更新cuda和pytorch后还是有兼容问题

[JiangXiaoDong123]

@pymumu 你好，大佬。现在我在尝试使用英伟达4060ti型号显卡，cuda版本为13.0，libtorch为2.1.0，python为3.8.0，然后重新编译了modelbox的代码，然后在运行yolox官方的yolox_s.pt（通过脚本从pth格式导出的）后modelbox崩溃。错误如下。请问我上面有哪些操作有问题么，还要注意哪些兼容性问题么？谢谢。如果方便能够更新下docker构建脚本就多些了。或者给与一些指导支持

[2025-12-18 08:13:08,566][ INFO][session_context.cc:40 ] session context start se id:49bba1b5-cda0-4c30-9fb1-5a72c24f28d0
[2025-12-18 08:13:08,700][DEBUG][flowunit_group.cc:388 ] yolox_post: opened.
[2025-12-18 08:13:08,700][ INFO][flowunit_group.cc:431 ] node: yolox_post get batch size is 1
terminate called after throwing an instance of 'c10::Error'
what(): isTuple()INTERNAL ASSERT FAILED at "../aten/src/ATen/core/ivalue_inl.h":1306, please report a bug to PyTorch. Expected Tuple but got String
Exception raised from toTuple at ../aten/src/ATen/core/ivalue_inl.h:1306 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits, std::allocator >) + 0x6b (0x7f5be52ec70b in /usr/local/lib/libc10.so)
frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&) + 0xce (0x7f5be52e83ee in /usr/local/lib/libc10.so)
frame #2: c10::detail::torchInternalAssertFail(char const*, char const*, unsigned int, char const*, std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&) + 0x4e (0x7f5be52ea14e in /usr/local/lib/libc10.so)
frame #3: + 0x3b9409f (0x7f5bd542809f in /usr/local/lib/libtorch_cpu.so)
frame #4: + 0x3b95655 (0x7f5bd5429655 in /usr/local/lib/libtorch_cpu.so)
frame #5: torch::jit::SourceRange::highlight(std::ostream&) const + 0x3d (0x7f5bd284e68d in /usr/local/lib/libtorch_cpu.so)
frame #6: torch::jit::ErrorReport::what() const + 0x351 (0x7f5bd2831781 in /usr/local/lib/libtorch_cpu.so)
frame #7: TorchInferenceFlowUnit::LoadModel(std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&, std::shared_ptrmodelbox::Configuration const&) + 0x8d9 (0x7f5be55c18af in /usr/local/lib/libmodelbox-unit-cuda-torch_inference.so.1.0.0)
frame #8: TorchInferenceFlowUnit::InitConfig(std::shared_ptrmodelbox::Configuration const&) + 0x103 (0x7f5be55c1be1 in /usr/local/lib/libmodelbox-unit-cuda-torch_inference.so.1.0.0)
frame #9: TorchInferenceFlowUnit::Open(std::shared_ptrmodelbox::Configuration const&) + 0x23a (0x7f5be55c2056 in /usr/local/lib/libmodelbox-unit-cuda-torch_inference.so.1.0.0)
frame #10: + 0xa7daa4 (0x7f5c7bac0aa4 in /usr/local/lib/libmodelbox.so.1)
frame #11: + 0xa8314a (0x7f5c7bac614a in /usr/local/lib/libmodelbox.so.1)
frame #12: + 0xa830cd (0x7f5c7bac60cd in /usr/local/lib/libmodelbox.so.1)
frame #13: + 0xa82ef8 (0x7f5c7bac5ef8 in /usr/local/lib/libmodelbox.so.1)
frame #14: + 0xa82847 (0x7f5c7bac5847 in /usr/local/lib/libmodelbox.so.1)
frame #15: + 0xa8242e (0x7f5c7bac542e in /usr/local/lib/libmodelbox.so.1)
frame #16: + 0xa8206c (0x7f5c7bac506c in /usr/local/lib/libmodelbox.so.1)
frame #17: + 0xa81cf5 (0x7f5c7bac4cf5 in /usr/local/lib/libmodelbox.so.1)
frame #18: + 0xa82983 (0x7f5c7bac5983 in /usr/local/lib/libmodelbox.so.1)
frame #19: + 0xa8251d (0x7f5c7bac551d in /usr/local/lib/libmodelbox.so.1)
frame #20: std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>::operator()() const + 0x4c (0x5601c8a1fe1a in /usr/local/bin/modelbox)
frame #21: std::__future_base::_State_baseV2::_M_do_set(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool) + 0x36 (0x5601c8a1e8f8 in /usr/local/bin/modelbox)
frame #22: void std::__invoke_impl<void, void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::__invoke_memfun_deref, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&) + 0xae (0x5601c8a23584 in /usr/local/bin/modelbox)
