Unable to load model using tensorflow hub.KerasLayer

TensorFlow General Discussion
1

I am getting this error, when I try to install the bert models:

Code:
bert_preprocess = hub.KerasLayer(‘TensorFlow | bert | Kaggle’)
bert_encoder = hub.KerasLayer(‘TensorFlow | bert | Kaggle’)

InvalidArgumentError: Multiple OpKernel registrations match NodeDef at the same priority ‘{{node AssignVariableOp}}’: ‘op: “AssignVariableOp” device_type: “GPU” constraint { name: “dtype” allowed_values { list { type: DT_INT64 } } } host_memory_arg: “resource”’ and ‘op: “AssignVariableOp” device_type: “GPU” constraint { name: “dtype” allowed_values { list { type: DT_INT64 } } } host_memory_arg: “resource”’
[[AssignVariableOp]] [Op:AssignVariableOp]

2

Hi @Abdul_Rahmaan, Before importing the bert model from tf hub you have to import tensorflow_text as text to registers the ops

import tensorflow_text as text
import tensorflow_hub as hub

preprocessor = hub.KerasLayer("https://tfhub.dev/tensorflow/bert_en_uncased_preprocess/3")

. Thank You.

1 Like
4

Yes, I have done that but still getting the same error.

5

HI Abdul,

I’ve just tested exactly the same code here on colab:

!pip install tensorflow_text

import tensorflow_text as text
import tensorflow_hub as hub

preprocessor = hub.KerasLayer("https://tfhub.dev/tensorflow/bert_en_uncased_preprocess/3")
bert_encoder = hub.KerasLayer("https://tfhub.dev/tensorflow/bert_en_uncased_L-12_H-768_A-12/4")

and everything works fine.
Maybe there’s something on the env you are running it?

6

Hey, I am not sure if there is env issue as I have created a new env. I am adding the logs on anaconda prompt. Hopefully that helps.

To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2023-08-18 19:31:57.154641: I tensorflow/c/logging.cc:34] DirectML: creating device on adapter 0 (NVIDIA GeForce GTX 1650)
2023-08-18 19:31:57.336691: I tensorflow/c/logging.cc:34] Successfully opened dynamic library Kernel32.dll
2023-08-18 19:31:57.340141: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:306] Could not identify NUMA node of platform GPU ID 0, defaulting to 0. Your kernel may not have been built with NUMA support.
2023-08-18 19:31:57.340511: W tensorflow/core/common_runtime/pluggable_device/pluggable_device_bfc_allocator.cc:28] Overriding allow_growth setting because force_memory_growth was requested by the device.
2023-08-18 19:31:57.341053: I tensorflow/core/common_runtime/pluggable_device/pluggable_device_factory.cc:272] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 10190 MB memory) → physical PluggableDevice (device: 0, name: DML, pci bus id: )
2023-08-18 19:31:59.144319: E tensorflow/core/grappler/clusters/utils.cc:87] Failed to get device properties, error code: 302
[I 2023-08-18 19:32:57.550 ServerApp] Saving file at /NNCodeBasics/Untitled1.ipynb

8

Hi @Abdul_Rahmaan, Could you please share the details of the environment like OS, Tensorflow version you are using, the platform which you are trying to execute the code, etc… Thank You.

10

Hey, I was able to solve the issue by deleting the environment and creating a new one.
Thank you.