Hugging Face's logo

Datasets:
hakari-bench
/
NanoR2MED

Modalities:
Text
Formats:
parquet
Languages:
English
Size:
10K - 100K
Tags:
information-retrieval
retrieval
nano
bm25
dense-retrieval
reranking
Libraries:
Datasets
pandas
Polars
Dataset card Data Studio Files
xet
 Community
hotchpotch commited on
Commit
d477980
·
verified ·
1 Parent(s): e4b2a56

Copy dataset from hotchpotch/NanoR2MED

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. NanoR2MEDBioinformatics/corpus/test.parquet +3 -0
  2. NanoR2MEDBioinformatics/metadata/test.json +46 -0
  3. NanoR2MEDBioinformatics/qrels/test.parquet +3 -0
  4. NanoR2MEDBioinformatics/queries/test.parquet +3 -0
  5. NanoR2MEDBiology/corpus/test.parquet +3 -0
  6. NanoR2MEDBiology/metadata/test.json +46 -0
  7. NanoR2MEDBiology/qrels/test.parquet +3 -0
  8. NanoR2MEDBiology/queries/test.parquet +3 -0
  9. NanoR2MEDIIYiClinical/corpus/test.parquet +3 -0
  10. NanoR2MEDIIYiClinical/metadata/test.json +46 -0
  11. NanoR2MEDIIYiClinical/qrels/test.parquet +3 -0
  12. NanoR2MEDIIYiClinical/queries/test.parquet +3 -0
  13. NanoR2MEDMedQADiag/corpus/test.parquet +3 -0
  14. NanoR2MEDMedQADiag/metadata/test.json +46 -0
  15. NanoR2MEDMedQADiag/qrels/test.parquet +3 -0
  16. NanoR2MEDMedQADiag/queries/test.parquet +3 -0
  17. NanoR2MEDMedXpertQAExam/corpus/test.parquet +3 -0
  18. NanoR2MEDMedXpertQAExam/metadata/test.json +46 -0
  19. NanoR2MEDMedXpertQAExam/qrels/test.parquet +3 -0
  20. NanoR2MEDMedXpertQAExam/queries/test.parquet +3 -0
  21. NanoR2MEDMedicalSciences/corpus/test.parquet +3 -0
  22. NanoR2MEDMedicalSciences/metadata/test.json +46 -0
  23. NanoR2MEDMedicalSciences/qrels/test.parquet +3 -0
  24. NanoR2MEDMedicalSciences/queries/test.parquet +3 -0
  25. NanoR2MEDPMCClinical/corpus/test.parquet +3 -0
  26. NanoR2MEDPMCClinical/metadata/test.json +46 -0
  27. NanoR2MEDPMCClinical/qrels/test.parquet +3 -0
  28. NanoR2MEDPMCClinical/queries/test.parquet +3 -0
  29. NanoR2MEDPMCTreatment/corpus/test.parquet +3 -0
  30. NanoR2MEDPMCTreatment/metadata/test.json +46 -0
  31. NanoR2MEDPMCTreatment/qrels/test.parquet +3 -0
  32. NanoR2MEDPMCTreatment/queries/test.parquet +3 -0
  33. README.md +174 -0
  34. bm25/NanoR2MEDBioinformatics.parquet +3 -0
  35. bm25/NanoR2MEDBiology.parquet +3 -0
  36. bm25/NanoR2MEDIIYiClinical.parquet +3 -0
  37. bm25/NanoR2MEDMedQADiag.parquet +3 -0
  38. bm25/NanoR2MEDMedXpertQAExam.parquet +3 -0
  39. bm25/NanoR2MEDMedicalSciences.parquet +3 -0
  40. bm25/NanoR2MEDPMCClinical.parquet +3 -0
  41. bm25/NanoR2MEDPMCTreatment.parquet +3 -0
  42. manifest.json +378 -0
  43. metadata/NanoR2MEDBioinformatics.json +46 -0
  44. metadata/NanoR2MEDBiology.json +46 -0
  45. metadata/NanoR2MEDIIYiClinical.json +46 -0
  46. metadata/NanoR2MEDMedQADiag.json +46 -0
  47. metadata/NanoR2MEDMedXpertQAExam.json +46 -0
  48. metadata/NanoR2MEDMedicalSciences.json +46 -0
  49. metadata/NanoR2MEDPMCClinical.json +46 -0
  50. metadata/NanoR2MEDPMCTreatment.json +46 -0
NanoR2MEDBioinformatics/corpus/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:66cbd907d354aea52bd4eab8bf93b0dc43e855013ff5d30c9d2cef3992877ce6
3
+ size 4003275
NanoR2MEDBioinformatics/metadata/test.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDBioinformatics",
3
+ "source_task": "R2MEDBioinformaticsRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/Bioinformatics",
7
+ "revision": "6021fce366892cbfd7837fa85a4128ea93315e18"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 77,
15
+ "corpus": 10000,
16
+ "qrels": 226,
17
+ "bm25_rows": 77,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 226,
28
+ "hits": 226,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.17857929009122858,
32
+ "ndcg_at_100": 0.35060738065375957,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 77,
36
+ "source_corpus_count": 47471,
37
+ "source_qrels_query_count": 77,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 6,
42
+ "qrels_rewritten_for_duplicate_text": 1,
43
+ "forced_queries": 47,
44
+ "forced_doc_count": 111,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
NanoR2MEDBioinformatics/qrels/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34adc31b633617898b95bb7d6b00a3817844047032b616d2851debad291e2a0c
3
+ size 4525
NanoR2MEDBioinformatics/queries/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c55a4b87bd4844601e7f4fa75ff7e13dd56284ce00bbbf8b4b73a526e2d0cf0
3
+ size 43229
NanoR2MEDBiology/corpus/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fdd35ea7f9b11dc15bd59e3fca1e9f9c2191fe85c53ae016d5a11fd7dee64faf
3
+ size 2803375
NanoR2MEDBiology/metadata/test.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDBiology",
3
+ "source_task": "R2MEDBiologyRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/Biology",
7
+ "revision": "8b9fec2db9eda4b5742d03732213fbaee8169556"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 103,
15
+ "corpus": 10000,
16
+ "qrels": 374,
17
+ "bm25_rows": 103,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 374,
28
+ "hits": 374,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.25128449065032643,
32
+ "ndcg_at_100": 0.4296835025091202,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 103,
36
+ "source_corpus_count": 57359,
37
+ "source_qrels_query_count": 103,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 1058,
42
+ "qrels_rewritten_for_duplicate_text": 9,
43
+ "forced_queries": 68,
44
+ "forced_doc_count": 167,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
NanoR2MEDBiology/qrels/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bcbc6fccf59451bb0faac43f844ed6eb1b57442ad247f8c0f668701c2c4c550d
3
+ size 7955
NanoR2MEDBiology/queries/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edb56f0ab211e1a4ac7b730c5ae738a55642273aea23197201fc95959ba3bb9c
3
+ size 37895
