Ill-Ness commited on
Commit
afdcfb2
·
verified ·
1 Parent(s): 01717ef

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +307 -3
README.md CHANGED
@@ -1,3 +1,307 @@
1
- ---
2
- license: apache-2.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ task_categories:
4
+ - text-generation
5
+ language:
6
+ - de
7
+ tags:
8
+ - synthetic
9
+ - personas
10
+ - german
11
+ - custom
12
+ - datadesigner-inspired
13
+ size_categories:
14
+ - n<1K
15
+ dataset_info:
16
+ features:
17
+ - name: uuid
18
+ dtype: string
19
+ - name: professional_persona
20
+ dtype: string
21
+ - name: sports_persona
22
+ dtype: string
23
+ - name: arts_persona
24
+ dtype: string
25
+ - name: travel_persona
26
+ dtype: string
27
+ - name: culinary_persona
28
+ dtype: string
29
+ - name: persona
30
+ dtype: string
31
+ - name: cultural_background
32
+ dtype: string
33
+ - name: skills_and_expertise
34
+ dtype: string
35
+ - name: skills_and_expertise_list
36
+ sequence: string
37
+ - name: hobbies_and_interests
38
+ dtype: string
39
+ - name: hobbies_and_interests_list
40
+ sequence: string
41
+ - name: career_goals_and_ambitions
42
+ dtype: string
43
+ - name: sex
44
+ dtype: string
45
+ - name: age
46
+ dtype: int64
47
+ - name: marital_status
48
+ dtype: string
49
+ - name: education_level
50
+ dtype: string
51
+ - name: bachelors_field
52
+ dtype: string
53
+ - name: occupation
54
+ dtype: string
55
+ - name: city
56
+ dtype: string
57
+ - name: state
58
+ dtype: string
59
+ - name: zipcode
60
+ dtype: string
61
+ - name: country
62
+ dtype: string
63
+ - name: language
64
+ dtype: string
65
+ - name: persona_region
66
+ dtype: string
67
+ - name: profile_tag
68
+ dtype: string
69
+ splits:
70
+ - name: train
71
+ num_bytes: 70746
72
+ num_examples: 240
73
+ download_size: 70746
74
+ dataset_size: 70746
75
+ configs:
76
+ - config_name: default
77
+ data_files:
78
+ - split: train
79
+ path: data/train-*
80
+ ---
81
+ Persona-German
82
+ =========================================================================
83
+
84
+ <center>
85
+ <p>
86
+ <em>Ein deutsches, synthetisches Persona-Dataset im Stil von Nemotron-Personas-Korea</em><br>
87
+ <em>A German synthetic persona dataset inspired by Nemotron-Personas-Korea</em>
88
+ </p>
89
+ </center>
90
+
91
+ # Dataset Overview
92
+
93
+ Persona-German is a custom synthetic German-language persona dataset built in the style of `nvidia/Nemotron-Personas-Korea`, but adapted for a smaller local dataset release. It contains German persona descriptions together with demographic and geographic context fields so the data can be filtered, conditioned, or expanded for downstream training and evaluation workflows.
94
+
95
+ The current release contains 240 records stored in 2 parquet files. Each record includes multiple persona views such as professional, sports, arts, travel, culinary, and a concise overall persona summary. Additional attributes such as cultural background, skills, hobbies, goals, age, education, occupation, and location are included to make the personas easier to target in synthetic-data or LLM experiments.
96
+
97
+ All records in this dataset are synthetic. The generator was written locally and includes a sentence-level deduplication check so the same sentence is not repeated across the persona text fields more than once.
98
+
99
+ --
100
+
101
+ Persona-German ist ein benutzerdefiniertes synthetisches deutschsprachiges Persona-Dataset, das sich strukturell an `nvidia/Nemotron-Personas-Korea` orientiert, aber als kleinere lokale Veröffentlichung für deutsche Personas umgesetzt wurde. Es enthält deutsche Personabeschreibungen zusammen mit demografischen und geografischen Kontextfeldern, damit die Daten für Training, Evaluation oder konditionierte Generierung gefiltert und erweitert werden können.
102
+
103
+ Die aktuelle Version enthält 240 Datensätze in 2 Parquet-Dateien. Jeder Datensatz umfasst mehrere Persona-Perspektiven wie Beruf, Sport, Kunst, Reisen, Kulinarik sowie eine kurze zusammenfassende Persona. Ergänzende Attribute wie kultureller Hintergrund, Fähigkeiten, Interessen, Ziele, Alter, Bildung, Beruf und Wohnort erleichtern die gezielte Nutzung in synthetischen Datenpipelines und LLM-Workflows.
104
+
105
+ Alle Einträge in diesem Dataset sind künstlich erzeugt. Der lokale Generator enthält außerdem eine Duplikatsprüfung auf Satzebene, sodass derselbe Satz in den Persona-Textfeldern nicht mehrfach vorkommt.
