eduagarcia commited on
Commit
d8ea41a
·
verified ·
1 Parent(s): 92c881f

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +360 -0
README.md ADDED
@@ -0,0 +1,360 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ task_categories:
3
+ - text-generation
4
+ language:
5
+ - en
6
+ - eng
7
+ - pt
8
+ - por
9
+ - es
10
+ - spa
11
+ - fr
12
+ - fra
13
+ - zh
14
+ - cmn
15
+ - ja
16
+ - jpn
17
+ - de
18
+ - deu
19
+ - ru
20
+ - rus
21
+ - ar
22
+ - arb
23
+ - it
24
+ - ita
25
+ - pl
26
+ - pol
27
+ - ko
28
+ - kor
29
+ - code
30
+ multilinguality:
31
+ - multilingual
32
+ pretty_name: Multilingual Tokenizer Benchmark
33
+ size_categories:
34
+ - 100K<n<1M
35
+ configs:
36
+ - config_name: default
37
+ data_files:
38
+ - split: train
39
+ path: "**/*.jsonl.gz"
40
+ - config_name: en
41
+ data_files:
42
+ - split: train
43
+ path: en/*.jsonl.gz
44
+ - config_name: pt
45
+ data_files:
46
+ - split: train
47
+ path: pt/*.jsonl.gz
48
+ - config_name: es
49
+ data_files:
50
+ - split: train
51
+ path: es/*.jsonl.gz
52
+ - config_name: fr
53
+ data_files:
54
+ - split: train
55
+ path: fr/*.jsonl.gz
56
+ - config_name: zh
57
+ data_files:
58
+ - split: train
59
+ path: zh/*.jsonl.gz
60
+ - config_name: ja
61
+ data_files:
62
+ - split: train
63
+ path: ja/*.jsonl.gz
64
+ - config_name: de
65
+ data_files:
66
+ - split: train
67
+ path: de/*.jsonl.gz
68
+ - config_name: ru
69
+ data_files:
70
+ - split: train
71
+ path: ru/*.jsonl.gz
72
+ - config_name: ar
73
+ data_files:
74
+ - split: train
75
+ path: ar/*.jsonl.gz
76
+ - config_name: it
77
+ data_files:
78
+ - split: train
79
+ path: it/*.jsonl.gz
80
+ - config_name: pl
81
+ data_files:
82
+ - split: train
83
+ path: pl/*.jsonl.gz
84
+ - config_name: ko
85
+ data_files:
86
+ - split: train
87
+ path: ko/*.jsonl.gz
88
+ - config_name: code
89
+ data_files:
90
+ - split: train
91
+ path: code/*.jsonl.gz
92
+ source_datasets:
93
+ - bigcode/the-stack-smol
94
+ - badranx/opus_raw
95
+ - HuggingFaceFW/fineweb-2
96
+ - rntc/pubmed_preprocess
97
+ - joelniklaus/Multi_Legal_Pile
98
+ - wikimedia/wikipedia
99
+ - almanach/HALvest
100
+ - premio-ai/TheArabicPile_Medical
101
+ - ImruQays/Rasaif-Classical-Arabic-English-Parallel-texts
102
+ - bigscience-data/roots_ar_ted_talks_iwslt
103
+ - premio-ai/TheArabicPile_Lyrics
104
+ - AIR-Bench/qa_news_ar
105
+ - premio-ai/TheArabicPile_Poetry
106
+ - bigscience-data/roots_ar_wikipedia
107
+ - bigcode/starcoderdata
108
+ - AIR-Bench/qa_news_de
