Fine-tuning the model
#4
by
danmanor
- opened
Hey,
Thank you for your work this model is very impressive. I am quite new in this field, and I want to try fine-tuning this model on different subjects to test its understanding capabilities. I understand the training method vary between different models/tasks (depend on the architecture and task we want to optimize results for), but for the task of understanding text documents and being able to answer questions on, how would you fine-tune this model ?
Thank you very much,
Dan
Hi Dan and Every one else,
I run into the same problem. How do you fine tune the model? And how many examples do you need? I have almost 30K Q&A. How do I fine tune the model?
Dan - did you manage?
Best,
Eli