README.md CHANGED
@@ -4,7 +4,7 @@ license: cc0-1.0
4
  task_categories:
5
  - image-classification
6
  size_categories:
7
- - 1K<n<10K
8
  configs:
9
  - config_name: image
10
  data_files:
@@ -27,110 +27,41 @@ dataset_info:
27
  dtype:
28
  class_label:
29
  names:
30
- '0': aeroform
31
- '1': billowing
32
- '2': collection
33
- '3': column
34
- '4': convergence
35
- '5': cool
36
- '6': crush
37
- '7': crystal
38
- '8': dispersion
39
- '9': earth
40
- '10': empower
41
- '11': entwine
42
- '12': fire
43
- '13': float
44
- '14': focus
45
- '15': gather
46
- '16': levitation
47
- '17': light
48
- '18': orb
49
- '19': pull
50
- '20': region
51
- '21': repetition
52
- '22': water
53
- '23': weave
54
- '24': wind-directs-air
55
- '25': wind-underfoot
56
  splits:
57
  - name: train
58
- num_bytes: 25373702
59
- num_examples: 7312
60
  - name: validation
61
- num_bytes: 5113048
62
- num_examples: 1829
63
- download_size: 27982004
64
- dataset_size: 30486750
65
  - config_name: vector
66
  features:
67
  - name: id
68
  dtype: string
69
- - name: sample_id
70
- dtype: string
71
  - name: sign
72
  dtype: string
73
  - name: data
74
  list:
75
  list:
76
  - name: t
77
- dtype: float64
78
  - name: x
79
  dtype: float64
80
- - name: y
81
  dtype: float64
82
- - name: pose
83
- struct:
84
- - name: center_x
85
- dtype: float64
86
- - name: center_y
87
- dtype: float64
88
- - name: scale_x
89
- dtype: float64
90
- - name: scale_y
91
- dtype: float64
92
- - name: angle
93
- dtype: float64
94
- - name: angle_sin
95
- dtype: float64
96
- - name: angle_cos
97
- dtype: float64
98
- - name: normalisation
99
- struct:
100
- - name: min_x
101
- dtype: float64
102
- - name: min_y
103
- dtype: float64
104
- - name: max_x
105
- dtype: float64
106
- - name: max_y
107
- dtype: float64
108
- - name: width
109
- dtype: float64
110
- - name: height
111
- dtype: float64
112
- - name: scale
113
- dtype: float64
114
- - name: source
115
- struct:
116
- - name: captured_at
117
- dtype: string
118
- - name: review_status
119
- dtype: string
120
- - name: canvas_width
121
- dtype: int64
122
- - name: canvas_height
123
- dtype: int64
124
- - name: device_pixel_ratio
125
- dtype: float64
126
- - name: reference_size
127
- dtype: int64
128
  splits:
129
  - name: train
130
- num_bytes: 19431967
131
- num_examples: 9141
132
- download_size: 19341868
133
- dataset_size: 19431967
 
 
 
134
  ---
135
  # WHA Spell Simulator Glyphs
136
 
@@ -146,46 +77,20 @@ preserving aspect ratio.
146
  One record per sample:
147
 
148
  - `id` — content hash of the raw sample
149
- - `sample_id` — database sample id
150
  - `sign` — glyph class label
151
  - `data` — list of strokes; each stroke is a list of `{t, x, y}` points
152
  (`t` in ms from first pen-down, `x`/`y` normalised to 0..1)
153
- - `pose` — normalized reference overlay center, scale, and rotation targets
154
 
155
  ```python
156
  from datasets import load_dataset
157
  vector = load_dataset("wha-spell-simulator/labelled-samples", "vector")
158
  ```
159
 
160
- | sign | train |
161
- |---|---|
162
- | aeroform | 331 |
163
- | billowing | 318 |
164
- | collection | 356 |
165
- | column | 382 |
166
- | convergence | 571 |
167
- | cool | 342 |
168
- | crush | 338 |
169
- | crystal | 333 |
170
- | dispersion | 343 |
171
- | earth | 326 |
172
- | empower | 333 |
173
- | entwine | 332 |
174
- | fire | 351 |
175
- | float | 385 |
176
- | focus | 330 |
177
- | gather | 343 |
178
- | levitation | 349 |
179
- | light | 323 |
180
- | orb | 340 |
181
- | pull | 327 |
182
- | region | 400 |
183
- | repetition | 348 |
184
- | water | 335 |
185
- | weave | 337 |
186
- | wind-directs-air | 343 |
187
- | wind-underfoot | 325 |
188
- | **total** | 9141 |
189
 
190
  ## `image` config
191
 
@@ -198,37 +103,12 @@ images = load_dataset("wha-spell-simulator/labelled-samples", "image")
198
 
