TF 2.15 object detection

When im trying to use research object_detections models it says AttributeError: module ‘tensorflow.python.ops.control_flow_ops’ has no attribute ‘case’. I found out that downgrading to tf 2.13 helps with that, but then it does see my gpu. I tried to install cuda manually and via package manager, nothing helps. nvcc as well says that i should install cuda toolkit, while its already installed. So my question is if it is possible to make object detection model via tensorflow 2.15?

1 Like

Hi @raaw ,

As per my understanding from above, yes it’s possible to make object detection models with tensorflow version 2.15.*. Before that you are facing issues with your environment setup, please follow these tensorflow setup and compatible versions for your hardware requirements.I think if you follow the instructions in the documentation then you will be able to train the models without any setup issues.Hope this helps!

There are couple of Tutorials on training and deploying machine learning models in the web browser.


Same issue. I get training to work only with TF2.13 but then it does not recognised the grapgic card in colab