Log in
Log into community
Find answers from the community
View all posts
Related posts
Did this answer your question?
π
π
π
Powered by
Hall
Inactive
Updated 2 years ago
0
Follow
load multiple embeddings files
load multiple embeddings files
Inactive
0
Follow
d
dev_blockchain
2 years ago
Β·
Hey guys is there a way we can load multiple embeddings files in one go, let's say my
index_file
is a kind of array or list or a directory.
Plain Text
Copy
index = GPTSimpleVectorIndex.load_from_disk(index_file, service_context)
d
L
5 comments
Share
Open in Discord
d
dev_blockchain
2 years ago
any help or reference @Logan M
L
Logan M
2 years ago
Sorry, load from disk is only made for one file at a time π€
d
dev_blockchain
2 years ago
So, do we have any other method to resolve this problem ?
L
Logan M
2 years ago
I think there's a few option
create all the embeddings in one file to start with (this is usually the easiest option tbh)
use each index in a composable index, maybe with a top level index (list, tree, etc,)
use each index with the new router abstraction (requires upgrading llama index, which will break a few things and change how some functions work)
create a custom retriever (again, requires updating)
There's maybe some other hacky ways, but tbh I would just recreate the index if possible
d
dev_blockchain
2 years ago
let me check this one
Add a reply
Sign up and join the conversation on Discord
Join on Discord