frame #23: std::__invoke_result<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>::type std::__invoke<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&) + 0x7e (0x5601c8a213ed in /usr/local/bin/modelbox)
frame #24: std::call_once<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::once_flag&, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&)::{lambda()#1}::operator()() const + 0x79 (0x5601c8a1f63b in /usr/local/bin/modelbox)
frame #25: std::call_once<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::once_flag&, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&)::{lambda()#2}::operator()() const + 0x36 (0x5601c8a1f766 in /usr/local/bin/modelbox)
frame #26: std::call_once<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::once_flag&, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&)::{lambda()#2}::_FUN() + 0x21 (0x5601c8a1f79e in /usr/local/bin/modelbox)
frame #27: + 0xf907 (0x7f5c7ae33907 in /lib/x86_64-linux-gnu/libpthread.so.0)
frame #28: + 0x213fee (0x5601c8a13fee in /usr/local/bin/modelbox)
frame #29: void std::call_once<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::once_flag&, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&) + 0xa6 (0x5601c8a1f892 in /usr/local/bin/modelbox)
frame #30: std::__future_base::_State_baseV2::_M_set_result(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool) + 0x8c (0x5601c8a1e382 in /usr/local/bin/modelbox)
frame #31: + 0xa81d7b (0x7f5c7bac4d7b in /usr/local/lib/libmodelbox.so.1)
frame #32: std::packaged_task<modelbox::Status ()>::operator()() + 0x42 (0x5601c8aded7e in /usr/local/bin/modelbox)
frame #33: + 0xa7f747 (0x7f5c7bac2747 in /usr/local/lib/libmodelbox.so.1)
frame #34: + 0xa80026 (0x7f5c7bac3026 in /usr/local/lib/libmodelbox.so.1)
frame #35: std::function<void ()>::operator()() const + 0x41 (0x5601c8b4a851 in /usr/local/bin/modelbox)
frame #36: modelbox::ThreadPool::RunWorker(modelbox::ThreadWorker*) + 0x151 (0x7f5c7bdb029d in /usr/local/lib/libmodelbox.so.1)
frame #37: modelbox::ThreadWorker::Run(modelbox::ThreadWorker*) + 0x53 (0x7f5c7bdaf645 in /usr/local/lib/libmodelbox.so.1)
frame #38: void std::__invoke_impl<void, void ()(modelbox::ThreadWorker), modelbox::ThreadWorker*>(std::__invoke_other, void (&&)(modelbox::ThreadWorker), modelbox::ThreadWorker*&&) + 0x4a (0x7f5c7bdb9ced in /usr/local/lib/libmodelbox.so.1)
frame #39: std::__invoke_result<void ()(modelbox::ThreadWorker), modelbox::ThreadWorker*>::type std::__invoke<void ()(modelbox::ThreadWorker), modelbox::ThreadWorker*>(void (&&)(modelbox::ThreadWorker), modelbox::ThreadWorker*&&) + 0x4e (0x7f5c7bdb952d in /usr/local/lib/libmodelbox.so.1)
frame #40: decltype (__invoke((_S_declval<0ul>)(), (_S_declval<1ul>)())) std::thread::_Invoker<std::tuple<void ()(modelbox::ThreadWorker), modelbox::ThreadWorker*> >::_M_invoke<0ul, 1ul>(std::_Index_tuple<0ul, 1ul>) + 0x59 (0x7f5c7bdbaccd in /usr/local/lib/libmodelbox.so.1)
frame #41: std::thread::_Invoker<std::tuple<void ()(modelbox::ThreadWorker), modelbox::ThreadWorker*> >::operator()() + 0x2c (0x7f5c7bdbaa06 in /usr/local/lib/libmodelbox.so.1)
frame #42: std::thread::_State_impl<std::thread::_Invoker<std::tuple<void ()(modelbox::ThreadWorker), modelbox::ThreadWorker*> > >::_M_run() + 0x2b (0x7f5c7bdba2f1 in /usr/local/lib/libmodelbox.so.1)
frame #43: + 0xbd6df (0x7f5c79b956df in /usr/lib/x86_64-linux-gnu/libstdc++.so.6)
frame #44: + 0x76db (0x7f5c7ae2b6db in /lib/x86_64-linux-gnu/libpthread.so.0)
frame #45: clone + 0x3f (0x7f5c795f061f in /lib/x86_64-linux-gnu/libc.so.6)

