Hugging Face's logo

NAMAA-Space
/
GATE-Reranker-V1

Text Classification
Transformers
Safetensors
sentence-transformers
Arabic
bert
reranking
Inference Endpoints
Model card Files Community
1
omarelshehy commited on
Commit
dfd15f6
·
verified ·
1 Parent(s): 05d3718

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +6 -5
README.md CHANGED
@@ -49,11 +49,12 @@ scores = model.predict([(Query, Paragraph1), (Query, Paragraph2)])
49
 
50
  We evaluate our model on two different datasets using the metrics **MAP**, **MRR** and **NDCG@10**:
51
 
52
- The purpose of this evaluation is to highlight the performance of our model with regards to: Relevant/Irrelevant labels and
53
 
54
- 1 - Dataset 1: [NAMAA-Space/Arabic-Reranking-Triplet-5-Eval](https://huggingface.co/datasets/NAMAA-Space/Arabic-Reranking-Triplet-5-Eval)
55
- 2 - Dataset 2: [NAMAA-Space/Ar-Reranking-Eval](https://huggingface.co/datasets/NAMAA-Space/Ar-Reranking-Eval)
56
 
57
- and compare it to other famous models on the hub
58
 
59
- ![Plot](https://huggingface.co/NAMAA-Space/GATE-Reranker-V1/resolve/main/Dataset1_Evaluation.jpg)
 
 
 
49
 
50
  We evaluate our model on two different datasets using the metrics **MAP**, **MRR** and **NDCG@10**:
51
 
52
+ The purpose of this evaluation is to highlight the performance of our model with regards to: Relevant/Irrelevant labels and positive/multiple negatives documents:
53
 
54
+ Dataset 1: [NAMAA-Space/Ar-Reranking-Eval](https://huggingface.co/datasets/NAMAA-Space/Ar-Reranking-Eval)
 
55
 
56
+ ![Plot](https://huggingface.co/NAMAA-Space/GATE-Reranker-V1/resolve/main/Dataset1_Evaluation.jpg)
57
 
58
+ Dataset 2: [NAMAA-Space/Arabic-Reranking-Triplet-5-Eval](https://huggingface.co/datasets/NAMAA-Space/Arabic-Reranking-Triplet-5-Eval)
59
+
60
+ and compare it to other famous models on the hub