Custom object detection, training with tflite-mobdel maker via Google Colab

TensorFlow General Discussion
,
1

Hi everyone!

I am in a school project right now, and I try to detect my custom object models according this tutorial: > https://www.youtube.com/watch?v=-ZyFYniGUsw

Google Colab link: Google Colab

Everything wents fine until “Step 2: Train a custom model with TensorFlow Lite Model Maker”.

I also use Google Colab as the tutorial mentions, but there are some huge error messages during the installation ( !pip install -q tflite-model-maker and !pip install -q tflite-support):

error: cannot install tflite-model-maker==0.1.2, tflite-model-maker==0.2.0, tflite-model-maker==0.2.1, tflite-model-maker==0.2.2, tflite-model-maker==0.2.3, tflite-model-maker==0.2.4, tflite-model-maker==0.2.5, tflite-model-maker==0.3.3, tflite-model-maker==0.3.4, tflite-model-maker==0.4.0, tflite-model-maker==0.4.1 and tflite-model-maker==0.4.2 because these package versions have conflicting dependencies.

Can you please help me out? There are couple of deprecated solutions (installing another model-maker, or use fallback at Google Colab /which is also not available/)

Thanks in advance

3

Hi @s4m4n

Welcome to the TensorFlow Forum!

There is a known issue of tflite model maker installation if you are using python >=3.10. Please use Python 3.9 or Mediapipe Model Maker as a workaround. Thank you.

4

Check a python notebook that I have created recently to see how you can use pytho 3.9 inside Colab environment. You can find it at the below blog post.