[2025-12-18 08:13:08,714][ERROR][ main.cc:93 ] Segment fault, Signal: 6, Addr: 0x117835, Code: -6, Caused by:
[2025-12-18 08:13:08,714][ERROR][ main.cc:98 ] CPU Register Info:
[R8]=0x0000000000000000 [R9]=0x00007f5b69ff8760 [R10]=0x0000000000000008 [R11]=0x00007f5b446ec220
[R12]=0x00007f5b69ff8b10 [R13]=0x0000000000000004 [R14]=0x00007f5b4409def0 [R15]=0x0000000000000002
[RDI]=0x00007f5b69ff8760 [RSI]=0x00007f5c798bb840 [RBP]=0x00007f5b4465eb80 [RBX]=0x0000000000000000
[RDX]=0x0000000000000000 [RAX]=0x00007f5c7950de87 [RCX]=0x00007f5b69ff8760 [RSP]=0x00007f5c7950de87
[RIP]=0x0000000000000246 [RFLAGS]=0xa82035b8b23b7700
[2025-12-18 08:13:08,727][FATAL][ main.cc:100 ] Stack:
#0 0x00005601c8a99e2f: ?? () from /usr/local/bin/modelbox+0x299e2f
#1 0x00007f5c7ae36980: ?? () from /lib/x86_64-linux-gnu/libpthread.so.0+0x12980
#2 0x00007f5c7950de87: gsignal from /lib/x86_64-linux-gnu/libc.so.6+0x3ee87
#3 0x00007f5c7950f7f1: abort from /lib/x86_64-linux-gnu/libc.so.6+0x407f1
#4 0x00007f5c79b64957: ?? () from /usr/lib/x86_64-linux-gnu/libstdc++.so.6+0x8c957
#5 0x00007f5c79b6aae6: ?? () from /usr/lib/x86_64-linux-gnu/libstdc++.so.6+0x92ae6
#6 0x00007f5c79b69b49: ?? () from /usr/lib/x86_64-linux-gnu/libstdc++.so.6+0x91b49
#7 0x00007f5c79b6a4b8: __gxx_personality_v0 from /usr/lib/x86_64-linux-gnu/libstdc++.so.6+0x924b8
#8 0x00007f5c798d0573: ?? () from /lib/x86_64-linux-gnu/libgcc_s.so.1+0x10573
#9 0x00007f5c798d0df5: _Unwind_Resume from /lib/x86_64-linux-gnu/libgcc_s.so.1+0x10df5
#10 0x00007f5bd5427c8e: ?? () from /usr/local/lib/libtorch_cpu.so+0x3b93c8e
#11 0x00007f5bd5429655: ?? () from /usr/local/lib/libtorch_cpu.so+0x3b95655
#12 0x00007f5bd284e68d: torch::jit::SourceRange::highlight(std::ostream&) const from /usr/local/lib/libtorch_cpu.so+0xfba68d
#13 0x00007f5bd2831781: torch::jit::ErrorReport::what() const from /usr/local/lib/libtorch_cpu.so+0xf9d781
#14 0x00007f5be55c18af: TorchInferenceFlowUnit::LoadModel(std::__cxx11::basic_string<char, std::char_traits, std::allocator > const&, std::shared_ptrmodelbox::Configuration const&) from /usr/local/lib/libmodelbox-unit-cuda-torch_inference.so.1.0.0+0x898af
#15 0x00007f5be55c1be1: TorchInferenceFlowUnit::InitConfig(std::shared_ptrmodelbox::Configuration const&) from /usr/local/lib/libmodelbox-unit-cuda-torch_inference.so.1.0.0+0x89be1
#16 0x00007f5be55c2056: TorchInferenceFlowUnit::Open(std::shared_ptrmodelbox::Configuration const&) from /usr/local/lib/libmodelbox-unit-cuda-torch_inference.so.1.0.0+0x8a056
#17 0x00007f5c7bac0aa4: ?? () from /usr/local/lib/libmodelbox.so.1+0xa7daa4
#18 0x00007f5c7bac614a: ?? () from /usr/local/lib/libmodelbox.so.1+0xa8314a