109
+ - orionweller/books_mds_incremental
110
+ - ElKulako/stocktwits-emoji
111
+ - Orion-zhen/dpo-codealpaca-emoji
112
+ - Orion-zhen/dpo-mathinstuct-emoji
113
+ - Orion-zhen/dpo-physical-reasoning-emoji
114
+ - AIR-Bench/qa_finance_en
115
+ - bigscience-data/roots_en_ted_talks_iwslt
116
+ - orionweller/cc_news_mds_incremental
117
+ - orionweller/arxiv_mds_incremental
118
+ - orionweller/pes2o_mds_incremental
119
+ - orionweller/reddit_mds_incremental
120
+ - HuggingFaceFW/fineweb
121
+ - orionweller/wikipedia_mds_incremental
122
+ - jorgeortizfuentes/spanish_books
123
+ - bigscience-data/roots_es_ted_talks_iwslt
124
+ - LeoCordoba/CC-NEWS-ES
125
+ - eduagarcia/scielo_abstracts
126
+ - bigscience-data/roots_es_wikipedia
127
+ - AIR-Bench/qa_finance_fr
128
+ - bigscience-data/roots_fr_ted_talks_iwslt
129
+ - AIR-Bench/qa_news_fr
130
+ - kamarko/ccnews-french-subset
131
+ - bigscience-data/roots_fr_wikipedia
132
+ - HiTZ/Multilingual-Medical-Corpus
133
+ - globis-university/aozorabunko-clean
134
+ - p1atdev/modern_haiku
135
+ - geniacllm/hanrei
136
+ - kajuma/CC-news-2024-July-October-cleaned
137
+ - kunishou/J-ResearchCorpus
138
+ - BCCard/BCCard-Finance-Kor-QnA
139
+ - jihye-moon/klac_legal_aid_counseling
140
+ - joonhok-exo-ai/korean_law_open_data_precedents
141
+ - AIR-Bench/qa_news_ko
142
+ - daekeun-ml/naver-news-summarization-ko
143
+ - WiktorS/polish-news
144
+ - winterkitsune/elka-pl-news
145
+ - eduagarcia/pd_books_pt
146
+ - eduagarcia/LegalPT_dedup
147
+ - bigscience-data/roots_pt_ted_talks_iwslt
148
+ - eduagarcia/cc_news_pt_v2
149
+ - eduagarcia/scielo_papers
150
+ - carolina-c4ai/corpus-carolina
151
+ - tallesl/quinquilharia
152
+ - bigscience-data/roots_pt_wikipedia
153
+ - blinoff/medical_qa_ru_data
154
+ - AIR-Bench/qa_news_ru
155
+ - IlyaGusev/ru_news
156
+ - mlsa-iai-msu-lab/ru_sci_bench
157
+ - DavidLanz/medical_pretrain
158
+ - Orion-zhen/dpo-emoji-zh
159
+ - Orion-zhen/dpo-ruozhiba-emoji
160
+ - shareAI/DPO-zh-en-emoji
161
+ - Duxiaoman-DI/FinCorpus
162
+ - TigerResearch/tigerbot-law-plugin
163
+ - bigscience-data/roots_zh_ted_talks_iwslt
164
+ - AIR-Bench/qa_news_zh
165
+ - liswei/news-collection-zhtw
166
+ - bigscience-data/roots_zh-cn_wikipedia
167
+ ---
168
+ # Multilingual Tokenizer Benchmark
169
+
170
+
171
+ More details of each subset like word count, character count, original sources, etc, can be found in the [dataset_meta.yaml](dataset_meta.yaml) file in the repository root.
