参考链接：
https://gitee.com/ascend/ModelZoo-PyTorch/tree/master/ACL_PyTorch/contrib/cv/classfication/MaskRcnn
https://gitee.com/ascend/msit/tree/master/msit/docs/debug/compare
一、问题现象（附报错日志上下文：
maskrcnn精度比对报错;然后使用onnx推理报同样的错**
报错打屏信息：
root@localhost:/home/maskrcnn# msit debug compare -gm ./model.onnx --om maskrcnn_npu.om [INFO] Start to converting the model to json
[INFO] ATC command line /usr/local/Ascend/ascend-toolkit/8.0.RC2/aarch64-linux/bin/atc --mode=1 --om=/home/maskrcnn/maskrcnn_npu.om --json=/home/maskrcnn/20241017175633/model/maskrcnn_npu.json
[INFO] Execute command:/usr/local/Ascend/ascend-toolkit/8.0.RC2/aarch64-linux/bin/atc --mode=1 --om=/home/maskrcnn/maskrcnn_npu.om --json=/home/maskrcnn/20241017175633/model/maskrcnn_npu.json [INFO] Complete model conversion to json /home/maskrcnn/20241017175633/model/maskrcnn_npu.json. [INFO] The /home/maskrcnn/20241017175633/model/maskrcnn_npu.json file is exists.
[INFO] The input of model is not dynamic.
[INFO] acl init success
[INFO] open device 0 success
[INFO] load model /home/maskrcnn/maskrcnn_npu.om success
[INFO] create model description success
[INFO] unload model success, model Id is 1
[ERROR] Please check onnx model can run in local env. Error: [ONNXRuntimeError] : 10 : INVALID_GRAPH : Load model from /home/maskrcnn/model.onnx failed:This is an invalid model. In Node, ("BatchMultiClassNMS_1351", BatchMultiClassNMS, "", -1) : ("1918": tensor(float16),"1921": tensor(float16),) -> ("1922","1923","1924","1925",) , Error No Op registered for BatchMultiClassNMS with domain_version of 11
Traceback (most recent call last):
File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/onnx_model/onnx_dump_data.py", line 141, in _load_session infersession = onnxruntime.InferenceSession(model_path, options) File "/usr/local/python3.9.0/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 383, in init self._create_inference_session(providers, provider_options, disabled_optimizers) File "/usr/local/python3.9.0/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 424, in _create_inference_session sess = C.InferenceSession(session_options, self._model_path, True, self._read_config_from_model) onnxruntime.capi.onnxruntime_pybind11_state.InvalidGraph: [ONNXRuntimeError] : 10 : INVALID_GRAPH : Load model from /home/maskrcnn/model.onnx failed:This is an invalid model. In Node, ("BatchMultiClassNMS_1351", BatchMultiClassNMS, "", -1) : ("1918": tensor(float16),"1921": tensor(float16),) -> ("1922","1923","1924","1925",) , Error No Op registered for BatchMultiClassNMS with domain_version of 11
During handling of the above exception, another exception occurred:
Traceback (most recent call last): File "/usr/local/python3.9.0/lib/python3.9/site-packages/components/main.py", line 38, in main args.handle(args) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/main.py", line 198, in handle cmp_process(cmp_args, True) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 200, in cmp_process raise error File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 194, in cmp_process check_and_run(args, use_cli) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 337, in check_and_run res = run(args, input_shape, original_out_path, use_cli) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 236, in run golden_dump = _generate_golden_data_model(args, npu_dump_npy_path) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 77, in _generate_golden_data_model return OnnxDumpData(args, npu_dump_npy_path) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/onnx_model/onnx_dump_data.py", line 90, in init self.model_with_inputs_session = self._load_session(self.model_path) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/onnx_model/onnx_dump_data.py", line 144, in _load_session raise utils.AccuracyCompareException(utils.ACCURACY_COMPARISON_MODEL_TYPE_ERROR) msquickcmp.common.utils.AccuracyCompareException
The above exception was the direct cause of the following exception:
Traceback (most recent call last): File "/usr/local/python3.9.0/bin/msit", line 8, in sys.exit(main()) File "/usr/local/python3.9.0/lib/python3.9/site-packages/components/main.py", line 40, in main raise Exception( Exception: [ERROR] Refer FAQ if a known issue: gitee repo: Ascend/msit, wiki [INFO] end to reset device 0 [INFO] end to finalize acl
二、软件版本: -- CANN 版本 : 6.0.1 / 8.0.RC2
--Pytorch 版本: 1.8 / 1.11 --Python 版本 (e.g., Python 3.7.5): 3.7.5 / 3.9.0
--操作系统版本 (e.g., Ubuntu 18.04): Ubuntu22.04
--onnx版本：1.8.0/1.12.0/1.15.0
三、测试步骤：
参考官网链接导出onnx模型
1、转换为om模型正常，可以正常推理
2、 执行精度比对报错
3、查询模型opset版本
当前的Opset版本: 11
Domain: ai.onnx, Opset Version: 11
四、日志信息:
点赞
收藏
回复
分享
举报
已解决
参考链接：
https://gitee.com/ascend/ModelZoo-PyTorch/tree/master/ACL_PyTorch/contrib/cv/classfication/MaskRcnn
https://gitee.com/ascend/msit/tree/master/msit/docs/debug/compare
一、问题现象（附报错日志上下文：
maskrcnn精度比对报错;然后使用onnx推理报同样的错**
报错打屏信息：
root@localhost:/home/maskrcnn# msit debug compare -gm ./model.onnx --om maskrcnn_npu.om [INFO] Start to converting the model to json
[INFO] ATC command line /usr/local/Ascend/ascend-toolkit/8.0.RC2/aarch64-linux/bin/atc --mode=1 --om=/home/maskrcnn/maskrcnn_npu.om --json=/home/maskrcnn/20241017175633/model/maskrcnn_npu.json
[INFO] Execute command:/usr/local/Ascend/ascend-toolkit/8.0.RC2/aarch64-linux/bin/atc --mode=1 --om=/home/maskrcnn/maskrcnn_npu.om --json=/home/maskrcnn/20241017175633/model/maskrcnn_npu.json [INFO] Complete model conversion to json /home/maskrcnn/20241017175633/model/maskrcnn_npu.json. [INFO] The /home/maskrcnn/20241017175633/model/maskrcnn_npu.json file is exists.