106
+
107
+ # What Is Not In The Dataset
108
+
109
+ This dataset is not an official NVIDIA release and is not a replication of the real `Nemotron-Personas-Korea` production pipeline. It does not include family, housing, district-level, or military-status fields because those columns are not present in the generated parquet schema for this local version.
110
+
111
+ The dataset is also much smaller than the Korean reference dataset and should be treated as a custom seed dataset or scaffold rather than a population-scale release.
112
+
113
+ --
114
+
115
+ Dieses Dataset ist keine offizielle NVIDIA-Veröffentlichung und bildet die Produktionspipeline von `Nemotron-Personas-Korea` nicht vollständig nach. Familien-, Wohnungs-, Militärstatus- oder kreisgenaue Regionalfelder sind hier nicht enthalten, weil diese Spalten im tatsächlich erzeugten Parquet-Schema dieser lokalen Version nicht vorkommen.
116
+
117
+ Außerdem ist das Dataset deutlich kleiner als das koreanische Referenz-Dataset und sollte eher als benutzerdefinierter Startpunkt oder als lokales Persona-Scaffold verstanden werden, nicht als bevölkerungsnahe Großveröffentlichung.
118
+
119
+ # Data Developer
120
+
121
+ Custom local dataset created in this workspace.
122
+
123
+ # Release Date
124
+
125
+ Released locally on April 26, 2026.
126
+
127
+ # Dataset Creation Date
128
+
129
+ April 26, 2026
130
+
131
+ # License / Terms of Use
132
+
133
+ This dataset is licensed under the [Creative Commons Attribution 4.0 International License](https://creativecommons.org/licenses/by/4.0/deed.en) (CC BY 4.0).
134
+
135
+ --
136
+
137
+ Dieses Dataset steht unter der [Creative Commons Attribution 4.0 International License](https://creativecommons.org/licenses/by/4.0/deed.en) (CC BY 4.0).
138
+
139
+ # Use Case
140
+
141
+ This dataset can be used for:
142
+
143
+ * German persona generation
144
+ * Synthetic data bootstrapping
145
+ * LLM fine-tuning experiments
146
+ * Prompt conditioning and retrieval tests
147
+ * Evaluation of persona diversity in German outputs
148
+
149
+ --
150
+
151
+ Dieses Dataset kann unter anderem für folgende Zwecke genutzt werden:
152
+
153
+ * deutsche Persona-Generierung
154
+ * Bootstrapping synthetischer Daten
155
+ * Fine-Tuning-Experimente für LLMs
156
+ * Prompt-Konditionierung und Retrieval-Tests
157
+ * Evaluation der Persona-Vielfalt in deutschen Modellausgaben
158
+
159
+ # Data Version
160
+
161
+ 1.0 (04/26/2026)
162
+
163
+ # Intended Use
164
+
165
+ Persona-German was created as a compact German counterpart inspired by the structure of `Nemotron-Personas-Korea`. It is intended for experimentation, prototyping, and small-scale dataset development where German-language personas with consistent structure are useful.
166
+
167
+ Because the dataset is synthetic and relatively small, it should not be treated as a faithful statistical representation of the entire German population. It is better suited to controlled testing, synthetic prompt generation, schema prototyping, and downstream expansion than to demographic research.
168
+
169
+ The dataset includes only adult personas aged 19 and above.
170
+
171
+ --
172
+
173
+ Persona-German wurde als kompakte deutsche Variante im Stil von `Nemotron-Personas-Korea` erstellt. Es ist für Experimente, Prototyping und kleinere Datensatz-Workflows gedacht, in denen strukturierte deutschsprachige Personas hilfreich sind.
174
+
175
+ Da das Dataset synthetisch und relativ klein ist, sollte es nicht als exakte statistische Abbildung der gesamten deutschen Bevölkerung verstanden werden. Es eignet sich eher für kontrollierte Tests, synthetische Prompt-Erzeugung, Schema-Prototyping und spätere Erweiterungen als für demografische Forschung.
176
+
177
+ Das Dataset enthält ausschließlich erwachsene Personas ab 19 Jahren.
