Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
# GeoReranker
|
| 2 |
|
| 3 |
The GeoReranker model is a critical component in Retrieval-Augmented Generation (RAG) systems, designed to refine the initial retrieval results by reordering candidate documents based on their semantic relevance to the query. Different from embedding model, reranker uses question and document as input and directly output similarity instead of embedding. You can get a relevance score by inputting query and passage to the reranker. And the score can be mapped to a float value in [0,1] by sigmoid function.
|
|
@@ -29,4 +35,4 @@ The model is not intended for use in any manner that violates applicable laws or
|
|
| 29 |
|
| 30 |
|
| 31 |
## Limitations
|
| 32 |
-
GeoReranker is trained on English datasets, and performance may be suboptimal for other languages.
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
base_model:
|
| 5 |
+
- BAAI/bge-m3
|
| 6 |
+
---
|
| 7 |
# GeoReranker
|
| 8 |
|
| 9 |
The GeoReranker model is a critical component in Retrieval-Augmented Generation (RAG) systems, designed to refine the initial retrieval results by reordering candidate documents based on their semantic relevance to the query. Different from embedding model, reranker uses question and document as input and directly output similarity instead of embedding. You can get a relevance score by inputting query and passage to the reranker. And the score can be mapped to a float value in [0,1] by sigmoid function.
|
|
|
|
| 35 |
|
| 36 |
|
| 37 |
## Limitations
|
| 38 |
+
GeoReranker is trained on English datasets, and performance may be suboptimal for other languages.
|