Kevin Hu
Kevin Hu
[May this help](https://stackoverflow.com/questions/22567306/how-to-upload-file-with-python-requests)
I guess that LlamaParse calls remote services, so it's not literally local parsing.
docker compose down -v (remove all the local data) docker compose up
What about removing the version numbers?
Is it locally deployed or could you share a sample of this text file? It's weird 4K text file will generate 3K chunks. I guess it stucked with ES indexing....
When the setup dialog is gona be used by API, it can deliver variables to fullfill the prompt, like /api/completion?gender=male. The prompt is like "You're a xxx. When the gender...
All the chat history will be transfered to LLM. I guess LLM thought the first answer was not good enough so came the same quesion again, and it's also confused...
Theoretically, you're not gona loss any data by upgrading system. In term of data security, just backup the docker volumes.
After code modification, run ""npm run build"" to make it work.
Max token of RAPTOR means the summary of a cluster should be in that maxium length. 1024 of chunk size is a reasonable choice.