172
+
173
+
174
+ ### Natural language word count functions
175
+ Download spacy models
176
+ ```bash
177
+ python -m spacy download ko_core_news_sm
178
+ python -m spacy download ja_core_news_sm
179
+ python -m spacy download zh_core_web_sm
180
+ ```
181
+
182
+ ```python
183
+ import nltk
184
+ nltk.download('punkt_tab')
185
+ from nltk import word_tokenize
186
+ import spacy
187
+
188
+ langs_nltk = {
189
+ 'en': 'english',
190
+ 'pt': 'portuguese',
191
+ 'it': 'italian',
192
+ 'pl': 'polish',
193
+ 'de': 'german',
194
+ 'es': 'spanish',
195
+ 'fr': 'french',
196
+ 'ru': 'russian',
197
+ 'ar': 'english' # Arabic will use the default config from nltk.word_tokenize
198
+ }
199
+
200
+ langs_spacy = {
201
+ 'ja': spacy.load('ja_core_news_sm', disable=['parser', 'ner', 'lemmatizer', 'tagger', 'attribute_ruler']),
202
+ 'zh': spacy.load('zh_core_web_sm', disable=['parser', 'ner', 'lemmatizer', 'tagger', 'attribute_ruler']),
203
+ 'ko': spacy.load('ko_core_news_sm', disable=['parser', 'ner', 'lemmatizer', 'tagger', 'attribute_ruler']),
204
+ }
205
+
206
+ def word_tokenize_spacy(text, model):
207
+ doc = model(text)
208
+ return [token.text for token in doc if not token.is_space]
209
+
210
+ #### Usage Example ####
211
+ word_count_pt = len(word_tokenize("Olá Mundo.", langs_nltk['pt']))
212
+ word_count_ko = len(word_tokenize_spacy("이것은 한국어 문장입니다.", langs_spacy['ko']))
213
+ ```
214
+
215
+ ### Code word count function
216
+ ```python
217
+ import nltk
218
+ nltk.download('punkt_tab')
219
+ from nltk import word_tokenize
220
+
221
+ from pygments import lex
222
+ from pygments.lexers import get_lexer_by_name
223
+ from pygments.token import Token
224
+
225
+ NATURAL_LANGUAGE_PARENT_TOKEN_TYPES = {
226
+ Token.Comment,
227
+ Token.Literal.String,
228
+ Token.Text,
229
+ Token.Generic.Heading, # For Markdown,
230
+ Token.Generic.Subheading, # For Markdown,
231
+ Token.Generic.Emph, # For Markdown *italic*
232
+ Token.Generic.Strong, # For Markdown **bold**
233
+ Token.Generic.Output, # For shell output,
234
+ Token.Generic.Error, # Error messages
235
+ Token.Generic.Traceback, # Traceback messages
236
+ }
237
+
238
+ def _is_natural_language_like(tok_type) -> bool:
239
+ for nl_parent_type in NATURAL_LANGUAGE_PARENT_TOKEN_TYPES:
240
+ if tok_type in nl_parent_type: # This checks if tok_type is nl_parent_type OR a child
241
+ return True
242
+ return False
243
+
244
+ def pygments_tokenize(code: str, language: str) -> list[str]:
245
+ lexer = get_lexer_by_name(language, stripnl=True, stripall=True)
246
+ tokens = lex(code, lexer)
247
+
248
+ result = []
249
+
250
+ for tok_type, value in tokens:
251
+ if _is_natural_language_like(tok_type):
252
+ # Recursively tokenize comment or string content. Use the default english tokenizer.
253
+ result.extend(word_tokenize(value))
254
+ else:
255
+ # Tokenize symbol/keyword/identifier at surface level
256
+ result.append(value)
257
+
258
+ return [t for t in result if t.strip() != '']
259
+
260
+ #### Usage Example ####
261
+ word_count_json = len(pygments_tokenize('{"hello": 1}', 'json'))
262
+ word_count_python = len(pygments_tokenize("lambda x: print('hello')", "python"))
