Comments (6)
You are right. We still don't yet implement a solution to just upsert the delta
changes, everything will be re-upsert again. This is the area we need to improve. We can implement a logic to compare the new embeddings to the cached embeddings, and upsert the new ones. Since we already have embeddings caching:
from flowise.
Thank you very much for the honest response,
Do you have any estimation for the fix you were proposing?
By the way I read over pinecone's documentation they do support upsert, why not use their API I think they should handle the logic of upset once it exists and not flowise
https://docs.pinecone.io/reference/upsert
Thanks
from flowise.
we are using the upsert API under the hood via langchain abstraction. however the ids are different each time you upsert, that causes duplication.
dont have any estimation, but its one of the higher priority item we are taking a look at
from flowise.
Thanks for your honesty @HenryHengZJ
One last question please
Untill you'll fix this issue, if I'll reinsert the same vectors over and over again, is that something which may effect the performance or the accuracy of my bot?
Thanks
from flowise.
seems like it is a very hard task to achieve, I recommend not using upsert, but rather do delete all documents and re-insert these. @HenryHengZJ
from flowise.
soupsert
cannot be used? im also noticing duplicates in my vector store. @HenryHengZJ your solution to use redis did not work
from flowise.
Related Issues (20)
- [question] socket.io sticky session / multiple nodes HOT 1
- Override Config not working
- [BUG?] Weaviate index name is converted to a name beginning with a capital letter HOT 1
- [BUG] Postgres AddDocumentStore migration appears to be broken HOT 2
- [BUG] Wrong CMD in Dockerfile making container fail to start HOT 2
- [BUG] CSV agent is not working with OpenAI ChatGPT HOT 2
- [BUG] CSV document loader marks text splitter as optional, but it fails if it's not specified HOT 1
- 404 by In-Memory Vectore Store
- Upsert button diseaper HOT 2
- [FEATURE] Support proxy for Azure OpenAI Chat model
- Error: chatflowsService.getAllChatflows - Error: getRunningExpressApp failed! [BUG] HOT 5
- Rate Limit Bug HOT 2
- [BUG] 1.7.1 Using Document Store with metadata corrupts chat flow HOT 9
- Setting UI to Dark Mode HOT 1
- [BUG] ConnectionError: Failed to connect to xxx.xxx.xxx.net:1433 - self-signed certificate
- [FEATURE] Support for PostgreSQL as a Vector Store in LlamaIndex
- Unable to load open AI models in my docker container [BUG]
- [BUG] overrideConfig for parameter in 2 nodes not working as documented HOT 1
- Set source name (via Metadata) from data when using document loaders
- Mandatory Fields Error using Apify via Document Store HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from flowise.