Runtime error for alexnet.

[arvindhbti]

@prasshantg i am facing the same issue when running with NN_L0_0_fbuf.

NvDlaSubmit: Error IOCTL failed (No such process)
(DLA_RUNTIME) Error 0x0003000f: (propagating from Runtime.cpp, function submitInternal(), line 666)
(DLA_TEST) Error 0x00000004: runtime->submit() failed (in RuntimeTest.cpp, function runTest(), line 387)
(DLA_TEST) Error 0x00000004: (propagating from RuntimeTest.cpp, function run(), line 440)

But when i am running my alexnet i got fast-math.nvdla and when i am running my input cat.pgm it is giving error like
./nvdla_runtime --loadable fast-math.nvdla --image cat.pgm

creating new runtime context...
Emulator starting
pgm2dimg 1 360 480 1 7264 1648928 16489280
(DLA_TEST) Error 0x00000004: Mismatched width: 480 != 227 (in TestUtils.cpp, function createImageCopy(), line 156)
(DLA_TEST) Error 0x00000004: (propagating from RuntimeTest.cpp, function copyImageToInputTensor(), line 100)
(DLA_TEST) Error 0x00000004: (propagating from RuntimeTest.cpp, function setupInputBuffer(), line 162)
(DLA_TEST) Error 0x00000004: (propagating from RuntimeTest.cpp, function runTest(), line 382)
(DLA_TEST) Error 0x00000004: (propagating from RuntimeTest.cpp, function run(), line 440)
Shutdown signal received, exiting
(DLA_TEST) Error 0x00000004: (propagating from main.cpp, function launchTest(), line 87)

Please @prasshantg help me out.Thanks

[prasshantg]

@arvindhbti you build alexnet for 227x227 input but using image of 480x360 you need to use same input image as input dimensions set in network

[arvindhbti]

Hi @prasshantg thanks for reply.

can you please explain me why i am getting this error in docker vp and local system.

creating new wisdom context...
opening wisdom context...
parsing caffe network...
[libprotobuf FATAL /usr/local/include/google/protobuf/repeated_field.h:886] CHECK failed: (index) < (size()):
terminate called after throwing an instance of 'google::protobuf::FatalException'
what(): CHECK failed: (index) < (size()):
Aborted (core dumped)

and @prasshantg please explain me how to solve this problem.Thanks.

[arvindhbti]

@arvindhbti you build alexnet for 227x227 input but using image of 480x360 you need to use same input image as input dimensions set in network

@prasshantg when i am using image with 480*360 then i am getting this error.

./nvdla_runtime --loadable fast-math.nvdla --image cat.pgm

creating new runtime context...
Emulator starting
(DLA_TEST) Error 0x0000000b: Unexpected PGM value: P6 (in DlaImageUtils.cpp, function parsePGMInfo(), line 122)
(DLA_TEST) Error 0x0000000b: (propagating from DlaImageUtils.cpp, function PGM2DIMG(), line 67)
(DLA_TEST) Error 0x0000000b: (propagating from RuntimeTest.cpp, function copyImageToInputTensor(), line 84)
(DLA_TEST) Error 0x0000000b: (propagating from RuntimeTest.cpp, function setupInputBuffer(), line 162)
(DLA_TEST) Error 0x0000000b: (propagating from RuntimeTest.cpp, function runTest(), line 382)
(DLA_TEST) Error 0x0000000b: (propagating from RuntimeTest.cpp, function run(), line 440)
Shutdown signal received, exiting
(DLA_TEST) Error 0x0000000b: (propagating from main.cpp, function launchTest(), line 87)

@prasshantg Thanks.

[arvindhbti]

Hi @prasshantg that AlexNet issue solved , it was asking 227*227 dim gray image.in last giving test pass also.

But I want to know about lenet model is not compiling as you can see about the error.

Please @prasshantg help me out in lenet model.Thanks.

[prasshantg]

@arvindhbti what error you get for lenet model?

[arvindhbti]

Hi @prasshantg thanks for reply.

This is the error.

creating new wisdom context...
opening wisdom context...
parsing caffe network...
[libprotobuf FATAL /usr/local/include/google/protobuf/repeated_field.h:886] CHECK failed: (index) < (size()):
terminate called after throwing an instance of 'google::protobuf::FatalException'
what(): CHECK failed: (index) < (size()):
Aborted (core dumped)

same error i got in docker vp and local system.

Please let me know the reason.
@prasshantg Thanks.

[arvindhbti]

hi @prasshantg find below lenet.prototxt and caffemodel.

caffemodel - lenet_iter_10000.caffemodel.

prototxt - deploy.prototxt

[Jian-PENG207]

Hi,I want to know which prototxt and caffemodel file of alexnet you used ? I have met a "unhandle level 2 translation fault " question

[arvindhbti]

Hi @prasshantg thanks for reply.

This is the error.

creating new wisdom context...
opening wisdom context...
parsing caffe network...
[libprotobuf FATAL /usr/local/include/google/protobuf/repeated_field.h:886] CHECK failed: (index) < (size()):
terminate called after throwing an instance of 'google::protobuf::FatalException'
what(): CHECK failed: (index) < (size()):
Aborted (core dumped)

same error i got in docker vp and local system.

Please let me know the reason.
@prasshantg Thanks.

Hi @prasshantg can you please explain the reason.