263
+ ```
264
+
265
+ ### Per language stats
266
+
267
+ | Code | Language | Num. of Subsets | Num. of Docs | Num. of Characters | Num. of Words | Data Types |
268
+ |:-------|:----------------------|------------------:|---------------:|---------------------:|----------------:|:---------------------------------------------------------------------------------------------------------------------------------|
269
+ | code | Programming Languages | 18 | 19820 | 90121858 | 18976907 | C, C++, CSS, Dockerfile, Go, HTML, JSON, Java, JavaScript, Lua, Markdown, PHP, Python, Rust, SQL, Shell, TeX, TypeScript |
270
+ | en | English | 12 | 38155 | 61999845 | 12361906 | Biomedical, Books, Emoji Heavy, Finance, Legal, Live Speech, News, Scientific Papers, Social Networks, Subtitles, Web, Wikipedia |
271
+ | pt | Portuguese | 10 | 30528 | 50195889 | 9603267 | Biomedical, Books, Legal, Live Speech, News, Scientific Papers, Social Networks, Subtitles, Web, Wikipedia |
272
+ | ar | Arabic | 9 | 50539 | 45233908 | 8911565 | Biomedical, Books, Live Speech, Lyrics, News, Poetry, Subtitles, Web, Wikipedia |
273
+ | es | Spanish | 9 | 25315 | 45168212 | 8420656 | Biomedical, Books, Legal, Live Speech, News, Scientific Papers, Subtitles, Web, Wikipedia |
274
+ | fr | French | 9 | 26934 | 45185903 | 8594454 | Biomedical, Finance, Legal, Live Speech, News, Scientific Papers, Subtitles, Web, Wikipedia |
275
+ | zh | Chinese | 9 | 47627 | 43209186 | 24276657 | Biomedical, Emoji Heavy, Finance, Legal, Live Speech, News, Subtitles, Web, Wikipedia |
276
+ | ja | Japanese | 8 | 35950 | 35945203 | 19903904 | Books, Haiku, Legal, News, Scientific Papers, Subtitles, Web, Wikipedia |
277
+ | de | German | 7 | 19740 | 35215590 | 5876322 | Biomedical, Legal, News, Scientific Papers, Subtitles, Web, Wikipedia |
278
+ | it | Italian | 6 | 29414 | 30160296 | 5462293 | Biomedical, Legal, Scientific Papers, Subtitles, Web, Wikipedia |
279
+ | ko | Korean | 6 | 28795 | 30376957 | 8362948 | Finance, Legal, News, Subtitles, Web, Wikipedia |
280
+ | ru | Russian | 6 | 24329 | 30360746 | 5250785 | Biomedical, News, Scientific Papers, Subtitles, Web, Wikipedia |
281
+ | pl | Polish | 5 | 15763 | 25168261 | 4131439 | Legal, News, Subtitles, Web, Wikipedia |
282
+
283
+ ### Sourced datasets:
284
+
285
+ | | HF Dataset | Num. Subsets | Languages | Data Types |
286
+ |---:|:---------------------------------------------------------------------------------------------------------------------------------------------------|--------:|:-----------------------------------------------|:-------------------------------------------------------------------------------------------------------------------|
287
+ | 0 | [bigcode/the-stack-smol](https://huggingface.co/datasets/bigcode/the-stack-smol) | 17 | code | C, C++, CSS, Dockerfile, Go, HTML, Java, JavaScript, Lua, Markdown, PHP, Python, Rust, SQL, Shell, TeX, TypeScript |
288
+ | 1 | [badranx/opus_raw](https://huggingface.co/datasets/badranx/opus_raw) | 13 | ar, de, en, es, fr, it, ja, ko, pl, pt, ru, zh | Subtitles |
289
+ | 2 | [HuggingFaceFW/fineweb-2](https://huggingface.co/datasets/HuggingFaceFW/fineweb-2) | 11 | ar, de, es, fr, it, ja, ko, pl, pt, ru, zh | Web |
