It’s for an online training website.
1- What is the best way to fine tune data used for online training?
2- Is it possible to update the data used for fine tuning or do I have to fine tune everything? How?
3- Is it possible to train the AI with let’s say 100 questions and add additional set later?
4- Let’s say I have DataSet1, DataSet2,… If I train my AI with different DataSets and want to replace/update one DataSet, do I have to train it again with all DataSets?
5- What is the best model for my use case? curie, davinci,…?