1602353775wzj commited on
Commit
6a10fd0
·
verified ·
1 Parent(s): 517b122

Add HaikuSeasonClassification (4 seasons; stratified 80/20)

Browse files
README.md ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ task_categories:
4
+ - text-classification
5
+ task_ids:
6
+ - multi-class-classification
7
+ language:
8
+ - ja
9
+ multilinguality:
10
+ - monolingual
11
+ size_categories:
12
+ - n<1K
13
+ pretty_name: HaikuSeasonClassification
14
+ tags:
15
+ - poetry
16
+ - japanese
17
+ - haiku
18
+ - season-classification
19
+ - kigo
20
+ - mteb
21
+ - poetrymteb
22
+ - embedding-evaluation
23
+ annotations_creators:
24
+ - expert-generated
25
+ source_datasets:
26
+ - shigr3/haiku
27
+ configs:
28
+ - config_name: default
29
+ data_files:
30
+ - split: train
31
+ path: data/train-*
32
+ - split: test
33
+ path: data/test-*
34
+ default: true
35
+ dataset_info:
36
+ - config_name: default
37
+ features:
38
+ - name: id
39
+ dtype: string
40
+ - name: poem
41
+ dtype: string
42
+ - name: author
43
+ dtype: string
44
+ - name: season_word
45
+ dtype: string
46
+ - name: label
47
+ dtype: int64
48
+ - name: label_name
49
+ dtype: string
50
+ splits:
51
+ - name: train
52
+ num_examples: 113
53
+ - name: test
54
+ num_examples: 29
55
+ ---
56
+
57
+ # HaikuSeasonClassification
58
+
59
+ Single-label **Japanese haiku season classification** (春/夏/秋/冬) for PoetryMTEB, derived from [shigr3/haiku](https://huggingface.co/datasets/shigr3/haiku).
60
+
61
+ ## Dataset Card
62
+
63
+ | Item | Description |
64
+ |------|-------------|
65
+ | **Source** | [shigr3/haiku](https://huggingface.co/datasets/shigr3/haiku) |
66
+ | **Languages** | Japanese (`ja`) |
67
+ | **Unit** | Haiku text (`poem`) |
68
+ | **Classes** | 4 seasons |
69
+ | **Size** | train=113; test=29 |
70
+ | **Splits** | Stratified by season ≈ 80% / 20%, seed=42 |
71
+ | **License** | [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/) (same as upstream) |
72
+ | **Evaluation metrics** | Classification on embeddings: **accuracy**, **macro/weighted F1** |
73
+
74
+ ## Label taxonomy (4)
75
+
76
+ | id | label_name | English | romanization | train | test | total |
77
+ |---:|------------|---------|--------------|------:|-----:|------:|
78
+ | 0 | `春` | Spring | haru | 28 | 7 | 35 |
79
+ | 1 | `夏` | Summer | natsu | 30 | 8 | 38 |
80
+ | 2 | `秋` | Autumn | aki | 27 | 7 | 34 |
81
+ | 3 | `冬` | Winter | fuyu | 28 | 7 | 35 |
82
+
83
+ ## Features
84
+
85
+ | Field | Type | Description |
86
+ |-------|------|-------------|
87
+ | `id` | string | Example id |
88
+ | `poem` | string | Haiku body |
89
+ | `author` | string | Poet |
90
+ | `season_word` | string | Kigo (季語); metadata, not the classification target |
91
+ | `label` | int64 | Season class 0–3 |
92
+ | `label_name` | string | `春` / `夏` / `秋` / `冬` |
93
+
94
+ ## Attribution
95
+
96
+ Dataset: Famous Japanese Haiku Dataset — author **shigr3** — https://huggingface.co/datasets/shigr3/haiku
data/test-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4ec8ecda5307a119792cfe908ac6a9f4630878407914de08203ddff1f712a83
3
+ size 4394
data/train-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4253bd629d357a81352d902f05bfe9055c9d10d40844908abec0e3e6a2982a8
3
+ size 8525
dataset_infos.yaml ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ configs:
2
+ - config_name: default
3
+ data_files:
4
+ - path: data/train-*.parquet
5
+ split: train
6
+ - path: data/test-*.parquet
7
+ split: test
8
+ default: true
9
+ dataset_info:
10
+ configs:
11
+ - config_name: default
12
+ dataset_size: 12919
13
+ download_size: 12919
14
+ features:
15
+ - dtype: string
16
+ name: id
17
+ - dtype: string
18
+ name: poem
19
+ - dtype: string
20
+ name: author
21
+ - dtype: string
22
+ name: season_word
23
+ - dtype: int64
24
+ name: label
25
+ - dtype: string
26
+ name: label_name
27
+ splits:
28
+ - name: train
29
+ num_bytes: 8525
30
+ num_examples: 113
31
+ - name: test
32
+ num_bytes: 4394
33
+ num_examples: 29
label_taxonomy.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "n_classes": 4,
3
+ "labels": [
4
+ {
5
+ "id": 0,
6
+ "name": "春",
7
+ "name_en": "Spring",
8
+ "romanization": "haru"
9
+ },
10
+ {
11
+ "id": 1,
12
+ "name": "夏",
13
+ "name_en": "Summer",
14
+ "romanization": "natsu"
15
+ },
16
+ {
17
+ "id": 2,
18
+ "name": "秋",
19
+ "name_en": "Autumn",
20
+ "romanization": "aki"
21
+ },
22
+ {
23
+ "id": 3,
24
+ "name": "冬",
25
+ "name_en": "Winter",
26
+ "romanization": "fuyu"
27
+ }
28
+ ],
29
+ "counts": {
30
+ "train": {
31
+ "夏": 30,
32
+ "春": 28,
33
+ "秋": 27,
34
+ "冬": 28
35
+ },
36
+ "test": {
37
+ "冬": 7,
38
+ "春": 7,
39
+ "秋": 7,
40
+ "夏": 8
41
+ }
42
+ }
43
+ }