Dear Developers,
I am a newbie on the TensorFlow business. I have however managed to install TensorFlow (macOS and metal) on M1 MacBook Air as detailed in this website (https://developer.apple.com/metal/tensorflow-plugin/)
However, when I try to run an example code (https://github.com/keras-team/keras-io/blob/master/examples/vision/mnist_convnet.py), I got issues that I cannot understand
`2021-08-01 00:28:43.807922: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:176] None of the MLIR Optimization Passes are enabled (registered 2)
2021-08-01 00:28:43.808097: W tensorflow/core/platform/profile_utils/cpu_utils.cc:128] Failed to get CPU frequency: 0 Hz
Epoch 1/15
2021-08-01 00:28:43.911127: I tensorflow/core/grappler/optimizers/custom_graph_optimizer_registry.cc:112] Plugin optimizer for device_type GPU is enabled.
2021-08-01 00:28:43.984 python[20105:3940931] -[MPSNDArrayIdentity encodeToCommandEncoder:commandBuffer:sourceArrays:resultState:destinationArray:kernelDAGObject:]: unrecognized selector sent to instance 0x1682e8560
2021-08-01 00:28:43.987 python[20105:3940931] *** Terminating app due to uncaught exception 'NSInvalidArgumentException', reason: '-[MPSNDArrayIdentity encodeToCommandEncoder:commandBuffer:sourceArrays:resultState:destinationArray:kernelDAGObject:]: unrecognized selector sent to instance 0x1682e8560'
*** First throw call stack:
(
0 CoreFoundation 0x0000000180de2320 __exceptionPreprocess + 240
1 libobjc.A.dylib 0x0000000180b10c04 objc_exception_throw + 60
2 CoreFoundation 0x0000000180e71020 -[NSObject(NSObject) __retain_OA] + 0
3 CoreFoundation 0x0000000180d44184 ___forwarding___ + 1444
4 CoreFoundation 0x0000000180d43b30 _CF_forwarding_prep_0 + 96
5 libmetal_plugin.dylib 0x0000000150a55db4 ___ZN12metal_plugin14MPSApplyAdamOpIfE7ComputeEPNS_15OpKernelContextE_block_invoke.115 + 92
6 libdispatch.dylib 0x0000000180abd420 _dispatch_client_callout + 20
7 libdispatch.dylib 0x0000000180acba98 _dispatch_lane_barrier_sync_invoke_and_complete + 60
8 libmetal_plugin.dylib 0x0000000150a533c0 _ZN12metal_plugin14MPSApplyAdamOpIfE7ComputeEPNS_15OpKernelContextE + 4884
9 libmetal_plugin.dylib 0x0000000150a51e78 _ZN12metal_pluginL15ComputeOpKernelINS_14MPSApplyAdamOpIfEEEEvPvP18TF_OpKernelContext + 44
10 _pywrap_tensorflow_internal.so 0x0000000130076460 _ZN10tensorflow15PluggableDevice7ComputeEPNS_8OpKernelEPNS_15OpKernelContextE + 148
11 libtensorflow_framework.2.dylib 0x0000000123a78664 _ZN10tensorflow12_GLOBAL__N_113ExecutorStateINS_15PropagatorStateEE7ProcessENS2_10TaggedNodeEx + 3080
12 libtensorflow_framework.2.dylib 0x0000000123a79dbc _ZNSt3__110__function6__funcIZN10tensorflow12_GLOBAL__N_113ExecutorStateINS2_15PropagatorStateEE7RunTaskIZNS6_13ScheduleReadyEPN4absl14lts_2020_09_2313InlinedVectorINS5_10TaggedNodeELm8ENS_9allocatorISB_EEEEPNS5_20TaggedNodeReadyQueueEEUlvE0_EEvOT_EUlvE_NSC_ISL_EEFvvEEclEv + 56
13 _pywrap_tensorflow_internal.so 0x00000001307aa230 _ZN5Eigen15ThreadPoolTemplIN10tensorflow6thread16EigenEnvironmentEE10WorkerLoopEi + 1508
14 _pywrap_tensorflow_internal.so 0x00000001307a9b28 _ZZN10tensorflow6thread16EigenEnvironment12CreateThreadENSt3__18functionIFvvEEEENKUlvE_clEv + 80
15 libtensorflow_framework.2.dylib 0x0000000124014b08 _ZN10tensorflow12_GLOBAL__N_17PThread8ThreadFnEPv + 120
16 libsystem_pthread.dylib 0x0000000180c6a06c _pthread_start + 320
17 libsystem_pthread.dylib 0x0000000180c64da0 thread_start + 8
)
libc++abi.dylib: terminating with uncaught exception of type NSException`
Do you have an idea what could be the reason? I would like also to benefit from GPU acceleration. I am running on macOS Big Sur 11.2.3.
Many thanks, All the best