#19 0x00007f5c7bac60cd: ?? () from /usr/local/lib/libmodelbox.so.1+0xa830cd
#20 0x00007f5c7bac5ef8: ?? () from /usr/local/lib/libmodelbox.so.1+0xa82ef8
#21 0x00007f5c7bac5847: ?? () from /usr/local/lib/libmodelbox.so.1+0xa82847
#22 0x00007f5c7bac542e: ?? () from /usr/local/lib/libmodelbox.so.1+0xa8242e
#23 0x00007f5c7bac506c: ?? () from /usr/local/lib/libmodelbox.so.1+0xa8206c
#24 0x00007f5c7bac4cf5: ?? () from /usr/local/lib/libmodelbox.so.1+0xa81cf5
#25 0x00007f5c7bac5983: ?? () from /usr/local/lib/libmodelbox.so.1+0xa82983
#26 0x00007f5c7bac551d: ?? () from /usr/local/lib/libmodelbox.so.1+0xa8251d
#27 0x00005601c8a1fe1a: std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>::operator()() const from /usr/local/bin/modelbox+0x21fe1a
#28 0x00005601c8a1e8f8: std::__future_base::_State_baseV2::_M_do_set(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool) from /usr/local/bin/modelbox+0x21e8f8
#29 0x00005601c8a23584: void std::__invoke_impl<void, void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::__invoke_memfun_deref, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&) from /usr/local/bin/modelbox+0x223584
#30 0x00005601c8a213ed: std::__invoke_result<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>::type std::__invoke<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&) from /usr/local/bin/modelbox+0x2213ed
#31 0x00005601c8a1f63b: std::call_once<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::once_flag&, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&)::{lambda()#1}::operator()() const from /usr/local/bin/modelbox+0x21f63b
#32 0x00005601c8a1f766: std::call_once<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::once_flag&, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&)::{lambda()#2}::operator()() const from /usr/local/bin/modelbox+0x21f766
#33 0x00005601c8a1f79e: std::call_once<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::once_flag&, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&)::{lambda()#2}::_FUN() from /usr/local/bin/modelbox+0x21f79e
#34 0x00007f5c7ae33907: ?? () from /lib/x86_64-linux-gnu/libpthread.so.0+0xf907
#35 0x00005601c8a13fee: ?? () from /usr/local/bin/modelbox+0x213fee
#36 0x00005601c8a1f892: void std::call_once<void (std::__future_base::_State_baseV2::)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool>(std::once_flag&, void (std::__future_base::_State_baseV2::&&)(std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>, bool*), std::__future_base::_State_baseV2*&&, std::function<std::unique_ptr<std::__future_base::_Result_base, std::__future_base::_Result_base::_Deleter> ()>&&, bool&&) from /usr/local/bin/modelbox+0x21f892
#37 0x00005601c8a1e382: std:...