[INFO] The input of model is not dynamic.
[INFO] acl init success
[INFO] open device 0 success
[INFO] load model /home/maskrcnn/maskrcnn_npu.om success
[INFO] create model description success
[INFO] unload model success, model Id is 1
[ERROR] Please check onnx model can run in local env. Error: [ONNXRuntimeError] : 10 : INVALID_GRAPH : Load model from /home/maskrcnn/model.onnx failed:This is an invalid model. In Node, ("BatchMultiClassNMS_1351", BatchMultiClassNMS, "", -1) : ("1918": tensor(float16),"1921": tensor(float16),) -> ("1922","1923","1924","1925",) , Error No Op registered for BatchMultiClassNMS with domain_version of 11
Traceback (most recent call last):
File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/onnx_model/onnx_dump_data.py", line 141, in _load_session infersession = onnxruntime.InferenceSession(model_path, options) File "/usr/local/python3.9.0/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 383, in init self._create_inference_session(providers, provider_options, disabled_optimizers) File "/usr/local/python3.9.0/lib/python3.9/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 424, in _create_inference_session sess = C.InferenceSession(session_options, self._model_path, True, self._read_config_from_model) onnxruntime.capi.onnxruntime_pybind11_state.InvalidGraph: [ONNXRuntimeError] : 10 : INVALID_GRAPH : Load model from /home/maskrcnn/model.onnx failed:This is an invalid model. In Node, ("BatchMultiClassNMS_1351", BatchMultiClassNMS, "", -1) : ("1918": tensor(float16),"1921": tensor(float16),) -> ("1922","1923","1924","1925",) , Error No Op registered for BatchMultiClassNMS with domain_version of 11
During handling of the above exception, another exception occurred:
Traceback (most recent call last): File "/usr/local/python3.9.0/lib/python3.9/site-packages/components/main.py", line 38, in main args.handle(args) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/main.py", line 198, in handle cmp_process(cmp_args, True) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 200, in cmp_process raise error File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 194, in cmp_process check_and_run(args, use_cli) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 337, in check_and_run res = run(args, input_shape, original_out_path, use_cli) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 236, in run golden_dump = _generate_golden_data_model(args, npu_dump_npy_path) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/cmp_process.py", line 77, in _generate_golden_data_model return OnnxDumpData(args, npu_dump_npy_path) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/onnx_model/onnx_dump_data.py", line 90, in init self.model_with_inputs_session = self._load_session(self.model_path) File "/usr/local/python3.9.0/lib/python3.9/site-packages/msquickcmp/onnx_model/onnx_dump_data.py", line 144, in _load_session raise utils.AccuracyCompareException(utils.ACCURACY_COMPARISON_MODEL_TYPE_ERROR) msquickcmp.common.utils.AccuracyCompareException
The above exception was the direct cause of the following exception:
Traceback (most recent call last): File "/usr/local/python3.9.0/bin/msit", line 8, in sys.exit(main()) File "/usr/local/python3.9.0/lib/python3.9/site-packages/components/main.py", line 40, in main raise Exception( Exception: [ERROR] Refer FAQ if a known issue: gitee repo: Ascend/msit, wiki [INFO] end to reset device 0 [INFO] end to finalize acl
二、软件版本: -- CANN 版本 : 6.0.1 / 8.0.RC2
--Pytorch 版本: 1.8 / 1.11 --Python 版本 (e.g., Python 3.7.5): 3.7.5 / 3.9.0
--操作系统版本 (e.g., Ubuntu 18.04): Ubuntu22.04
--onnx版本：1.8.0/1.12.0/1.15.0
三、测试步骤：
参考官网链接导出onnx模型
1、转换为om模型正常，可以正常推理
2、 执行精度比对报错
3、查询模型opset版本
当前的Opset版本: 11
Domain: ai.onnx, Opset Version: 11
四、日志信息: