Hugging Face's logo

Datasets:
neuclir
/
bench

Tasks:
Text Retrieval
Text Ranking
Modalities:
Tabular
Text
Sub-tasks:
document-retrieval
Languages:
English
Chinese
Persian
Size:
100K - 1M
ArXiv:
License:
Dataset card Data Studio Files
xet
Community
1
eugene-yang commited on
Commit
6ffdcec
·
verified ·
1 Parent(s): c00d0f6

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +86 -1
README.md CHANGED
@@ -66,4 +66,89 @@ configs:
66
  dtype: int
67
  ---
68
 
69
- # NeuCLIR News Topics
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
66
  dtype: int
67
  ---
68
 
69
+ # NeuCLIRBench Topics and Queries
70
+
71
+ NeuCLIRBench is an evaluation benchmark for monolingual, cross-language, and multilingual adhoc retrieval.
72
+
73
+ The document collection can be found at [neuclir/neuclir1](https://huggingface.co/datasets/neuclir/neuclir1).
74
+
75
+ ## Supporting Tasks and Corresponding Data
76
+
77
+ NeuCLIRBench supports three types of tasks: monolingual, cross-language, and multilingual adhoc retrieval.
78
+ The following specifies the documents, queries, and qrels (labels) that should be used for each task.
79
+
80
+ Please report nDCG@20 for all tasks.
81
+
82
+ *We use `:` to indicate different subset under the dataset.*
83
+
84
+ ### Monolingual Retrieval (`mono`)
85
+
86
+ | Language | Documents | Queries | Qrels |
87
+ | --- | --- | --- | --- |
88
+ | English | All splits under `neuclir/neuclir1:mt_docs` | `eng` split of `neuclir/bench-queries:queries` | `mlir` split of `neuclir/bench-queries:qrels` |
89
+ | Persian | `fas` split of `neuclir/neuclir1:default` | `fas` split of `neuclir/bench-queries:queries` | `fas` split of `neuclir/bench-queries:qrels` |
90
+ | Russian | `rus` split of `neuclir/neuclir1:default` | `rus` split of `neuclir/bench-queries:queries` | `rus` split of `neuclir/bench-queries:qrels` |
91
+ | Chinese | `zho` split of `neuclir/neuclir1:default` | `zho` split of `neuclir/bench-queries:queries` | `zho` split of `neuclir/bench-queries:qrels` |
92
+
93
+
94
+ ### Cross-Language Retrieval (`clir`)
95
+
96
+ | Language | Documents | Queries | Qrels |
97
+ | --- | --- | --- | --- |
98
+ | Persian | `fas` split of `neuclir/neuclir1:default` | `eng` split of `neuclir/bench-queries:queries` | `fas` split of `neuclir/bench-queries:qrels` |
99
+ | Russian | `rus` split of `neuclir/neuclir1:default` | `eng` split of `neuclir/bench-queries:queries` | `rus` split of `neuclir/bench-queries:qrels` |
100
+ | Chinese | `zho` split of `neuclir/neuclir1:default` | `eng` split of `neuclir/bench-queries:queries` | `zho` split of `neuclir/bench-queries:qrels` |
101
+
102
+
103
+ ### Multilingual Retrieval (`mlir`)
104
+
105
+ | Language | Documents | Queries | Qrels |
106
+ | --- | --- | --- | --- |
107
+ | English | All splits under `neuclir/neuclir1:default` | `eng` split of `neuclir/bench-queries:queries` | `mlir` split of `neuclir/bench-queries:qrels` |
108
+
109
+ ## Baseline Retrieval Results and Run Files
110
+
111
+ We also provide all reported baseline retrieval results in the NeuCLIRBench paper.
112
+ Please refer to the paper for the detailed descriptions of each model.
113
+
114
+ ![Monolingual Results](https://cdn-uploads.huggingface.co/production/uploads/63a0c07a3c8841cfe2cd1e70/EdfQAcTgwsC4P4lk5lQTh.png)
115
+
116
+ ![Cross-Language and Multilngual Results](https://cdn-uploads.huggingface.co/production/uploads/63a0c07a3c8841cfe2cd1e70/CCs5PCzBgiPsjZsfG1iM6.png)
117
+
118
+ ### Run Names
119
+
120
+ Please refer to the `runs` subset in this dataset to find all the runs.
121
+ Files follow the naming scheme of `{run_handle}_{task:mono/clir/mlir}_{lang}.trec`. Please refer to the task section for the details.
122
+
123
+ | Run Handle | Model Type | Model Name |
124
+ |:------------------|:-------------------|:----------------------|
125
+ | bm25 | Lexical | BM25 |
126
+ | bm25dt | Lexical | BM25 w/ DT |
127
+ | bm25qt | Lexical | BM25 w/ QT |
128
+ | iclr | Bi-Encoder | MILCO |
129
+ | plaidx | Bi-Encoder | PLAID-X |
130
+ | qwen8b | Bi-Encoder | Qwen3 8B Embed |
131
+ | qwen4b | Bi-Encoder | Qwen3 4B Embed |
132
+ | qwen600m | Bi-Encoder | Qwen3 0.6B Embed |
133
+ | arctic | Bi-Encoder | Arctic-Embed Large v2 |
134
+ | splade | Bi-Encoder | SPLADEv3 |
135
+ | fusion3 | Bi-Encoder | Fusion |
136
+ | repllama | Bi-Encoder | RepLlama |
137
+ | me5large | Bi-Encoder | e5 Large |
138
+ | jinav3 | Bi-Encoder | JinaV3 |
139
+ | bgem3sparse | Bi-Encoder | BGE-M3 Sparse |
140
+ | mt5 | Pointwise Reranker | Mono-mT5XXL |
141
+ | qwen3-0.6b-rerank | Pointwise Reranker | Qwen3 0.6B Rerank |
142
+ | qwen3-4b-rerank | Pointwise Reranker | Qwen3 4B Rerank |
143
+ | qwen3-8b-rerank | Pointwise Reranker | Qwen3 8B Rerank |
144
+ | jina-rerank | Pointwise Reranker | Jina Reranker |
145
+ | searcher-rerank | Pointwise Reranker | SEARCHER Reranker |
146
+ | rank1 | Pointwise Reranker | Rank1 |
147
+ | qwq | Listwise Reranker | Rank-K (QwQ) |
148
+ | rankzephyr | Listwise Reranker | RankZephyr 7B |
149
+ | firstqwen | Listwise Reranker | FIRST Qwen3 8B |
150
+ | rq32b | Listwise Reranker | RankQwen-32B |
151
+
152
+
153
+ ## Citation
154
+ TBA