[pymumu]

模型和pytroch的兼容性问题吧

[JiangXiaoDong123]

模型和pytroch的兼容性问题吧

yolo11可以支持么，现在正在测试yolo11

模型和pytroch的兼容性问题吧

这个是我的环境
PyTorch版本: 2.4.1+cu121
CUDA可用: True
CUDA版本: 12.1
GPU数量: 1

重新编译了工程和训练的模型，现在运行模型会报下面的错误。能帮忙分析下么

[2025-12-19 08:04:36,711][DEBUG][ any.h:185 ] Key reset_flag not found
[2025-12-19 08:04:36,713][DEBUG][ node.cc:735 ] node: normalize, resource state after run, input stream 1, data ctx 1, output stream 1
[2025-12-19 08:04:36,713][DEBUG][flow_scheduler.cc:192 ] run normalize end
[2025-12-19 08:04:36,716][DEBUG][ any.h:185 ] Key reset_flag not found
[2025-12-19 08:04:36,716][DEBUG][ any.h:185 ] Key reset_flag not found
[2025-12-19 08:04:36,717][DEBUG][ any.h:185 ] Key reset_flag not found
[2025-12-19 08:04:36,717][DEBUG][ any.h:185 ] Key reset_flag not found
[2025-12-19 08:04:36,874][ERROR][torch_inference_flowunit.cc:600 ] model inference error, isTuple()INTERNAL ASSERT FAILED at "../aten/src/ATen/core/ivalue_inl.h":1306, please report a bug to PyTorch. Expected Tuple but got String
[2025-12-19 08:04:36,916][ERROR][torch_inference_flowunit.cc:600 ] model inference error, isTuple()INTERNAL ASSERT FAILED at "../aten/src/ATen/core/ivalue_inl.h":1306, please report a bug to PyTorch. Expected Tuple but got String
[2025-12-19 08:04:36,957][ERROR][torch_inference_flowunit.cc:600 ] model inference error, isTuple()INTERNAL ASSERT FAILED at "../aten/src/ATen/core/ivalue_inl.h":1306, please report a bug to PyTorch. Expected Tuple but got String
[2025-12-19 08:04:36,981][ERROR][torch_inference_flowunit.cc:604 ] other inference error, The following operation failed in the TorchScript interpreter.
Traceback of TorchScript (most recent call last):
File "code/torch/ultralytics/nn/tasks.py", line 64, in forward
model23 = self.model
_0 = getattr(model23, "0")
_25 = (_1).forward(act, (_0).forward(act, x, ), )
~~~~~~~~~~~ <--- HERE
_26 = (_3).forward(act, (_2).forward(act, _25, ), )
_27 = (_4).forward(act, _26, )
File "code/torch/ultralytics/nn/modules/conv.py", line 12, in forward
x: Tensor) -> Tensor:
conv = self.conv
_0 = (argument_1).forward((conv).forward(x, ), )
~~~~~~~~~~~~~ <--- HERE
return _0
class Concat(Module):
File "code/torch/torch/nn/modules/conv.py", line 12, in forward
bias = self.bias
weight = self.weight
input = torch._convolution(x, weight, bias, [2, 2], [1, 1], [1, 1], False, [0, 0], 1, False, False, True, True)
~~~~~~~~~~~~~~~~~~ <--- HERE
return input
RuntimeError: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
[2025-12-19 08:04:37,005][ERROR][torch_inference_flowunit.cc:604 ] other inference error, The following operation failed in the TorchScript interpreter.
Traceback of TorchScript (most recent call last):
File "code/torch/ultralytics/nn/tasks.py", line 64, in forward
model23 = self.model
_0 = getattr(model23, "0")
_25 = (_1).forward(act, (_0).forward(act, x, ), )
~~~~~~~~~~~ <--- HERE
_26 = (_3).forward(act, (_2).forward(act, _25, ), )
_27 = (_4).forward(act, _26, )
File "code/torch/ultralytics/nn/modules/conv.py", line 12, in forward
x: Tensor) -> Tensor:
conv = self.conv
_0 = (argument_1).forward((conv).forward(x, ), )
~~~~~~~~~~~~~ <--- HERE
return _0
class Concat(Module):
File "code/torch/torch/nn/modules/conv.py", line 12, in forward
bias = self.bias
weight = self.weight
input = torch._convolution(x, weight, bias, [2, 2], [1, 1], [1, 1], False, [0, 0], 1, False, False, True, True)
~~~~~~~~~~~~~~~~~~ <--- HERE
return input
RuntimeError: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.
