Issue saving and converting the Gemma 2 model after training

[sparsh35]

Describe the bug I am getting this error to convert the model 'Gemma2' with SFT trainer to pytorch after training , Lm.head weights are missing

To Reproduce use this script just replace the training model with Gemma 2 9b pt , Link to kaggle code of SFT example

[sparsh35]

This is saving screenshot of saving of the file

[sparsh35]

Got it , i didn't read complete docs , for those who may have this problem correct code is here , Code to SFT trainer with logic to save with tie word embeddings