KeyError: 'TensorArrayV3' while import_graph_def
algoking opened this issue · 4 comments
my environment is python 3.5 ubuntu 16.04, AMD RX480
It's work fine with basic calculation in GPU, such as:
with tf.device('/gpu:0'):
a = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[2, 3], name='a')
b = tf.constant([1.0, 2.0, 3.0, 4.0, 5.0, 6.0], shape=[3, 2], name='b')
c = tf.matmul(a, b)
but when I use object detection,
with detection_graph.as_default():
od_graph_def = tf.GraphDef()
with tf.gfile.GFile(PATH_TO_CKPT, 'r') as fid:
serialized_graph = fid.read()
od_graph_def.ParseFromString(serialized_graph)
tf.import_graph_def(od_graph_def, name='')
exception by KeyError: 'TensorArrayV3', I found a issue in tensorflow related to "TensorArrayV3",
but there is no any solution on it. Some guy said that tensorflow write in py2.7, when migrate to
py3, there will have lots of issue need to fix, such as annoying encode, decode stuff.
But my AMD only can work by your open-cl version tensorflow, and it's py3, is it possible to
compile tf-coriander to py2.7? or any work around? Great thanks, I stuck with AMD OpenCl Tensorflow in past few weeks.
Below it's my clinfo:
Number of platforms 1
Platform Name AMD Accelerated Parallel Processing
Platform Vendor Advanced Micro Devices, Inc.
Platform Version OpenCL 2.0 AMD-APP (2442.7)
Platform Profile FULL_PROFILE
Platform Extensions cl_khr_icd cl_amd_event_callback cl_amd_offline_devices
Platform Extensions function suffix AMD
Platform Name AMD Accelerated Parallel Processing
Number of devices 1
Device Name Ellesmere
Device Vendor Advanced Micro Devices, Inc.
Device Vendor ID 0x1002
Device Version OpenCL 1.2 AMD-APP (2442.7)
Driver Version 2442.7
Device OpenCL C Version OpenCL C 1.2
Device Type GPU
Device Profile FULL_PROFILE
Device Board Name (AMD) AMD Radeon (TM) RX 480 Graphics
Device Topology (AMD) PCI-E, 01:00.0
Max compute units 36
SIMD per compute unit (AMD) 4
SIMD width (AMD) 16
SIMD instruction width (AMD) 1
Max clock frequency 1291MHz
Graphics IP (AMD) 8.0
Device Partition (core)
Max number of sub-devices 36
Supported partition types none specified
Max work item dimensions 3
Max work item sizes 256x256x256
Max work group size 256
Preferred work group size multiple 64
Wavefront width (AMD) 64
Preferred / native vector sizes
char 4 / 4
short 2 / 2
int 1 / 1
long 1 / 1
half 1 / 1 (cl_khr_fp16)
float 1 / 1
double 1 / 1 (cl_khr_fp64)
Half-precision Floating-point support (cl_khr_fp16)
Denormals No
Infinity and NANs No
Round to nearest No
Round to zero No
Round to infinity No
IEEE754-2008 fused multiply-add No
Support is emulated in software No
Correctly-rounded divide and sqrt operations No
Single-precision Floating-point support (core)
Denormals No
Infinity and NANs Yes
Round to nearest Yes
Round to zero Yes
Round to infinity Yes
IEEE754-2008 fused multiply-add Yes
Support is emulated in software No
Correctly-rounded divide and sqrt operations Yes
Double-precision Floating-point support (cl_khr_fp64)
Denormals Yes
Infinity and NANs Yes
Round to nearest Yes
Round to zero Yes
Round to infinity Yes
IEEE754-2008 fused multiply-add Yes
Support is emulated in software No
Correctly-rounded divide and sqrt operations No
Address bits 64, Little-Endian
Global memory size 3392765952 (3.16GiB)
Global free memory (AMD) 3291040 (3.139GiB)
Global memory channels (AMD) 8
Global memory banks per channel (AMD) 16
Global memory bank width (AMD) 256 bytes
Error Correction support No
Max memory allocation 2415919104 (2.25GiB)
Unified memory for Host and Device No
Minimum alignment for any data type 128 bytes
Alignment of base address 2048 bits (256 bytes)
Global Memory cache type Read/Write
Global Memory cache size 16384
Global Memory cache line 64 bytes
Image support Yes
Max number of samplers per kernel 16
Max size for 1D images from buffer 134217728 pixels
Max 1D or 2D image array size 2048 images
Base address alignment for 2D image buffers 256 bytes
Pitch alignment for 2D image buffers 256 bytes
Max 2D image size 16384x16384 pixels
Max 3D image size 2048x2048x2048 pixels
Max number of read image args 128
Max number of write image args 8
Local memory type Local
Local memory size 32768 (32KiB)
Local memory syze per CU (AMD) 65536 (64KiB)
Local memory banks (AMD) 32
Max constant buffer size 2415919104 (2.25GiB)
Max number of constant args 8
Max size of kernel argument 1024
Queue properties
Out-of-order execution No
Profiling Yes
Prefer user sync for interop Yes
Profiling timer resolution 1ns