NanoR2MEDIIYiClinical/corpus/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04a14ca43061ef14f687e7fdf5f2d707f0f3909a5db837a48f09afcce7261b72
3
+ size 23769652
NanoR2MEDIIYiClinical/metadata/test.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDIIYiClinical",
3
+ "source_task": "R2MEDIIYiClinicalRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/IIYi-Clinical",
7
+ "revision": "974abbc9bc281c3169180a6aa5d7586cfd2f5877"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 129,
15
+ "corpus": 10000,
16
+ "qrels": 457,
17
+ "bm25_rows": 129,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 457,
28
+ "hits": 457,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.12460354977311297,
32
+ "ndcg_at_100": 0.33378196865460924,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 129,
36
+ "source_corpus_count": 10449,
37
+ "source_qrels_query_count": 129,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 0,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 113,
44
+ "forced_doc_count": 268,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
NanoR2MEDIIYiClinical/qrels/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4555d23602a158d788d57a34042266eeb4e55f5691607da0ec8f6b42ea4506b
3
+ size 5067
NanoR2MEDIIYiClinical/queries/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b80100d0ff50038645dfa432d8ab155c323cb2e2b34349e8d1956a5cdcb7b4b6
3
+ size 162080
NanoR2MEDMedQADiag/corpus/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65b946c3c648f256441629d3d07378a12ddeb1e8edf079d856a70d6e852b3bfd
3
+ size 5034064
NanoR2MEDMedQADiag/metadata/test.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDMedQADiag",
3
+ "source_task": "R2MEDMedQADiagRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/MedQA-Diag",
7
+ "revision": "78b585990279cc01a493f876c1b0cf09557fba57"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 118,
15
+ "corpus": 10000,
16
+ "qrels": 522,
17
+ "bm25_rows": 118,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 522,
28
+ "hits": 522,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.028104212132602936,
32
+ "ndcg_at_100": 0.2642743613844765,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 118,
36
+ "source_corpus_count": 56194,
37
+ "source_qrels_query_count": 118,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 6,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 109,
44
+ "forced_doc_count": 440,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
NanoR2MEDMedQADiag/qrels/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fea87dc86c668c656ffed195d4715c626b36eea8cfa6179f7ca4353ab00fc07c
3
+ size 7024
NanoR2MEDMedQADiag/queries/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:041e38a1fc0c1b4f1c28b2e08d581b4ad0a06ff40605869f79c4242e90d3c0c0
3
+ size 50503
NanoR2MEDMedXpertQAExam/corpus/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb0464aa006f75360cfac5fc8718284adb2d76a25b7d8d66a5c7e750a8517cb8
3
+ size 3983178
NanoR2MEDMedXpertQAExam/metadata/test.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDMedXpertQAExam",
3
+ "source_task": "R2MEDMedXpertQAExamRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/MedXpertQA-Exam",
7
+ "revision": "b457ea43db9ae5db74c3a3e5be0a213d0f85ac3a"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 97,
15
+ "corpus": 10000,
16
+ "qrels": 292,
17
+ "bm25_rows": 97,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 292,
28
+ "hits": 292,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.024462598222211914,
32
+ "ndcg_at_100": 0.2304796798768815,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 97,
36
+ "source_corpus_count": 61379,
37
+ "source_qrels_query_count": 97,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 0,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 90,
44
+ "forced_doc_count": 255,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
NanoR2MEDMedXpertQAExam/qrels/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8f58a934770ef049f8b6d77bff2f8c81284b4498dc5d133a7457d5424bfecbe
3
+ size 5056
NanoR2MEDMedXpertQAExam/queries/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5c211f08c29a509e77a95da974b27d05f0cf6dde3b7b226a4871b379e021a46c
3
+ size 55643
NanoR2MEDMedicalSciences/corpus/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:663524f25b90d466de174e4b475cef62e46c975be63b1ecd676808894c4ef696
3
+ size 4450523
NanoR2MEDMedicalSciences/metadata/test.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDMedicalSciences",
3
+ "source_task": "R2MEDMedicalSciencesRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/Medical-Sciences",
7
+ "revision": "1b48911514c80bf9182222d99752ad75e23b4b47"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 88,
15
+ "corpus": 10000,
16
+ "qrels": 244,
17
+ "bm25_rows": 88,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 244,
28
+ "hits": 244,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.10428646688120163,
32
+ "ndcg_at_100": 0.29337732163029967,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 88,
36
+ "source_corpus_count": 34792,
37
+ "source_qrels_query_count": 88,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 66,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 62,
44
+ "forced_doc_count": 138,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
NanoR2MEDMedicalSciences/qrels/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2f99064aceee7932ac35e302f971bf912cbb5e076f6bacadf5da19e01999e1e2
3
+ size 5727
NanoR2MEDMedicalSciences/queries/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77519074c0c5d238cfa33bde72d483ae159fc6668c43737da6013c8964e84beb
3
+ size 29066
NanoR2MEDPMCClinical/corpus/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a77852ff8dc01e3625c7e92aa3fba4ce15a8d5971d466bbde2d6a43736dac84e
3
+ size 12284775
NanoR2MEDPMCClinical/metadata/test.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDPMCClinical",
3
+ "source_task": "R2MEDPMCClinicalRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/PMC-Clinical",
7
+ "revision": "812829522f7eaa407ef82b96717be85788a50f7e"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 114,
15
+ "corpus": 10000,
16
+ "qrels": 248,
17
+ "bm25_rows": 114,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 248,
28
+ "hits": 248,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.3277008274619314,
32
+ "ndcg_at_100": 0.4621547795535059,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 114,
36
+ "source_corpus_count": 60406,
37
+ "source_qrels_query_count": 114,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 0,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 46,
44
+ "forced_doc_count": 63,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
NanoR2MEDPMCClinical/qrels/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:692a0677367c3efca9c31e3d06b2104c1289d5662fe273160d2ac59576eff4a8
3
+ size 4840
NanoR2MEDPMCClinical/queries/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dd47c2619058e22902e20b2912e8d1f8d220b929a9323c388312ce33e70a7cae
3
+ size 59255
NanoR2MEDPMCTreatment/corpus/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dc88079036bbb43b6dac1a6c097556973d73a744770ad565db00b62ff41f14ec
3
+ size 4116589
NanoR2MEDPMCTreatment/metadata/test.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDPMCTreatment",
3
+ "source_task": "R2MEDPMCTreatmentRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/PMC-Treatment",
7
+ "revision": "53c489a44a3664ba352c07550b72b4525a5968d5"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 150,
15
+ "corpus": 10000,
16
+ "qrels": 315,
17
+ "bm25_rows": 150,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 315,
28
+ "hits": 315,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.01798988334744109,
32
+ "ndcg_at_100": 0.2202813875703761,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 150,
36
+ "source_corpus_count": 28809,
37
+ "source_qrels_query_count": 150,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 6,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 106,
44
+ "forced_doc_count": 186,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
NanoR2MEDPMCTreatment/qrels/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9599792893fd20fd02273853d8b9e9fbd16b91dc3829918cc8b1f75a8c3ecd83
3
+ size 5587
NanoR2MEDPMCTreatment/queries/test.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23f397f3e0ccc2646e4bcbab6c22f0274822c84ad59dfc6765c01922b7eba040
3
+ size 139877
README.md ADDED
@@ -0,0 +1,174 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ configs:
3
+ - config_name: bm25
4
+ data_files:
5
+ - split: NanoR2MEDBiology
6
+ path: bm25/NanoR2MEDBiology.parquet
7
+ - split: NanoR2MEDBioinformatics
8
+ path: bm25/NanoR2MEDBioinformatics.parquet
9
+ - split: NanoR2MEDMedicalSciences
10
+ path: bm25/NanoR2MEDMedicalSciences.parquet
11
+ - split: NanoR2MEDMedXpertQAExam
12
+ path: bm25/NanoR2MEDMedXpertQAExam.parquet
13
+ - split: NanoR2MEDMedQADiag
14
+ path: bm25/NanoR2MEDMedQADiag.parquet
15
+ - split: NanoR2MEDPMCTreatment
16
+ path: bm25/NanoR2MEDPMCTreatment.parquet
17
+ - split: NanoR2MEDPMCClinical
18
+ path: bm25/NanoR2MEDPMCClinical.parquet
19
+ - split: NanoR2MEDIIYiClinical
20
+ path: bm25/NanoR2MEDIIYiClinical.parquet
21
+ - config_name: corpus
22
+ data_files:
23
+ - split: NanoR2MEDBiology
24
+ path: NanoR2MEDBiology/corpus/test.parquet
25
+ - split: NanoR2MEDBioinformatics
26
+ path: NanoR2MEDBioinformatics/corpus/test.parquet
27
+ - split: NanoR2MEDMedicalSciences
28
+ path: NanoR2MEDMedicalSciences/corpus/test.parquet
29
+ - split: NanoR2MEDMedXpertQAExam
30
+ path: NanoR2MEDMedXpertQAExam/corpus/test.parquet
31
+ - split: NanoR2MEDMedQADiag
32
+ path: NanoR2MEDMedQADiag/corpus/test.parquet
33
+ - split: NanoR2MEDPMCTreatment
34
+ path: NanoR2MEDPMCTreatment/corpus/test.parquet
35
+ - split: NanoR2MEDPMCClinical
36
+ path: NanoR2MEDPMCClinical/corpus/test.parquet
37
+ - split: NanoR2MEDIIYiClinical
38
+ path: NanoR2MEDIIYiClinical/corpus/test.parquet
39
+ - config_name: qrels
40
+ data_files:
41
+ - split: NanoR2MEDBiology
42
+ path: NanoR2MEDBiology/qrels/test.parquet
43
+ - split: NanoR2MEDBioinformatics
44
+ path: NanoR2MEDBioinformatics/qrels/test.parquet
45
+ - split: NanoR2MEDMedicalSciences
46
+ path: NanoR2MEDMedicalSciences/qrels/test.parquet
47
+ - split: NanoR2MEDMedXpertQAExam
48
+ path: NanoR2MEDMedXpertQAExam/qrels/test.parquet
49
+ - split: NanoR2MEDMedQADiag
50
+ path: NanoR2MEDMedQADiag/qrels/test.parquet
51
+ - split: NanoR2MEDPMCTreatment
52
+ path: NanoR2MEDPMCTreatment/qrels/test.parquet
53
+ - split: NanoR2MEDPMCClinical
54
+ path: NanoR2MEDPMCClinical/qrels/test.parquet
55
+ - split: NanoR2MEDIIYiClinical
56
+ path: NanoR2MEDIIYiClinical/qrels/test.parquet
57
+ - config_name: queries
58
+ data_files:
59
+ - split: NanoR2MEDBiology
60
+ path: NanoR2MEDBiology/queries/test.parquet
61
+ - split: NanoR2MEDBioinformatics
62
+ path: NanoR2MEDBioinformatics/queries/test.parquet
63
+ - split: NanoR2MEDMedicalSciences
64
+ path: NanoR2MEDMedicalSciences/queries/test.parquet
65
+ - split: NanoR2MEDMedXpertQAExam
66
+ path: NanoR2MEDMedXpertQAExam/queries/test.parquet
67
+ - split: NanoR2MEDMedQADiag
68
+ path: NanoR2MEDMedQADiag/queries/test.parquet
69
+ - split: NanoR2MEDPMCTreatment
70
+ path: NanoR2MEDPMCTreatment/queries/test.parquet
71
+ - split: NanoR2MEDPMCClinical
72
+ path: NanoR2MEDPMCClinical/queries/test.parquet
73
+ - split: NanoR2MEDIIYiClinical
74
+ path: NanoR2MEDIIYiClinical/queries/test.parquet
75
+ default: true
76
+ language:
77
+ - en
78
+ tags:
79
+ - information-retrieval
80
+ - retrieval
81
+ - nano
82
+ - bm25
83
+ ---
84
+
85
+ # NanoR2MED
86
+
87
+ This dataset is a Nano-style retrieval dataset. Nano-series evaluation can be run easily with the [HAKARI-Bench](https://github.com/hotchpotch/hakari-bench).
