Unsupported opset => Gather-11 (ai.onnx)

[lzlwakeup]

IR Version: v6 Producer: pytorch 1.11.0 Domain: Imports: ai.onnx v11 Conv_0: Conv-11 (ai.onnx) Inputs: input.1: float32[1 x 3 x 352 x 352] = [...] onnx::Conv_760: float32[24 x 3 x 3 x 3] = [...] onnx::Conv_761: float32[24] = [...] Outputs: input.4: float32[1 x 24 x 176 x 176] = [...]

...

Concat_264: Concat-11 (ai.onnx) Inputs: onnx::Concat_744: float32[1 x 1 x 22 x 22] = [...] onnx::Concat_960: float32[1 x 4 x 22 x 22] = [...] onnx::Concat_757: float32[1 x 80 x 22 x 22] = [...] Outputs: 758: float32[1 x 85 x 22 x 22] = [...] 110592 Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Gather-11 (ai.onnx) Unsupported opset => Resize-11 (ai.onnx) Unsupported opset => Pad-11 (ai.onnx)

Issue: Using https://github.com/dog-qiuqiu/FastestDet Fastest yolo, unsupported opset. Onnx model in /FastestDet/example/onnx-runtime.

算子不支持，在yolotiny测试中没问题，采用另一种轻量级模型出现该问题，想问下老哥是算子实现的问题还是别的。

[jianjunjiang]

还没支持Gather-11这个算子

[mjq2020]

还没支持Gather-11这个算子 还没支持Gather-11这个算子 Unsupported opset => Pad-8 (ai.onnx) Unsupported opset => Conv-8 (ai.onnx) Unsupported opset => LeakyRelu-8 (ai.onnx) Unsupported opset => Pad-8 (ai.onnx) Unsupported opset => Conv-8 (ai.onnx) Unsupported opset => LeakyRelu-8 (ai.onnx)

我用了一些简单的算子，但同样报错，我看程序示例的算子集是v8,所以在导出的时候也是用的v8算子集，但是还是报错。区别的是IR Version是不同的，我的是v7。请问这种是因为IR Version 不同导致的吗？或者遇到这种情况的时候应该怎么解决？