Profiling timer offset since Epoch (AMD) 1505395864920699653ns (Thu Sep 14 21:31:04 2017)
Execution capabilities
Run OpenCL kernels Yes
Run native kernels No
Thread trace supported (AMD) Yes
SPIR versions 1.2
printf() buffer size 1048576 (1024KiB)
Built-in kernels
Device Available Yes
Compiler Available Yes
Linker Available Yes
Device Extensions cl_khr_fp64 cl_amd_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_fp16 cl_khr_gl_sharing cl_amd_device_attribute_query cl_amd_vec3 cl_amd_printf cl_amd_media_ops cl_amd_media_ops2 cl_amd_popcnt cl_khr_image2d_from_buffer cl_khr_spir cl_khr_gl_event
NULL platform behavior
clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...) No platform
clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...) No platform
clCreateContext(NULL, ...) [default] No platform
clCreateContext(NULL, ...) [other] Success [AMD]
clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU) No platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU) No platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR) No platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM) No platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL) No platform
I think tf-coriander should be buildable for python 2.7. You will need to replace python3 with python2 in the scripts inside util directory, something like:
cd util
sed -i -e 's/python3/python2/g' *
I'm not sure how far this will get you. Let me know how this goes please, pasting the full output logs to a https://gist.github.com , when you hit the first issue/error.
fix bunch of error when build from the source, then stuck here
~/tf-coriander ~/tf-coriander
No Google Cloud Platform support will be enabled for TensorFlow
No Hadoop File System support will be enabled for TensorFlow
Found possible Python library paths:
/usr/local/lib/python2.7/dist-packages
/usr/lib/python2.7/dist-packages
Please input the desired Python library path to use. Default is [/usr/local/lib/python2.7/dist-packages]
/usr/local/lib/python2.7/dist-packages
No GPU support will be enabled for TensorFlow
checking operating system
Configuration finished
INFO: Starting clean (this may take a while). Consider using --expunge_async if the clean takes more than several minutes.
........
____Loading package: tensorflow/tools/pip_package
____Downloading http://github.com/google/protobuf/archive/v3.1.0.tar.gz via codeload.github.com: 13,700 bytes
____Downloading http://github.com/google/protobuf/archive/v3.1.0.tar.gz via codeload.github.com: 40,240 bytes
____Downloading http://github.com/google/protobuf/archive/v3.1.0.tar.gz via codeload.github.com: 96,186 bytes
____Downloading http://github.com/google/protobuf/archive/v3.1.0.tar.gz via codeload.github.com: 206,932 bytes
____Downloading http://github.com/google/protobuf/archive/v3.1.0.tar.gz via codeload.github.com: 331,378 bytes
____Downloading http://github.com/google/protobuf/archive/v3.1.0.tar.gz via codeload.github.com: 659,002 bytes
____Downloading http://github.com/google/protobuf/archive/v3.1.0.tar.gz via codeload.github.com: 943,010 bytes
____Downloading http://github.com/google/protobuf/archive/v3.1.0.tar.gz via codeload.github.com: 1,665,388 bytes
____Downloading http://github.com/google/protobuf/archive/v3.1.0.tar.gz via codeload.github.com: 2,847,776 bytes
INFO: All external dependencies fetched successfully.
+ util/build_tf.sh
++ uname
+ [[ Linux == Darwin ]]
+ bash util/build_tf_u1604.sh
+ echo BUILDTYPE
BUILDTYPE
+ bazel build @grpc//:grpc_cpp_plugin
INFO: Found 1 target...
ERROR: /home/davinci/.cache/bazel/_bazel_davinci/aaa0efdc1b9af1206a63d0304b90b478/external/protobuf/BUILD:73:1: undeclared inclusion(s) in rule '@protobuf//:protobuf_lite':
this rule is missing dependency declarations for the following files included by 'external/protobuf/src/google/protobuf/stubs/atomicops_internals_x86_msvc.cc':
'/usr/lib/gcc/x86_64-linux-gnu/4.9/include/stdarg.h'
'/usr/lib/gcc/x86_64-linux-gnu/4.9/include/stddef.h'
'/usr/lib/gcc/x86_64-linux-gnu/4.9/include/stdint.h'
'/usr/lib/gcc/x86_64-linux-gnu/4.9/include-fixed/limits.h'
'/usr/lib/gcc/x86_64-linux-gnu/4.9/include-fixed/syslimits.h'.
Target @grpc//:grpc_cpp_plugin failed to build
Use --verbose_failures to see the command lines of failed build steps.
INFO: Elapsed time: 2.092s, Critical Path: 0.36s
@algoking and @devkicks which version of tensorflow did you build? Google's object detection API requires a minimum of tensorflow 1.5.
Has the tf-corrainder project been updated for more recent builds (1.5+)?