88
+
89
+ NanoR2MED is derived from R2MED. It follows the Hugging Face Datasets layout convention used by [sentence-transformers/NanoBEIR-en](https://huggingface.co/datasets/sentence-transformers/NanoBEIR-en): each Nano split has separate `corpus`, `queries`, and `qrels` tables, and BM25 candidates are provided separately in a `bm25` table. This layout follows the NanoBEIR-style evaluation approach summarized in [NanoBEIR](https://huggingface.co/blog/sionic-ai/eval-sionic-nano-beir).
90
+
91
+ NanoR2MED contains 8 Nano retrieval splits derived from R2MED. Each split keeps up to 200 eligible queries and up to 10000 corpus documents, with exact duplicate query and document text removed where the generator records that policy.
92
+
93
+ ## Source Links
94
+
95
+ - Source benchmark: `R2MED`
96
+ - `R2MED/Bioinformatics`: https://huggingface.co/datasets/R2MED/Bioinformatics
97
+ - `R2MED/Biology`: https://huggingface.co/datasets/R2MED/Biology
98
+ - `R2MED/IIYi-Clinical`: https://huggingface.co/datasets/R2MED/IIYi-Clinical
99
+ - `R2MED/MedQA-Diag`: https://huggingface.co/datasets/R2MED/MedQA-Diag
100
+ - `R2MED/MedXpertQA-Exam`: https://huggingface.co/datasets/R2MED/MedXpertQA-Exam
101
+ - `R2MED/Medical-Sciences`: https://huggingface.co/datasets/R2MED/Medical-Sciences
102
+ - `R2MED/PMC-Clinical`: https://huggingface.co/datasets/R2MED/PMC-Clinical
103
+ - `R2MED/PMC-Treatment`: https://huggingface.co/datasets/R2MED/PMC-Treatment
104
+
105
+ ## Data Layout
106
+
107
+ This dataset uses four Hugging Face Datasets configs:
108
+
109
+ - `corpus`: documents with `_id` and `text`
110
+ - `queries`: queries with `_id` and `text`
111
+ - `qrels`: positive relevance labels with `query-id` and `corpus-id`
112
+ - `bm25`: BM25 candidate lists with `query-id` and `corpus-ids`
113
+
114
+ Each config uses the same Nano split names. If the actual generated dataset uses a different schema, config name, path layout, or field name, revise this section before publishing the README.
115
+
116
+ ## Construction Steps
117
+
118
+ This dataset was built as follows. If the actual generation procedure differs, revise this section before publishing the README.
119
+
120
+ 1. Use R2MED as the upstream benchmark or dataset family.
121
+ 2. Load the source datasets recorded in `manifest.json` and per-split metadata files.
122
+ 3. Use the source benchmark evaluation split, preferring `test` when available as the source evaluation split policy.
123
+ 4. Create one Nano split for each selected source retrieval task.
124
+ 5. Keep up to 200 eligible queries per Nano split.
125
+ 6. Include qrels-positive documents for the selected queries.
126
+ 7. Fill the corpus from source corpus order up to 10000 documents.
127
+ 8. Remove exact duplicate document text within each split. If a removed duplicate was referenced by qrels, rewrite qrels to the kept document id when the generator records that policy.
128
+ 9. Store document title and body as a single `text` field when the source provides both.
129
+ 10. Generate BM25 top-100 candidates with the tokenization policy recorded per split.
130
+ 11. If a qrels-positive document is missing from the raw BM25 result, insert it into the final `bm25` candidate list by replacing a tail non-positive candidate.
131
+
132
+ ## BM25 Subset Policy
133
+
134
+ The `bm25` config is a candidate subset for first-stage retrieval and reranking. It is not a separate source dataset. Each row contains one query id and a ranked list of corpus ids.
135
+
136
+ BM25 candidates are generated from the selected corpus for each split. The configured candidate cap is top-100. When a qrels-positive document is not present in the raw BM25 result, the missing positive is forced into the final candidate list by replacing a tail candidate that is not positive for that query. Candidate ids are kept unique after replacement.
137
+
138
+ ## Split Mapping
139
+
140
+ | Nano split | Source task | Source dataset | Queries | Corpus | Qrels |
141
+ |---|---|---|---:|---:|---:|
142
+ | `NanoR2MEDBiology` | `R2MEDBiologyRetrieval` | `R2MED/Biology` | 103 | 10000 | 374 |
143
+ | `NanoR2MEDBioinformatics` | `R2MEDBioinformaticsRetrieval` | `R2MED/Bioinformatics` | 77 | 10000 | 226 |
144
+ | `NanoR2MEDMedicalSciences` | `R2MEDMedicalSciencesRetrieval` | `R2MED/Medical-Sciences` | 88 | 10000 | 244 |
145
+ | `NanoR2MEDMedXpertQAExam` | `R2MEDMedXpertQAExamRetrieval` | `R2MED/MedXpertQA-Exam` | 97 | 10000 | 292 |
146
+ | `NanoR2MEDMedQADiag` | `R2MEDMedQADiagRetrieval` | `R2MED/MedQA-Diag` | 118 | 10000 | 522 |
147
+ | `NanoR2MEDPMCTreatment` | `R2MEDPMCTreatmentRetrieval` | `R2MED/PMC-Treatment` | 150 | 10000 | 315 |
148
+ | `NanoR2MEDPMCClinical` | `R2MEDPMCClinicalRetrieval` | `R2MED/PMC-Clinical` | 114 | 10000 | 248 |
149
+ | `NanoR2MEDIIYiClinical` | `R2MEDIIYiClinicalRetrieval` | `R2MED/IIYi-Clinical` | 129 | 10000 | 457 |
150
+
151
+ ## BM25 nDCG@10
152
+
153
+ `nDCG@10` is computed from the included BM25 ranking against the included qrels.
154
+
155
+ Tokenizer policy summary: `stemmer:en`.