[2025-12-19 08:04:37,029][ERROR][torch_inference_flowunit.cc:604 ] other inference error, The following operation failed in the TorchScript interpreter.
Traceback of TorchScript (most recent call last):
File "code/torch/ultralytics/nn/tasks.py", line 64, in forward
model23 = self.model
_0 = getattr(model23, "0")
_25 = (_1).forward(act, (_0).forward(act, x, ), )
~~~~~~~~~~~ <--- HERE
_26 = (_3).forward(act, (_2).forward(act, _25, ), )
_27 = (_4).forward(act, _26, )
File "code/torch/ultralytics/nn/modules/conv.py", line 12, in forward
x: Tensor) -> Tensor:
conv = self.conv
_0 = (argument_1).forward((conv).forward(x, ), )
~~~~~~~~~~~~~ <--- HERE
return _0
class Concat(Module):
File "code/torch/torch/nn/modules/conv.py", line 12, in forward
bias = self.bias
weight = self.weight
input = torch._convolution(x, weight, bias, [2, 2], [1, 1], [1, 1], False, [0, 0], 1, False, False, True, True)
~~~~~~~~~~~~~~~~~~ <--- HERE
return input
RuntimeError: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.

[pymumu]

模型参数校验不过吧，看看模型的配置文件，toml文件里面

[JiangXiaoDong123]

模型参数校验不过吧，看看模型的配置文件，toml文件里面

这是工程toml的2个截图

2025-12-19 15:41:33,581][DEBUG][ node.cc:735 ] node: normalize, resource state after run, input stream 1, data ctx 1, output stream 1
[2025-12-19 15:41:33,581][DEBUG][flow_scheduler.cc:192 ] run normalize end
[2025-12-19 15:41:33,581][DEBUG][flow_scheduler.cc:243 ] begin run node normalize for type: 0
[2025-12-19 15:41:33,581][DEBUG][flow_scheduler.cc:190 ] run normalize begin
[2025-12-19 15:41:33,703][ERROR][torch_inference_flowunit.cc:600 ] model inference error, isTuple()INTERNAL ASSERT FAILED at "../aten/src/ATen/core/ivalue_inl.h":1306, please report a bug to PyTorch. Expected Tuple but got String
[2025-12-19 15:41:33,744][ERROR][torch_inference_flowunit.cc:600 ] model inference error, isTuple()INTERNAL ASSERT FAILED at "../aten/src/ATen/core/ivalue_inl.h":1306, please report a bug to PyTorch. Expected Tuple but got String
[2025-12-19 15:41:33,785][ERROR][torch_inference_flowunit.cc:600 ] model inference error, isTuple()INTERNAL ASSERT FAILED at "../aten/src/ATen/core/ivalue_inl.h":1306, please report a bug to PyTorch. Expected Tuple but got String
[2025-12-19 15:41:33,808][ERROR][torch_inference_flowunit.cc:604 ] other inference error, The following operation failed in the TorchScript interpreter.
Traceback of TorchScript (most recent call last):
File "code/torch/ultralytics/nn/tasks.py", line 64, in forward
model23 = self.model
_0 = getattr(model23, "0")
_25 = (_1).forward(act, (_0).forward(act, x, ), )
~~~~~~~~~~~ <--- HERE
_26 = (_3).forward(act, (_2).forward(act, _25, ), )
_27 = (_4).forward(act, _26, )
File "code/torch/ultralytics/nn/modules/conv.py", line 12, in forward
x: Tensor) -> Tensor:
conv = self.conv
_0 = (argument_1).forward((conv).forward(x, ), )
~~~~~~~~~~~~~ <--- HERE
return _0
class Concat(Module):
File "code/torch/torch/nn/modules/conv.py", line 12, in forward
bias = self.bias
weight = self.weight
input = torch._convolution(x, weight, bias, [2, 2], [1, 1], [1, 1], False, [0, 0], 1, False, False, True, True)
~~~~~~~~~~~~~~~~~~ <--- HERE
return input
RuntimeError: CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call,so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1.

上面是执行时的错误信息

这个是torch的d调用例子。

这是对应处理功能单元的源码

看了源码部分确实应该是设置的参数格式有问题，但是现在不确定要如何设置，可以给下指导么？我在网上找的yolo11的调用方式，看样子是需要传递多个值，但是现在看配置只给了一个输入值，要如何增加和传递设置呢呢？谢谢

[JiangXiaoDong123]

模型参数校验不过吧，看看模型的配置文件，toml文件里面

现在想进行调试，但是使用vscode始终运行不到上面的报错位置，现在这个库是动态加载的吧，理论上应该可以打断点吧

[pymumu]

可以的，你可以编译debug版本。
不行可以加点log