(inputs) < left or torch.max(inputs) > right:
/home/ubuntu/dev/vits-onnx/vits/transforms.py:111: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
if min_bin_width * num_bins > 1.0:
/home/ubuntu/dev/vits-onnx/vits/transforms.py:113: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
if min_bin_height * num_bins > 1.0:
/home/ubuntu/dev/vits-onnx/vits/transforms.py:165: TracerWarning: Converting a tensor to a Python boolean might cause the trace to be incorrect. We can't record the data flow of Python values, so this value will be treated as a constant in the future. This means that the trace might not generalize to other inputs!
assert (discriminant >= 0).all()
/home/ubuntu/miniconda3/envs/pt112_py39/lib/python3.9/site-packages/torch/jit/_trace.py:992: TracerWarning: Trace had nondeterministic nodes. Did you forget call .eval() on your model? Nodes:
%7760 : Float(1, 192, 60, strides=[11520, 1, 192], requires_grad=0, device=cpu) = aten::randn_like(%m_p, %7755, %7756, %7757, %7758, %7759) # /home/ubuntu/dev/vits-onnx/vits/models.py:649:0
This may cause errors in trace checking. To disable trace checking, pass check_trace=False to torch.jit.trace()
_check_trace(
/home/ubuntu/miniconda3/envs/pt112_py39/lib/python3.9/site-packages/torch/jit/_trace.py:992: TracerWarning: Output nr 1. of the traced function does not match the corresponding output of the Python function. Detailed error:
The values for attribute 'shape' do not match: torch.Size([1, 1, 16896]) != torch.Size([1, 1, 20480]).
_check_trace(
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Warning: Constant folding - Only steps=1 can be constant folded for opset >= 10 onnx::Slice op. Constant folding not applied.
Floating point exception (core dumped)