Chat with RTX no longer re-trains with new data

I was able to train and use the model very well until yesterday when I added one more file to my dataset and retrained it. It got stuck at “Generating new values”, I tried to remove that last file I added and only try with old ones, deleting the vector embeddings and also trying with a different directory but none worked. It doesn’t continue from this part:

[03/18/2024-12:05:39] No sentence-transformers model found with name C:\Users\User/.cache\torch\sentence_transformers\WhereIsAI_UAE-Large-V1. Creating a new one with MEAN pooling.
Generating new values

RTX 4070 mobile and latest game ready drivers (551.76), Windows 11 Pro