GeoGPT-Research-Project commited on
Commit
21dc66c
·
verified ·
1 Parent(s): f1adcb4

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +6 -4
README.md CHANGED
@@ -1,8 +1,8 @@
1
- # Model Card for GeoReranker
2
 
3
- The reranker 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.
4
 
5
- ## Quick Start
6
 
7
  To load the GeoReranker model with HuggingFace, use the following snippet:
8
  ```python
@@ -21,9 +21,11 @@ scores = reranker.compute_score([['what is panda?', 'hi'], ['what is panda?', 'T
21
 
22
 
23
  ## License and Uses
24
- GeoReranker is licensed under the MIT License Agreement. The primary use of GeoGPT models is to support geoscience research, providing geoscientists with innovative tools and capabilities enhanced by large language models. It is specifically designed for non-commercial research and educational purposes.
25
 
26
  The model is not intended for use in any manner that violates applicable laws or regulations, nor for any activities prohibited by the license agreement. Additionally, it should not be used in languages other than those explicitly supported, as outlined in this model card.
27
 
 
 
28
  ## Limitations
29
  GeoReranker is trained on English datasets, and performance may be suboptimal for other languages.
 
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.
4
 
5
+ ## Quickstart
6
 
7
  To load the GeoReranker model with HuggingFace, use the following snippet:
8
  ```python
 
21
 
22
 
23
  ## License and Uses
24
+ GeoReranker is licensed under the [MIT License](https://github.com/GeoGPT-Research-Project/GeoGPT-RAG/blob/master/MIT_LICENSE). GeoReranker is trained on the foundation of [BGE-M3](https://huggingface.co/BAAI/bge-m3), which is licensed under the [MIT License](https://github.com/FlagOpen/FlagEmbedding?tab=MIT-1-ov-file). It is your responsibility to ensure that your use of GeoReranker adheres to the terms of both the GeoReranker model and its upstream dependency, [BGE-M3](https://huggingface.co/BAAI/bge-m3).
25
 
26
  The model is not intended for use in any manner that violates applicable laws or regulations, nor for any activities prohibited by the license agreement. Additionally, it should not be used in languages other than those explicitly supported, as outlined in this model card.
27
 
28
+
29
+
30
  ## Limitations
31
  GeoReranker is trained on English datasets, and performance may be suboptimal for other languages.