Thrsu opened a new issue, #15271: URL: https://github.com/apache/tvm/issues/15271
From the [official ONNX Changelog document](https://github.com/onnx/onnx/blob/main/docs/Changelog.md#ReduceL2-18), we can see that in Opset 18 of ONNX, the ReduceL2 operator has been modified to accept the axis parameter from an input tensor instead of an attribute. The same issue exists for ReduceL1, ReduceL3, ReduceSumSquare, and ReduceLogSum operators. ### Actual behavior The crash message is as below: ``` Traceback (most recent call last): ...... irmod, params = relay.frontend.from_onnx(model, shape_dict, opset=18) File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/onnx.py", line 7346, in from_onnx mod, params = g.from_onnx(graph, opset) File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/onnx.py", line 6963, in from_onnx self._construct_nodes(graph) File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/onnx.py", line 7078, in _construct_nodes op = self._convert_operator(op_name, inputs, attr, self.opset) File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/onnx.py", line 7204, in _convert_operator sym = convert_map[op_name](inputs, attrs, self._params) File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/onnx.py", line 3266, in _impl_v1 out = AttrCvt("sum")(inputs, attr) File "/workplace/software/tvm/tvm_/python/tvm/relay/frontend/common.py", line 453, in __call__ return get_relay_op(op_name)(*inputs, **new_attrs) TypeError: sum() got multiple values for argument 'axis' ``` ### Environment TVM version: 0.13.dev0 ONNX version: 1.14.0 ONNX Opset version: 18 ### Steps to reproduce You can download the [onnx model](https://drive.google.com/file/d/1F42GQK78WS8OTHNGSfDkWQcQE747CStC/view?usp=share_link) and use the script below to load it: ``` import onnx import tvm import tvm.relay as relay shape_dict = {'data': [3, 2, 2], 'axes': [0]} irmod, params = relay.frontend.from_onnx(model, shape_dict, opset=18) ``` ### Triage * frontend:onnx * needs-triage -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