199
  | sign | train | validation |
200
  |---|---|---|
201
- | aeroform | 265 | 66 |
202
- | billowing | 254 | 64 |
203
- | collection | 285 | 71 |
204
- | column | 306 | 76 |
205
- | convergence | 457 | 114 |
206
- | cool | 274 | 68 |
207
- | crush | 270 | 68 |
208
- | crystal | 266 | 67 |
209
- | dispersion | 274 | 69 |
210
- | earth | 261 | 65 |
211
- | empower | 266 | 67 |
212
- | entwine | 266 | 66 |
213
- | fire | 281 | 70 |
214
- | float | 308 | 77 |
215
- | focus | 264 | 66 |
216
- | gather | 274 | 69 |
217
- | levitation | 279 | 70 |
218
- | light | 258 | 65 |
219
- | orb | 272 | 68 |
220
- | pull | 262 | 65 |
221
- | region | 320 | 80 |
222
- | repetition | 278 | 70 |
223
- | water | 268 | 67 |
224
- | weave | 270 | 67 |
225
- | wind-directs-air | 274 | 69 |
226
- | wind-underfoot | 260 | 65 |
227
- | **total** | 7312 | 1829 |
228
 
229
  ## Provenance & licensing
230
 
231
  Collected via the [WHA spell simulator](https://github.com/cosykid/wha-spell-simulator) Sample Maker.
232
- Samples contain no personal information: only ids, class labels, stroke
233
- geometry, normalized pose targets, and non-identifying capture metadata
234
- survive the export pipeline.
 
4
  task_categories:
5
  - image-classification
6
  size_categories:
7
+ - n<1K
8
  configs:
9
  - config_name: image
10
  data_files:
 
27
  dtype:
28
  class_label:
29
  names:
30
+ '0': asclepio
31
+ '1': oruferia
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
32
  splits:
33
  - name: train
34
+ num_bytes: 2137
35
+ num_examples: 6
36
  - name: validation
37
+ num_bytes: 2137
38
+ num_examples: 6
39
+ download_size: 10140
40
+ dataset_size: 4274
41
  - config_name: vector
42
  features:
43
  - name: id
44
  dtype: string
 
 
45
  - name: sign
46
  dtype: string
47
  - name: data
48
  list:
49
  list:
50
  - name: t
51
+ dtype: int64
52
  - name: x
53
  dtype: float64
54
+ - name: 'y'
55
  dtype: float64
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
56
  splits:
57
  - name: train
58
+ num_bytes: 462
59
+ num_examples: 6
60
+ - name: validation
61
+ num_bytes: 158
62
+ num_examples: 2
63
+ download_size: 6114
64
+ dataset_size: 620
65
  ---
66
  # WHA Spell Simulator Glyphs
67
 
 
77
  One record per sample:
78
 
79
  - `id` — content hash of the raw sample
 
80
  - `sign` — glyph class label
81
  - `data` — list of strokes; each stroke is a list of `{t, x, y}` points
82
  (`t` in ms from first pen-down, `x`/`y` normalised to 0..1)
 
83
 
84
  ```python
85
  from datasets import load_dataset
86
  vector = load_dataset("wha-spell-simulator/labelled-samples", "vector")
87
  ```
88
 
89
+ | sign | train | validation |
90
+ |---|---|---|
91
+ | asclepio | 3 | 0 |
92
+ | oruferia | 3 | 2 |
93
+ | **total** | 6 | 2 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
94
 
95
  ## `image` config
96
 
 
103
 
104
  | sign | train | validation |
105
  |---|---|---|
106
+ | asclepio | 3 | 3 |
107
+ | oruferia | 3 | 3 |
108
+ | **total** | 6 | 6 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
109
 
110
  ## Provenance & licensing
111
 
112
  Collected via the [WHA spell simulator](https://github.com/cosykid/wha-spell-simulator) Sample Maker.
113
+ Samples contain no personal information: only the content hash, class
114
+ label, and stroke geometry survive the export pipeline.
 
image/train-00000-of-00001.parquet CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f7b50e8997fc09ce969c43adcce0fff81deecda0261a699b974d8e461e49bb6b
3
- size 22375633
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a235d774134e3145732b8986c13332be778399a688a5d5df02955a007fe288a5
3
+ size 5070
image/validation-00000-of-00001.parquet CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:be65ff8b975715d89d45189b68f962179a85c71ff0f022f91080f102ab9f4ab0
3
- size 5606371
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a235d774134e3145732b8986c13332be778399a688a5d5df02955a007fe288a5
3
+ size 5070
vector/train-00000-of-00001.parquet CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8900f772ea708d194b3e924734fd808e873ae47cfe72a274576a50bf1929c38a
3
- size 19341868
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b562089c4abff58a191dfdb09b9522e641ae3da673990dd92da7a7fba9dbea7
3
+ size 3174
vector/validation-00000-of-00001.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:766cccd7bb1e08ff79f1467c537c0b06485df6eca334f6269bfe8bd8b0b60e10
3
+ size 2940