156
+
157
+ | Nano split | Tokenizer | Forced BM25 positives | BM25 nDCG@10 |
158
+ |---|---|---:|---:|
159
+ | `NanoR2MEDBiology` | `stemmer:en` | 167 | 0.2513 |
160
+ | `NanoR2MEDBioinformatics` | `stemmer:en` | 111 | 0.1786 |
161
+ | `NanoR2MEDMedicalSciences` | `stemmer:en` | 138 | 0.1043 |
162
+ | `NanoR2MEDMedXpertQAExam` | `stemmer:en` | 255 | 0.0245 |
163
+ | `NanoR2MEDMedQADiag` | `stemmer:en` | 440 | 0.0281 |
164
+ | `NanoR2MEDPMCTreatment` | `stemmer:en` | 186 | 0.0180 |
165
+ | `NanoR2MEDPMCClinical` | `stemmer:en` | 63 | 0.3277 |
166
+ | `NanoR2MEDIIYiClinical` | `stemmer:en` | 268 | 0.1246 |
167
+
168
+ ## Skipped Tasks
169
+
170
+ No source tasks were skipped.
171
+
172
+ ## License
173
+
174
+ NanoR2MED is a derived dataset. Users must comply with the licenses, terms, and attribution requirements of the upstream datasets and benchmarks.
bm25/NanoR2MEDBioinformatics.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:06c4aede877e64495167f2fb864fb644c44b72b46891395340eca804ee5e3b72
3
+ size 33054
bm25/NanoR2MEDBiology.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16ea88ee18fa34d84080e018034c90ffa30613f41781445ea9af7fe6d1044149
3
+ size 47992
bm25/NanoR2MEDIIYiClinical.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9cee09602743c441a5ee1e8f02e9f3f03dd49a849fa31e46285d3bd59199db99
3
+ size 45535
bm25/NanoR2MEDMedQADiag.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:079d797b31b8091910e9df85718e2bc9c94f025260cdd813c32b406ea77a7f6b
3
+ size 45525
bm25/NanoR2MEDMedXpertQAExam.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3509b7b4a24c5a397cbda095e952144eb118a577bc5cabd5d185bc9c0cbb032c
3
+ size 35307
bm25/NanoR2MEDMedicalSciences.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4173726766ffdc206d02a98ce45ee8e73b18716c9a96c8a2b5a96f79b4f6ce47
3
+ size 48847
bm25/NanoR2MEDPMCClinical.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c6260b155323d88df4fcd3ec4dbe0379b8df46f0beb2205278cc746add3b6fc6
3
+ size 58427
bm25/NanoR2MEDPMCTreatment.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08abc0b532ce843a3a44f3ffc1b53e66e7f674d11fc0ab09bae1bca8454ce5f7
3
+ size 49117
manifest.json ADDED
@@ -0,0 +1,378 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "dataset": "NanoR2MED",
3
+ "benchmark": "R2MED",
4
+ "query_limit": 200,
5
+ "doc_limit": 10000,
6
+ "bm25_top_k": 100,
7
+ "splits": [
8
+ {
9
+ "split": "NanoR2MEDBiology",
10
+ "source_task": "R2MEDBiologyRetrieval",
11
+ "source_type": "Retrieval",
12
+ "source_dataset": {
13
+ "path": "R2MED/Biology",
14
+ "revision": "8b9fec2db9eda4b5742d03732213fbaee8169556"
15
+ },
16
+ "source_eval_splits": [
17
+ "test"
18
+ ],
19
+ "query_limit": 200,
20
+ "doc_limit": 10000,
21
+ "queries": 103,
22
+ "corpus": 10000,
23
+ "qrels": 374,
24
+ "bm25_rows": 103,
25
+ "bm25_top_k": 100,
26
+ "bm25_tokenization": {
27
+ "mode": "stemmer",
28
+ "language": "en",
29
+ "stemmer_algorithm": "english",
30
+ "tokenizer_name": null,
31
+ "reason": "language 'en' mapped to stemmer 'english'"
32
+ },
33
+ "qrels_coverage": {
34
+ "total": 374,
35
+ "hits": 374,
36
+ "recall": 1.0
37
+ },
38
+ "ndcg_at_10": 0.25128449065032643,
39
+ "ndcg_at_100": 0.4296835025091202,
40
+ "source_container": "task.corpus/queries/relevant_docs",
41
+ "source_eval_split_used": "test",
42
+ "source_query_count": 103,
43
+ "source_corpus_count": 57359,
44
+ "source_qrels_query_count": 103,
45
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
46
+ "skipped_queries_missing_text_or_qrels": 0,
47
+ "skipped_duplicate_query_texts": 0,
48
+ "duplicate_doc_texts_removed": 1058,
49
+ "qrels_rewritten_for_duplicate_text": 9,
50
+ "forced_queries": 68,
51
+ "forced_doc_count": 167,
52
+ "missing_positive_doc_count_after_forcing": 0
53
+ },
54
+ {
55
+ "split": "NanoR2MEDBioinformatics",
56
+ "source_task": "R2MEDBioinformaticsRetrieval",
57
+ "source_type": "Retrieval",
58
+ "source_dataset": {
59
+ "path": "R2MED/Bioinformatics",
60
+ "revision": "6021fce366892cbfd7837fa85a4128ea93315e18"
61
+ },
62
+ "source_eval_splits": [
63
+ "test"
64
+ ],
65
+ "query_limit": 200,
66
+ "doc_limit": 10000,
67
+ "queries": 77,
68
+ "corpus": 10000,
69
+ "qrels": 226,
70
+ "bm25_rows": 77,
71
+ "bm25_top_k": 100,
72
+ "bm25_tokenization": {
73
+ "mode": "stemmer",
74
+ "language": "en",
75
+ "stemmer_algorithm": "english",
76
+ "tokenizer_name": null,
77
+ "reason": "language 'en' mapped to stemmer 'english'"
78
+ },
79
+ "qrels_coverage": {
80
+ "total": 226,
81
+ "hits": 226,
82
+ "recall": 1.0
83
+ },
84
+ "ndcg_at_10": 0.17857929009122858,
85
+ "ndcg_at_100": 0.35060738065375957,
86
+ "source_container": "task.corpus/queries/relevant_docs",
87
+ "source_eval_split_used": "test",
88
+ "source_query_count": 77,
89
+ "source_corpus_count": 47471,
90
+ "source_qrels_query_count": 77,
91
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
92
+ "skipped_queries_missing_text_or_qrels": 0,
93
+ "skipped_duplicate_query_texts": 0,
94
+ "duplicate_doc_texts_removed": 6,
95
+ "qrels_rewritten_for_duplicate_text": 1,
96
+ "forced_queries": 47,
97
+ "forced_doc_count": 111,
98
+ "missing_positive_doc_count_after_forcing": 0
99
+ },
100
+ {
101
+ "split": "NanoR2MEDMedicalSciences",
102
+ "source_task": "R2MEDMedicalSciencesRetrieval",
103
+ "source_type": "Retrieval",
104
+ "source_dataset": {
105
+ "path": "R2MED/Medical-Sciences",
106
+ "revision": "1b48911514c80bf9182222d99752ad75e23b4b47"
