whether train_initialization.py and train_fine_tune.py use the same training data?

I am a little confuse.
If they use the same training data, In fine_tune phase just train the last few layer.
I think this operating is a little redundant with train_initialization.py

Looking forward to your guidance.

Start by reading train_initialization.py and train_fine_tune.py, then trace the data-loading paths and layer-freezing configuration in each script. Compare which datasets and training stages they use; the issue is resolved when the relationship between the two phases and any redundancy is clearly documented.

Written by the indexing model from the issue text.