290
+ | 3 | [rntc/pubmed_preprocess](https://huggingface.co/datasets/rntc/pubmed_preprocess) | 8 | de, en, es, fr, it, pt, ru, zh | Biomedical |
291
+ | 4 | [joelniklaus/Multi_Legal_Pile](https://huggingface.co/datasets/joelniklaus/Multi_Legal_Pile) | 6 | de, en, es, fr, it, pl | Legal |
292
+ | 5 | [wikimedia/wikipedia](https://huggingface.co/datasets/wikimedia/wikipedia) | 6 | de, it, ja, ko, pl, ru | Wikipedia |
293
+ | 6 | [almanach/HALvest](https://huggingface.co/datasets/almanach/HALvest) | 5 | de, es, fr, it, ru | Scientific Papers |
294
+ | 7 | [premio-ai/TheArabicPile_Medical](https://huggingface.co/datasets/premio-ai/TheArabicPile_Medical) | 1 | ar | Biomedical |
295
+ | 8 | [ImruQays/Rasaif-Classical-Arabic-English-Parallel-texts](https://huggingface.co/datasets/ImruQays/Rasaif-Classical-Arabic-English-Parallel-texts) | 1 | ar | Books |
296
+ | 9 | [bigscience-data/roots_ar_ted_talks_iwslt](https://huggingface.co/datasets/bigscience-data/roots_ar_ted_talks_iwslt) | 1 | ar | Live Speech |
297
+ | 10 | [premio-ai/TheArabicPile_Lyrics](https://huggingface.co/datasets/premio-ai/TheArabicPile_Lyrics) | 1 | ar | Lyrics |
298
+ | 11 | [AIR-Bench/qa_news_ar](https://huggingface.co/datasets/AIR-Bench/qa_news_ar) | 1 | ar | News |
299
+ | 12 | [premio-ai/TheArabicPile_Poetry](https://huggingface.co/datasets/premio-ai/TheArabicPile_Poetry) | 1 | ar | Poetry |
300
+ | 13 | [bigscience-data/roots_ar_wikipedia](https://huggingface.co/datasets/bigscience-data/roots_ar_wikipedia) | 1 | ar | Wikipedia |
301
+ | 14 | [bigcode/starcoderdata](https://huggingface.co/datasets/bigcode/starcoderdata) | 1 | code | JSON |
302
+ | 15 | [AIR-Bench/qa_news_de](https://huggingface.co/datasets/AIR-Bench/qa_news_de) | 1 | de | News |
303
+ | 16 | [orionweller/books_mds_incremental](https://huggingface.co/datasets/orionweller/books_mds_incremental) | 1 | en | Books |
304
+ | 17 | [ElKulako/stocktwits-emoji](https://huggingface.co/datasets/ElKulako/stocktwits-emoji) | 1 | en | Emoji Heavy |
305
+ | 18 | [Orion-zhen/dpo-codealpaca-emoji](https://huggingface.co/datasets/Orion-zhen/dpo-codealpaca-emoji) | 1 | en | Emoji Heavy |
306
+ | 19 | [Orion-zhen/dpo-mathinstuct-emoji](https://huggingface.co/datasets/Orion-zhen/dpo-mathinstuct-emoji) | 1 | en | Emoji Heavy |
307
+ | 20 | [Orion-zhen/dpo-physical-reasoning-emoji](https://huggingface.co/datasets/Orion-zhen/dpo-physical-reasoning-emoji) | 1 | en | Emoji Heavy |
308
+ | 21 | [AIR-Bench/qa_finance_en](https://huggingface.co/datasets/AIR-Bench/qa_finance_en) | 1 | en | Finance |
309
+ | 22 | [bigscience-data/roots_en_ted_talks_iwslt](https://huggingface.co/datasets/bigscience-data/roots_en_ted_talks_iwslt) | 1 | en | Live Speech |
310
+ | 23 | [orionweller/cc_news_mds_incremental](https://huggingface.co/datasets/orionweller/cc_news_mds_incremental) | 1 | en | News |
311
+ | 24 | [orionweller/arxiv_mds_incremental](https://huggingface.co/datasets/orionweller/arxiv_mds_incremental) | 1 | en | Scientific Papers |
312
+ | 25 | [orionweller/pes2o_mds_incremental](https://huggingface.co/datasets/orionweller/pes2o_mds_incremental) | 1 | en | Scientific Papers |