107
+ },
108
+ "source_eval_splits": [
109
+ "test"
110
+ ],
111
+ "query_limit": 200,
112
+ "doc_limit": 10000,
113
+ "queries": 88,
114
+ "corpus": 10000,
115
+ "qrels": 244,
116
+ "bm25_rows": 88,
117
+ "bm25_top_k": 100,
118
+ "bm25_tokenization": {
119
+ "mode": "stemmer",
120
+ "language": "en",
121
+ "stemmer_algorithm": "english",
122
+ "tokenizer_name": null,
123
+ "reason": "language 'en' mapped to stemmer 'english'"
124
+ },
125
+ "qrels_coverage": {
126
+ "total": 244,
127
+ "hits": 244,
128
+ "recall": 1.0
129
+ },
130
+ "ndcg_at_10": 0.10428646688120163,
131
+ "ndcg_at_100": 0.29337732163029967,
132
+ "source_container": "task.corpus/queries/relevant_docs",
133
+ "source_eval_split_used": "test",
134
+ "source_query_count": 88,
135
+ "source_corpus_count": 34792,
136
+ "source_qrels_query_count": 88,
137
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
138
+ "skipped_queries_missing_text_or_qrels": 0,
139
+ "skipped_duplicate_query_texts": 0,
140
+ "duplicate_doc_texts_removed": 66,
141
+ "qrels_rewritten_for_duplicate_text": 0,
142
+ "forced_queries": 62,
143
+ "forced_doc_count": 138,
144
+ "missing_positive_doc_count_after_forcing": 0
145
+ },
146
+ {
147
+ "split": "NanoR2MEDMedXpertQAExam",
148
+ "source_task": "R2MEDMedXpertQAExamRetrieval",
149
+ "source_type": "Retrieval",
150
+ "source_dataset": {
151
+ "path": "R2MED/MedXpertQA-Exam",
152
+ "revision": "b457ea43db9ae5db74c3a3e5be0a213d0f85ac3a"
153
+ },
154
+ "source_eval_splits": [
155
+ "test"
156
+ ],
157
+ "query_limit": 200,
158
+ "doc_limit": 10000,
159
+ "queries": 97,
160
+ "corpus": 10000,
161
+ "qrels": 292,
162
+ "bm25_rows": 97,
163
+ "bm25_top_k": 100,
164
+ "bm25_tokenization": {
165
+ "mode": "stemmer",
166
+ "language": "en",
167
+ "stemmer_algorithm": "english",
168
+ "tokenizer_name": null,
169
+ "reason": "language 'en' mapped to stemmer 'english'"
170
+ },
171
+ "qrels_coverage": {
172
+ "total": 292,
173
+ "hits": 292,
174
+ "recall": 1.0
175
+ },
176
+ "ndcg_at_10": 0.024462598222211914,
177
+ "ndcg_at_100": 0.2304796798768815,
178
+ "source_container": "task.corpus/queries/relevant_docs",
179
+ "source_eval_split_used": "test",
180
+ "source_query_count": 97,
181
+ "source_corpus_count": 61379,
182
+ "source_qrels_query_count": 97,
183
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
184
+ "skipped_queries_missing_text_or_qrels": 0,
185
+ "skipped_duplicate_query_texts": 0,
186
+ "duplicate_doc_texts_removed": 0,
187
+ "qrels_rewritten_for_duplicate_text": 0,
188
+ "forced_queries": 90,
189
+ "forced_doc_count": 255,
190
+ "missing_positive_doc_count_after_forcing": 0
191
+ },
192
+ {
193
+ "split": "NanoR2MEDMedQADiag",
194
+ "source_task": "R2MEDMedQADiagRetrieval",
195
+ "source_type": "Retrieval",
196
+ "source_dataset": {
197
+ "path": "R2MED/MedQA-Diag",
198
+ "revision": "78b585990279cc01a493f876c1b0cf09557fba57"
199
+ },
200
+ "source_eval_splits": [
201
+ "test"
202
+ ],
203
+ "query_limit": 200,
204
+ "doc_limit": 10000,
205
+ "queries": 118,
206
+ "corpus": 10000,
207
+ "qrels": 522,
208
+ "bm25_rows": 118,
209
+ "bm25_top_k": 100,
210
+ "bm25_tokenization": {
211
+ "mode": "stemmer",
212
+ "language": "en",
213
+ "stemmer_algorithm": "english",
214
+ "tokenizer_name": null,
215
+ "reason": "language 'en' mapped to stemmer 'english'"
216
+ },
217
+ "qrels_coverage": {
218
+ "total": 522,
219
+ "hits": 522,
220
+ "recall": 1.0
221
+ },
222
+ "ndcg_at_10": 0.028104212132602936,
223
+ "ndcg_at_100": 0.2642743613844765,
224
+ "source_container": "task.corpus/queries/relevant_docs",
225
+ "source_eval_split_used": "test",
226
+ "source_query_count": 118,
227
+ "source_corpus_count": 56194,
228
+ "source_qrels_query_count": 118,
229
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
230
+ "skipped_queries_missing_text_or_qrels": 0,
231
+ "skipped_duplicate_query_texts": 0,
232
+ "duplicate_doc_texts_removed": 6,
233
+ "qrels_rewritten_for_duplicate_text": 0,
234
+ "forced_queries": 109,
235
+ "forced_doc_count": 440,
236
+ "missing_positive_doc_count_after_forcing": 0
237
+ },
238
+ {
239
+ "split": "NanoR2MEDPMCTreatment",
240
+ "source_task": "R2MEDPMCTreatmentRetrieval",
241
+ "source_type": "Retrieval",
242
+ "source_dataset": {
243
+ "path": "R2MED/PMC-Treatment",
244
+ "revision": "53c489a44a3664ba352c07550b72b4525a5968d5"
245
+ },
246
+ "source_eval_splits": [
247
+ "test"
248
+ ],
249
+ "query_limit": 200,
250
+ "doc_limit": 10000,
251
+ "queries": 150,
252
+ "corpus": 10000,
253
+ "qrels": 315,
254
+ "bm25_rows": 150,
255
+ "bm25_top_k": 100,
256
+ "bm25_tokenization": {
257
+ "mode": "stemmer",
258
+ "language": "en",
259
+ "stemmer_algorithm": "english",
260
+ "tokenizer_name": null,
261
+ "reason": "language 'en' mapped to stemmer 'english'"
262
+ },
263
+ "qrels_coverage": {
264
+ "total": 315,
265
+ "hits": 315,
266
+ "recall": 1.0
267
+ },
268
+ "ndcg_at_10": 0.01798988334744109,
269
+ "ndcg_at_100": 0.2202813875703761,
270
+ "source_container": "task.corpus/queries/relevant_docs",
271
+ "source_eval_split_used": "test",
272
+ "source_query_count": 150,
273
+ "source_corpus_count": 28809,
274
+ "source_qrels_query_count": 150,
275
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
276
+ "skipped_queries_missing_text_or_qrels": 0,
277
+ "skipped_duplicate_query_texts": 0,
278
+ "duplicate_doc_texts_removed": 6,
279
+ "qrels_rewritten_for_duplicate_text": 0,
280
+ "forced_queries": 106,
