Systematic way to freeze parameters for fine-tuning.

I want to fine-tune the model on my own dataset. But I need to perform training on specific layers.

I want to know if there is an attribute we can change or add in the yaml configuration file, to freeze the parameters of different layers?

Start by locating the YAML configuration used for fine-tuning and the code that constructs model parameters for training. Determine how selected layers should be specified and verify that those parameters remain frozen while other layers train.

Written by the indexing model from the issue text.