Hello Christiopher!
The difference is I used the ATL method to fine tune GPT2-medium and improve it’s responses drastically. Using deep learning, I added another layer of improvement to GPT2′s parameters in order to simulate an alignment solution.
Thanks!
Hello Christiopher!
The difference is I used the ATL method to fine tune GPT2-medium and improve it’s responses drastically. Using deep learning, I added another layer of improvement to GPT2′s parameters in order to simulate an alignment solution.
Thanks!