178
+
179
+ # Dataset Details
180
+
181
+ The dataset contains:
182
+
183
+ * 240 records across 2 parquet files
184
+ * 26 fields: 6 core persona fields, 4 additional persona attribute text fields, 15 demographic and contextual fields, and 1 unique identifier-style tag
185
+ * German-language persona descriptions
186
+ * Adult personas only
187
+ * Sentence-level deduplication across persona text fields
188
+
189
+ --
190
+
191
+ Das Dataset enthält:
192
+
193
+ * 240 Datensätze in 2 Parquet-Dateien
194
+ * 26 Felder: 6 zentrale Persona-Felder, 4 ergänzende textuelle Persona-Attributfelder, 15 demografische und kontextuelle Felder sowie 1 identifier-ähnliches Tag
195
+ * deutschsprachige Personabeschreibungen
196
+ * ausschließlich erwachsene Personas
197
+ * Satzebenen-Deduplizierung über die Persona-Textfelder
198
+
199
+ # Schema
200
+
201
+ The dataset includes the following columns:
202
+
203
+ * `uuid`
204
+ * `professional_persona`
205
+ * `sports_persona`
206
+ * `arts_persona`
207
+ * `travel_persona`
208
+ * `culinary_persona`
209
+ * `persona`
210
+ * `cultural_background`
211
+ * `skills_and_expertise`
212
+ * `skills_and_expertise_list`
213
+ * `hobbies_and_interests`
214
+ * `hobbies_and_interests_list`
215
+ * `career_goals_and_ambitions`
216
+ * `sex`
217
+ * `age`
218
+ * `marital_status`
219
+ * `education_level`
220
+ * `bachelors_field`
221
+ * `occupation`
222
+ * `city`
223
+ * `state`
224
+ * `zipcode`
225
+ * `country`
226
+ * `language`
227
+ * `persona_region`
228
+ * `profile_tag`
229
+
230
+ # Seed And Generation Notes
231
+
232
+ The dataset was generated from local templates and controlled attribute pools in `generate_persona_german.py`. It does not claim alignment to official German census distributions. Instead, it follows a structured synthetic generation approach with rotating attribute pools and schema-consistent output suitable for local experimentation.
233
+
234
+ The generator also validates that no full sentence is repeated across the main persona text fields.
235
+
236
+ --
237
+
238
+ Das Dataset wurde aus lokalen Vorlagen und kontrollierten Attribut-Pools in `generate_persona_german.py` erzeugt. Es erhebt keinen Anspruch auf eine exakte Ausrichtung an offiziellen deutschen Bevölkerungsverteilungen. Stattdessen folgt es einem strukturierten synthetischen Generierungsansatz mit rotierenden Attribut-Pools und schema-konsistentem Output für lokale Experimente.
239
+
240
+ Der Generator validiert außerdem, dass in den zentralen Persona-Textfeldern kein vollständiger Satz doppelt vorkommt.
241
+
242
+ # How To Use It
243
+
244
+ ```python
245
+ from datasets import load_dataset
246
+
247
+ persona_german = load_dataset("8F-ai/Persona-German")
248
+ ```
249
+
250
+ If you are loading it locally from files instead of from the Hub:
251
+
252
+ ```python
253
+ from datasets import load_dataset
254
+
255
+ persona_german = load_dataset(
256
+ "parquet",
257
+ data_files={"train": "Persona-German/data/train-*.parquet"},
258
+ )
259
+ ```
260
+
261
+ # Dataset Characterization
262
+
263
+ ## Data Collection Method
264
+
265
+ * Synthetic
266
+
267
+ ## Labeling Method
268
+
269
+ * Not applicable
270
+
271
+ ## Dataset Format
272
+
273
+ * Text
274
+ * Parquet
275
+
276
+ ## Dataset Quantification
277
+
278
+ * Number of records: 240
279
+ * Number of parquet files: 2
280
+ * Total storage size: 70,746 bytes
281
+
282
+ # Ethical Considerations
283
+
284
+ This dataset is fully synthetic, but synthetic personas can still encode design biases from template choices, category pools, and coverage decisions. Users should review the data before using it in production systems or in any application that may affect real people.
285
+
286
+ The dataset should not be interpreted as a statistically complete or authoritative representation of Germany, German regions, or any demographic group.
287
+
288
+ --
289
+
290
+ Dieses Dataset ist vollständig synthetisch, dennoch können auch synthetische Personas Verzerrungen aus Vorlagen, Kategorien und Abdeckungsentscheidungen übernehmen. Vor einem produktiven Einsatz oder vor Anwendungen mit Auswirkungen auf reale Personen sollte der Inhalt daher sorgfältig geprüft werden.
291
+
292
+ Das Dataset sollte nicht als statistisch vollständige oder autoritative Darstellung Deutschlands, deutscher Regionen oder einzelner Bevölkerungsgruppen interpretiert werden.
293
+
294
+ # Citation
295
+
296
+ If you find this dataset useful, you can cite it as:
297
+
298
+ ```bibtex
299
+ @dataset{persona_german_2026,
300
+ title = {Persona-German},
301
+ author = {Ill-Ness and Selen Martin},
302
+ year = {2026},
303
+ month = {April},
304
+ note = {Synthetic German persona dataset inspired by Nemotron-Personas-Korea},
305
+ url = {https://huggingface.co/datasets/8F-ai/Persona-German}
306
+ }
307
+ ```