Save or backup huge pdf embeddings #1726
Replies: 3 comments 1 reply
-
Yea, I was wondering if there is a way to push ingested files to a long term memory on the hard drive that can be queried and loaded into vRAM in a pre-ingested state. |
Beta Was this translation helpful? Give feedback.
-
If you restart your computer and the privateGPT system, it should behave as if the listed files are ingested. I just tested this explicitly with a couple of short files describing both of my dogs, and having the chat make up a cute story about them. It worked as I expected. I've also observed that ingested files are resilient across restarts previously, but wanted to make double sure for you. I'll make a video about this soon! My system is in a Hyper-V virtual machine and doesn't have a GPU, but I'd expect a system with a GPU would reload these at startup. |
Beta Was this translation helpful? Give feedback.
-
The vector is by default Qdrant as filebased. So persistent, unless if you use Docker without volume. |
Beta Was this translation helpful? Give feedback.
-
Hello
Is there any way to save or backup embeddings or data from already embedded huge pdf file and re use/ load again when required?
Beta Was this translation helpful? Give feedback.
All reactions