By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement . We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account c++ 推理错误 NotFoundError: Operator (transpose2) does not have kernel for data_type[double]:data_layout #6638 c++ 推理错误 NotFoundError: Operator (transpose2) does not have kernel for data_type[double]:data_layout #6638 dtiny opened this issue Jun 20, 2022 · 3 comments 训练算法：starnet_r34_vd_tps_bilstm_ctc.yml
训练语言：python
模型导出：参考python3 tools/export_model.py -c configs/rec/rec_r34_vd_tps_bilstm_ctc.yml -o Global.pretrained_model=./rec_r34_vd_tps_bilstm_ctc_v2.0_train/best_accuracy Global.save_inference_dir=./inference/rec_starnet
不同点在于我训练的时候使用图片输入尺寸为[3,24,192]，可以不报异常导出模型参数。
python推理测试：使用导出模型参数可以正确得到推理结果。
c++推理：按照demo，更改了输入尺寸为[3,24,192]，模型的初始化，数据处理都正常，但在执行this->predictor_->Run()这句时候保存。错误信息如下：
C++ Traceback (most recent call last):
0 paddle::AnalysisPredictor::ZeroCopyRun()
1 paddle::framework::NaiveExecutor::Run()
2 paddle::framework::OperatorBase::Run(paddle::framework::Scope const&, paddle::platform::Place const&)
3 paddle::framework::OperatorWithKernel::RunImpl(paddle::framework::Scope const&, paddle::platform::Place const&) const
4 paddle::framework::OperatorWithKernel::RunImpl(paddle::framework::Scope const&, paddle::platform::Place const&, paddle::framework::RuntimeContext*) const
5 paddle::framework::OperatorWithKernel::ChooseKernel(paddle::framework::RuntimeContext const&, paddle::framework::Scope const&, paddle::platform::Place const&) const
6 paddle::platform::EnforceNotMet::EnforceNotMet(paddle::platform::ErrorSummary const&, char const*, int)
7 paddle::platform::GetCurrentTraceBackString[abi:cxx11]
NotFoundError: Operator (transpose2) does not have kernel for data_type[double]:data_layout[ANY_LAYOUT]:place[CPUPlace]:library_type[PLAIN].
[Hint: Expected kernel_iter != kernels.end(), but received kernel_iter == kernels.end().] (at /paddle/paddle/fluid/framework/operator.cc:1236)
[operator < transpose2 > error]
Aborted (core dumped)

另外，官方文档说支持转onnx，但是实际测试下来，rec_r34_vd_none_bilstm_ctc_v2.0是成功的，但是rec_r34_vd_tps_bilstm_ctc_v2.0在转换的时候也是失败的，转换命令和错误信息如下：
命令：paddle2onnx --model_dir ./pp_models/rec_r34_vd_tps_bilstm_ctc_v2.0_train
--model_filename pp_models/rec_r34_vd_tps_bilstm_ctc_v2.0_train/inference.pdmodel
--params_filename pp_models/rec_r34_vd_tps_bilstm_ctc_v2.0_train/inference.pdiparams
--save_file ./pp_models/rec_r34_vd_tps_bilstm_ctc_v2.0_train/model.onnx
--opset_version 10
--input_shape_dict="{'x':[-1,3,-1,-1]}"
--enable_onnx_checker True
2022-06-16 16:02:39 [WARNING] [Deprecated] paddle2onnx.command.program2onnx will be deprecated in the future version, the recommended usage is paddle2onnx.export
Traceback (most recent call last):
File "/root/anaconda3/envs/paddle_det/bin/paddle2onnx", line 8, in
sys.exit(main())
File "/root/anaconda3/envs/paddle_det/lib/python3.8/site-packages/paddle2onnx/command.py", line 214, in main
program2onnx(
File "/root/anaconda3/envs/paddle_det/lib/python3.8/site-packages/paddle2onnx/command.py", line 145, in program2onnx
return program2onnx(model_dir, save_file, model_filename, params_filename,
File "/root/anaconda3/envs/paddle_det/lib/python3.8/site-packages/paddle2onnx/legacy/command.py", line 199, in program2onnx
program.blocks[0].ops[i].desc.infer_shape(program.blocks[0].desc)
ValueError: (InvalidArgument) Dims of all Inputs(X) must be the same, but received input 1 dim is:10 not equal to input 0 dim:-1.
[Hint: Expected input_dims[i] == input_dims[0], but received input_dims[i]:10 != input_dims[0]:-1.] (at /paddle/paddle/phi/infermeta/multiary.cc:1803)
[operator < stack > error]