281
+ "forced_doc_count": 186,
282
+ "missing_positive_doc_count_after_forcing": 0
283
+ },
284
+ {
285
+ "split": "NanoR2MEDPMCClinical",
286
+ "source_task": "R2MEDPMCClinicalRetrieval",
287
+ "source_type": "Retrieval",
288
+ "source_dataset": {
289
+ "path": "R2MED/PMC-Clinical",
290
+ "revision": "812829522f7eaa407ef82b96717be85788a50f7e"
291
+ },
292
+ "source_eval_splits": [
293
+ "test"
294
+ ],
295
+ "query_limit": 200,
296
+ "doc_limit": 10000,
297
+ "queries": 114,
298
+ "corpus": 10000,
299
+ "qrels": 248,
300
+ "bm25_rows": 114,
301
+ "bm25_top_k": 100,
302
+ "bm25_tokenization": {
303
+ "mode": "stemmer",
304
+ "language": "en",
305
+ "stemmer_algorithm": "english",
306
+ "tokenizer_name": null,
307
+ "reason": "language 'en' mapped to stemmer 'english'"
308
+ },
309
+ "qrels_coverage": {
310
+ "total": 248,
311
+ "hits": 248,
312
+ "recall": 1.0
313
+ },
314
+ "ndcg_at_10": 0.3277008274619314,
315
+ "ndcg_at_100": 0.4621547795535059,
316
+ "source_container": "task.corpus/queries/relevant_docs",
317
+ "source_eval_split_used": "test",
318
+ "source_query_count": 114,
319
+ "source_corpus_count": 60406,
320
+ "source_qrels_query_count": 114,
321
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
322
+ "skipped_queries_missing_text_or_qrels": 0,
323
+ "skipped_duplicate_query_texts": 0,
324
+ "duplicate_doc_texts_removed": 0,
325
+ "qrels_rewritten_for_duplicate_text": 0,
326
+ "forced_queries": 46,
327
+ "forced_doc_count": 63,
328
+ "missing_positive_doc_count_after_forcing": 0
329
+ },
330
+ {
331
+ "split": "NanoR2MEDIIYiClinical",
332
+ "source_task": "R2MEDIIYiClinicalRetrieval",
333
+ "source_type": "Retrieval",
334
+ "source_dataset": {
335
+ "path": "R2MED/IIYi-Clinical",
336
+ "revision": "974abbc9bc281c3169180a6aa5d7586cfd2f5877"
337
+ },
338
+ "source_eval_splits": [
339
+ "test"
340
+ ],
341
+ "query_limit": 200,
342
+ "doc_limit": 10000,
343
+ "queries": 129,
344
+ "corpus": 10000,
345
+ "qrels": 457,
346
+ "bm25_rows": 129,
347
+ "bm25_top_k": 100,
348
+ "bm25_tokenization": {
349
+ "mode": "stemmer",
350
+ "language": "en",
351
+ "stemmer_algorithm": "english",
352
+ "tokenizer_name": null,
353
+ "reason": "language 'en' mapped to stemmer 'english'"
354
+ },
355
+ "qrels_coverage": {
356
+ "total": 457,
357
+ "hits": 457,
358
+ "recall": 1.0
359
+ },
360
+ "ndcg_at_10": 0.12460354977311297,
361
+ "ndcg_at_100": 0.33378196865460924,
362
+ "source_container": "task.corpus/queries/relevant_docs",
363
+ "source_eval_split_used": "test",
364
+ "source_query_count": 129,
365
+ "source_corpus_count": 10449,
366
+ "source_qrels_query_count": 129,
367
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
368
+ "skipped_queries_missing_text_or_qrels": 0,
369
+ "skipped_duplicate_query_texts": 0,
370
+ "duplicate_doc_texts_removed": 0,
371
+ "qrels_rewritten_for_duplicate_text": 0,
372
+ "forced_queries": 113,
373
+ "forced_doc_count": 268,
374
+ "missing_positive_doc_count_after_forcing": 0
375
+ }
376
+ ],
377
+ "skipped": []
378
+ }
metadata/NanoR2MEDBioinformatics.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDBioinformatics",
3
+ "source_task": "R2MEDBioinformaticsRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/Bioinformatics",
7
+ "revision": "6021fce366892cbfd7837fa85a4128ea93315e18"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 77,
15
+ "corpus": 10000,
16
+ "qrels": 226,
17
+ "bm25_rows": 77,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 226,
28
+ "hits": 226,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.17857929009122858,
32
+ "ndcg_at_100": 0.35060738065375957,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 77,
36
+ "source_corpus_count": 47471,
37
+ "source_qrels_query_count": 77,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 6,
42
+ "qrels_rewritten_for_duplicate_text": 1,
43
+ "forced_queries": 47,
44
+ "forced_doc_count": 111,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
metadata/NanoR2MEDBiology.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDBiology",
3
+ "source_task": "R2MEDBiologyRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/Biology",
7
+ "revision": "8b9fec2db9eda4b5742d03732213fbaee8169556"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 103,
15
+ "corpus": 10000,
16
+ "qrels": 374,
17
+ "bm25_rows": 103,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 374,
28
+ "hits": 374,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.25128449065032643,
32
+ "ndcg_at_100": 0.4296835025091202,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 103,
36
+ "source_corpus_count": 57359,
37
+ "source_qrels_query_count": 103,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 1058,
42
+ "qrels_rewritten_for_duplicate_text": 9,
43
+ "forced_queries": 68,
44
+ "forced_doc_count": 167,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
metadata/NanoR2MEDIIYiClinical.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDIIYiClinical",
3
+ "source_task": "R2MEDIIYiClinicalRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/IIYi-Clinical",
7
+ "revision": "974abbc9bc281c3169180a6aa5d7586cfd2f5877"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 129,
15
+ "corpus": 10000,
16
+ "qrels": 457,
17
+ "bm25_rows": 129,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 457,
28
+ "hits": 457,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.12460354977311297,
32
