Spaces:
Build error
Similar tool, different dataset
Hi, I'm pretty new here, so please forgive any silly question. I come from the business side and see a great potential in something like this using other datasets. How could I create a chatbot exactly like this, but with different dataset? For example, a dataset that is composed of the full text of 4 books from the same author (considering I have the rights to use that text).
Hi, The basic answer if you're familiar with git and python involves using GPT_Index and LangChain. If you provide more info I could be more specific however. Another approach if you're interested in working with books is this respository Dr Doc search.
Thanks for you reply @EmbeddedAndrew !
The idea is not limited to books. I gave that example to make it easy to understand. In fact, it would be mostly about online courses (videos classes and written material). I envision implementations that could assist online courses producers to create new content and offer streamlined support to students. (Therefore, the Dr Doc approach is not the way forward)
As I said, I come from business and humanities backgroud. No real experience with git and python. However, I'm sure I can learn a bit and find partners that have that technical knowledge, to make it work. Huge market and business opportunities here.
Your advice was already super helpful. Thanks again!
Hi, Since this is not a general interest discussion, feel free to email me at recursive@altmails.com.