Cannot assign a device for operation M1 Macbook pro Monterey 12.1 Tried to run Image segmentation from tensorflow tutorial

Question

Created Jan ’22

Replies 6

Boosts 2

Views 2.0k

Participants 6

InvalidArgumentError: Cannot assign a device for operation augment/random_flip/stateful_uniform_full_int/RngReadAndSkip: Could not satisfy explicit device specification '' because the node {{colocation_node augment/random_flip/stateful_uniform_full_int/RngReadAndSkip}} was colocated with a group of nodes that required incompatible device '/job:localhost/replica:0/task:0/device:GPU:0'. All available devices [/job:localhost/replica:0/task:0/device:CPU:0, /job:localhost/replica:0/task:0/device:GPU:0]. Colocation Debug Info: Colocation group had the following types and supported devices: Root Member(assigned_device_name_index_=2 requested_device_name_='/job:localhost/replica:0/task:0/device:GPU:0' assigned_device_name_='/job:localhost/replica:0/task:0/device:GPU:0' resource_device_name_='/job:localhost/replica:0/task:0/device:GPU:0' supported_device_types_=[CPU] possible_devices_=[] RngReadAndSkip: CPU _Arg: GPU CPU

Colocation members, user-requested devices, and framework assigned devices, if any: augment_random_flip_stateful_uniform_full_int_rngreadandskip_resource (_Arg) framework assigned device=/job:localhost/replica:0/task:0/device:GPU:0 augment/random_flip/stateful_uniform_full_int/RngReadAndSkip (RngReadAndSkip)

 [[{{node augment/random_flip/stateful_uniform_full_int/RngReadAndSkip}}]] [Op:MakeIterator]

Boost

Answer 1

Frameworks Engineer OP

Apple

Jan ’22

Hi @RahulKaranam

Thanks for reporting this issue. The RngReadAndSkip op implementation for the GPU is currently missing from Tensorflow-metal. I will update here once we have the op available. In the meanwhile unfortunately the random_flip augmentation can only be used when running on the CPU.

0

Answer 2

mdrame OP

Mar ’22

Any update on this? I'm facing the same issue

1

Answer 3

LukasM OP

May ’22

Furthermore RandomRotation seems to have the same issue

1

Answer 4

pndy19 OP

Jul ’22

Has there been any update on this ? What is point of using M1 if we are not able to use GPU to utilise in high complex tasks. I am not able to use many functionalities of Tensorflow for which I bought M1.

1

Answer 5

mjacques OP

Jul ’22

Any updates on this? Quite frustrating not being able to use M1 Pro to train my models.

0

Answer 6

Frameworks Engineer OP

Apple

Aug ’22

RngReadAndSkip is registered on the GPU in tensorflow-metal==0.5.1.

2