+ "ndcg_at_100": 0.33378196865460924,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 129,
36
+ "source_corpus_count": 10449,
37
+ "source_qrels_query_count": 129,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 0,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 113,
44
+ "forced_doc_count": 268,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
metadata/NanoR2MEDMedQADiag.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDMedQADiag",
3
+ "source_task": "R2MEDMedQADiagRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/MedQA-Diag",
7
+ "revision": "78b585990279cc01a493f876c1b0cf09557fba57"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 118,
15
+ "corpus": 10000,
16
+ "qrels": 522,
17
+ "bm25_rows": 118,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 522,
28
+ "hits": 522,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.028104212132602936,
32
+ "ndcg_at_100": 0.2642743613844765,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 118,
36
+ "source_corpus_count": 56194,
37
+ "source_qrels_query_count": 118,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 6,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 109,
44
+ "forced_doc_count": 440,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
metadata/NanoR2MEDMedXpertQAExam.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDMedXpertQAExam",
3
+ "source_task": "R2MEDMedXpertQAExamRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/MedXpertQA-Exam",
7
+ "revision": "b457ea43db9ae5db74c3a3e5be0a213d0f85ac3a"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 97,
15
+ "corpus": 10000,
16
+ "qrels": 292,
17
+ "bm25_rows": 97,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 292,
28
+ "hits": 292,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.024462598222211914,
32
+ "ndcg_at_100": 0.2304796798768815,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 97,
36
+ "source_corpus_count": 61379,
37
+ "source_qrels_query_count": 97,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 0,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 90,
44
+ "forced_doc_count": 255,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
metadata/NanoR2MEDMedicalSciences.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDMedicalSciences",
3
+ "source_task": "R2MEDMedicalSciencesRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/Medical-Sciences",
7
+ "revision": "1b48911514c80bf9182222d99752ad75e23b4b47"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 88,
15
+ "corpus": 10000,
16
+ "qrels": 244,
17
+ "bm25_rows": 88,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 244,
28
+ "hits": 244,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.10428646688120163,
32
+ "ndcg_at_100": 0.29337732163029967,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 88,
36
+ "source_corpus_count": 34792,
37
+ "source_qrels_query_count": 88,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 66,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 62,
44
+ "forced_doc_count": 138,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
metadata/NanoR2MEDPMCClinical.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDPMCClinical",
3
+ "source_task": "R2MEDPMCClinicalRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/PMC-Clinical",
7
+ "revision": "812829522f7eaa407ef82b96717be85788a50f7e"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 114,
15
+ "corpus": 10000,
16
+ "qrels": 248,
17
+ "bm25_rows": 114,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 248,
28
+ "hits": 248,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.3277008274619314,
32
+ "ndcg_at_100": 0.4621547795535059,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 114,
36
+ "source_corpus_count": 60406,
37
+ "source_qrels_query_count": 114,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 0,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 46,
44
+ "forced_doc_count": 63,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }
metadata/NanoR2MEDPMCTreatment.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "split": "NanoR2MEDPMCTreatment",
3
+ "source_task": "R2MEDPMCTreatmentRetrieval",
4
+ "source_type": "Retrieval",
5
+ "source_dataset": {
6
+ "path": "R2MED/PMC-Treatment",
7
+ "revision": "53c489a44a3664ba352c07550b72b4525a5968d5"
8
+ },
9
+ "source_eval_splits": [
10
+ "test"
11
+ ],
12
+ "query_limit": 200,
13
+ "doc_limit": 10000,
14
+ "queries": 150,
15
+ "corpus": 10000,
16
+ "qrels": 315,
17
+ "bm25_rows": 150,
18
+ "bm25_top_k": 100,
19
+ "bm25_tokenization": {
20
+ "mode": "stemmer",
21
+ "language": "en",
22
+ "stemmer_algorithm": "english",
23
+ "tokenizer_name": null,
24
+ "reason": "language 'en' mapped to stemmer 'english'"
25
+ },
26
+ "qrels_coverage": {
27
+ "total": 315,
28
+ "hits": 315,
29
+ "recall": 1.0
30
+ },
31
+ "ndcg_at_10": 0.01798988334744109,
32
+ "ndcg_at_100": 0.2202813875703761,
33
+ "source_container": "task.corpus/queries/relevant_docs",
34
+ "source_eval_split_used": "test",
35
+ "source_query_count": 150,
36
+ "source_corpus_count": 28809,
37
+ "source_qrels_query_count": 150,
38
+ "skipped_queries_with_more_than_bm25_top_k_positives": 0,
39
+ "skipped_queries_missing_text_or_qrels": 0,
40
+ "skipped_duplicate_query_texts": 0,
41
+ "duplicate_doc_texts_removed": 6,
42
+ "qrels_rewritten_for_duplicate_text": 0,
43
+ "forced_queries": 106,
44
+ "forced_doc_count": 186,
45
+ "missing_positive_doc_count_after_forcing": 0
46
+ }