Unable to run on coral edge tpu

Hi all, I am trying to run on coral tpu over RPC server, I am able to rum on the cpu, bur when I switch to target = “edge_tpu” (mobilenet_v2_1.0_224_quant_edgetpu.tflite) I get the next error:

on host:

Traceback (most recent call last): File “/home/user/projects/tvm/testcompile/main.py”, line 19, in runtime = tflite_runtime.create(f.read(), ctx, runtime_target=target) File “/home/user/tvm/python/tvm/contrib/tflite_runtime.py”, line 49, in create return TFLiteModule(fcreate(bytearray(tflite_model_bytes), ctx)) File “/home/user/tvm/python/tvm/_ffi/_ctypes/packed_func.py”, line 225, in call raise get_last_ffi_error() tvm._ffi.base.TVMError: Traceback (most recent call last): [bt] (5) /home/user/tvm/build/libtvm.so(TVMFuncCall+0x65) [0x7f80efeafc85] [bt] (4) /home/user/tvm/build/libtvm.so(std::_Function_handler<void (tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*), tvm::runtime::RPCModuleNode::WrapRemoteFunc(void*)::{lambda(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*)#1}>::_M_invoke(std::_Any_data const&, tvm::runtime::TVMArgs&&, tvm::runtime::TVMRetValue*&&)+0x33) [0x7f80eff2d273] [bt] (3) /home/user/tvm/build/libtvm.so(tvm::runtime::RPCWrappedFunc::operator()(tvm::runtime::TVMArgs, tvm::runtime::TVMRetValue*) const+0x3c5) [0x7f80eff2ccd5] [bt] (2) /home/user/tvm/build/libtvm.so(tvm::runtime::RPCClientSession::CallFunc(void*, TVMValue const*, int const*, int, std::function<void (tvm::runtime::TVMArgs)> const&)+0x57) [0x7f80eff1d937] [bt] (1) /home/user/tvm/build/libtvm.so(tvm::runtime::RPCEndpoint::CallFunc(void*, TVMValue const*, int const*, int, std::function<void (tvm::runtime::TVMArgs)>)+0x2a3) [0x7f80eff15293] [bt] (0) /home/user/tvm/build/libtvm.so(+0x10f9232) [0x7f80eff12232] File “/home/user/tvm/src/runtime/rpc/rpc_endpoint.cc”, line 799 TVMError: Check failed: code == RPCCode: :kReturn: code=1

on the RP4:

INFO:RPCServer:connection from (‘’, 41442) INFO:RPCServer:Finish serving (‘’, 41442) INFO:RPCServer:connection from (‘’, 41526) INFO:RPCServer:Finish serving (‘’, 41526) INFO:RPCServer:connection from (‘’, 41530) ERROR: Internal: Unsupported data type in custom op handler: 756172288 ERROR: Node number 0 (edgetpu-custom-op) failed to prepare.

I am using edge tpu version 14 and tensorflow/tensorflow@d855adf

as was said here https://github.com/google-coral/edgetpu/issues/137

does anyone know how to solve that problem? thanks

1 Like

hi, i got a same issue when trying to run inference on coral tpu. have you solved the problem? i tried to run inference on the board directly without TVM, it works fine.

1 Like