-
### Description
The system shall allow the user to present the sparse vector representation of the query to the system and perform an efficient and effective search of the inverted index to identify …
-
In my opinion, sparse vectors can solve two different problems:
1. Text search (TFIDF, BM25, SPLADE, etc.)
2. Weighted-keywords search
In both cases, we can have a sparse vector representation:…
-
### Description
The system shall build an inverted index on the sparse vector embedding that are loaded by the user into the system, in an efficient manner.
### Acceptance criteria
1. The time re…
-
-
### Description
Learned Sparse Vectors claim to combine the benefits of sparse (i.e. lexical) and dense (i.e. vector) representations
From https://en.wikipedia.org/wiki/Learned_sparse_retrieval:…
-
First of all, thank you for putting time and effort in this project, and maybe this feature request is not at all the way you want to go with this project.
But as I'm planning to transition my pro…
-
### Question Validation
- [X] I have searched both the documentation and discord for an answer.
### Question
I'm running Qdrant on hybrid mode:
```
vector_store = QdrantVectorStore(
collec…
-
## Current Behavior
I've implemented hybrid search with filters and I sometimes get high-scoring results that do not closely match my query. My collections have both sparse and dense vectors an…
-
### Description
The system shall allow the user to assign weights to the results from the sparse vector search and dense vector search, fuse them based on the weighted score of the two search methods…
-
**Describe the bug**
Some cases of appropriate sparse matrix broadcasts error on CUDA sparse matrices
**To reproduce**
The Minimal Working Example (MWE) for this bug:
```julia
using CUDA
…