Я получил ошибку, когда импортировал обученную модель в плату разработки Baidu Edgeboard DK-1B ⇐ Python

[Гость]

--- speedLow:0.8m/s | speedHigh:0.8m/s
tar: lib0.o: time stamp 2025-05-08 19:45:17 is 68940855.041611686 s in the future
tar: devc.o: time stamp 2025-05-08 19:45:17 is 68940855.039901724 s in the future
compile done.
tar: nna_mbs.VhNyv3: time stamp 2025-05-08 19:50:25 is 68941162.704824917 s in the future
INFO: Heap :ocm (0x18)
INFO: Heap :anonymous (0x2)
INFO: Heap :dmabuf (0x2)
INFO: Heap :unified (0x5)
[ WARN:0] OpenCV | GStreamer warning: Embedded video playback halted; module source reported: Could not read from resource.
[ WARN:0] OpenCV | GStreamer warning: unable to start pipeline
[ WARN:0] OpenCV | GStreamer warning: GStreamer: pipeline have not been created
--------------[等待按键发车!]-------------------
terminate called after throwing an instance of 'tvm::runtime::InternalError'
what(): [21:31:15] /home/edgeboard/work/nfs/PPNC-ToolChain/src/runtime/ndarray.cc:260:
---------------------------------------------------------------
An error occurred during the execution of TVM.
For more information, please see: https://tvm.apache.org/docs/errors.html
---------------------------------------------------------------

Check failed: from_size == to_size (1228800 vs. 4435968) : TVMArrayCopyFromTo: The size must exactly match
Stack trace:
[bt] (0) /usr/local/lib/ppnc/libppnc_runtime.so(tvm::runtime::Backtrace[abi:cxx11]()+0x30) [0x7fa607cca0]
[bt] (1) /usr/local/lib/ppnc/libppnc_runtime.so(tvm::runtime::detail::LogFatal::Entry::Finalize()+0x88) [0x7fa605d098]
[bt] (2) /usr/local/lib/ppnc/libppnc_runtime.so(tvm::runtime::NDArray::CopyFromTo(DLTensor const*, DLTensor*, void*)+0x280) [0x7fa608df00]
[bt] (3) /usr/local/lib/ppnc/libppnc_runtime.so(tvm::runtime::GraphExecutor::SetInput(int, DLTensor*)+0x70) [0x7fa60fdd50]
[bt] (4) /usr/local/lib/ppnc/libppnc_runtime.so(+0x11b724) [0x7fa6102724]
[bt] (5) /usr/local/lib/ppnc/libppnc_runtime.so(ppnc::api::PPNCPredictor::set_inputs(std::unordered_map const&)+0xfc) [0x7fa61208e4]
[bt] (6) ./icar(+0x1788c) [0x55677e688c]
[bt] (7) ./icar(+0x7e24) [0x55677d6e24]
[bt] (8) /lib/aarch64-linux-gnu/libc.so.6(__libc_start_main+0xe0) [0x7fa62167a0]
< /code>
После того, как я завершил обучение модели, я собрал модель и импортировал скомпилированную подход в плату Baidu и сообщил об ошибке
i на основе на основе на основе на основе на основе https://pan.baidu.com/s/1uipdet_or8p8nl ... %E4%BA%8B8, 7%94%A8%E5%8D%A1-%E6%A8%A1%E5%9E%8B%E7%BC%96%E8%AF%91%E6%95%99%E7%A8%8B & ParentPath =%2FSHARELINK. Сасу

Подробнее здесь: https://stackoverflow.com/questions/796 ... dk-1b-deve