I am using Keras in Google Colab. I am doing transfer learning: trying different models/weights available in keras.applications. I am not developing a new model, just importing layers (except last one) and weights, and adding one final layer.
I need to get reproducible results: every time I run the same code I get the same results. I have tried different options (links below), but I cannot get reproducible results.
Please, any suggestions or code for this?
How to get reproducible results in keras
Tensorflow-Keras reproducibility problem on Google Colab
Tensorflow-Keras reproducibility problem on Google Colab
Thanks