313
+ | 26 | [orionweller/reddit_mds_incremental](https://huggingface.co/datasets/orionweller/reddit_mds_incremental) | 1 | en | Social Networks |
314
+ | 27 | [HuggingFaceFW/fineweb](https://huggingface.co/datasets/HuggingFaceFW/fineweb) | 1 | en | Web |
315
+ | 28 | [orionweller/wikipedia_mds_incremental](https://huggingface.co/datasets/orionweller/wikipedia_mds_incremental) | 1 | en | Wikipedia |
316
+ | 29 | [jorgeortizfuentes/spanish_books](https://huggingface.co/datasets/jorgeortizfuentes/spanish_books) | 1 | es | Books |
317
+ | 30 | [bigscience-data/roots_es_ted_talks_iwslt](https://huggingface.co/datasets/bigscience-data/roots_es_ted_talks_iwslt) | 1 | es | Live Speech |
318
+ | 31 | [LeoCordoba/CC-NEWS-ES](https://huggingface.co/datasets/LeoCordoba/CC-NEWS-ES) | 1 | es | News |
319
+ | 32 | [eduagarcia/scielo_abstracts](https://huggingface.co/datasets/eduagarcia/scielo_abstracts) | 1 | es | Scientific Papers |
320
+ | 33 | [bigscience-data/roots_es_wikipedia](https://huggingface.co/datasets/bigscience-data/roots_es_wikipedia) | 1 | es | Wikipedia |
321
+ | 34 | [AIR-Bench/qa_finance_fr](https://huggingface.co/datasets/AIR-Bench/qa_finance_fr) | 1 | fr | Finance |
322
+ | 35 | [bigscience-data/roots_fr_ted_talks_iwslt](https://huggingface.co/datasets/bigscience-data/roots_fr_ted_talks_iwslt) | 1 | fr | Live Speech |
323
+ | 36 | [AIR-Bench/qa_news_fr](https://huggingface.co/datasets/AIR-Bench/qa_news_fr) | 1 | fr | News |
324
+ | 37 | [kamarko/ccnews-french-subset](https://huggingface.co/datasets/kamarko/ccnews-french-subset) | 1 | fr | News |
325
+ | 38 | [bigscience-data/roots_fr_wikipedia](https://huggingface.co/datasets/bigscience-data/roots_fr_wikipedia) | 1 | fr | Wikipedia |
326
+ | 39 | [HiTZ/Multilingual-Medical-Corpus](https://huggingface.co/datasets/HiTZ/Multilingual-Medical-Corpus) | 1 | it | Biomedical |
327
+ | 40 | [globis-university/aozorabunko-clean](https://huggingface.co/datasets/globis-university/aozorabunko-clean) | 1 | ja | Books |
328
+ | 41 | [p1atdev/modern_haiku](https://huggingface.co/datasets/p1atdev/modern_haiku) | 1 | ja | Haiku |
329
+ | 42 | [geniacllm/hanrei](https://huggingface.co/datasets/geniacllm/hanrei) | 1 | ja | Legal |
330
+ | 43 | [kajuma/CC-news-2024-July-October-cleaned](https://huggingface.co/datasets/kajuma/CC-news-2024-July-October-cleaned) | 1 | ja | News |
331
+ | 44 | [kunishou/J-ResearchCorpus](https://huggingface.co/datasets/kunishou/J-ResearchCorpus) | 1 | ja | Scientific Papers |
332
+ | 45 | [BCCard/BCCard-Finance-Kor-QnA](https://huggingface.co/datasets/BCCard/BCCard-Finance-Kor-QnA) | 1 | ko | Finance |
333
+ | 46 | [jihye-moon/klac_legal_aid_counseling](https://huggingface.co/datasets/jihye-moon/klac_legal_aid_counseling) | 1 | ko | Legal |
334
+ | 47 | [joonhok-exo-ai/korean_law_open_data_precedents](https://huggingface.co/datasets/joonhok-exo-ai/korean_law_open_data_precedents) | 1 | ko | Legal |
335
+ | 48 | [AIR-Bench/qa_news_ko](https://huggingface.co/datasets/AIR-Bench/qa_news_ko) | 1 | ko | News |
336
+ | 49 | [daekeun-ml/naver-news-summarization-ko](https://huggingface.co/datasets/daekeun-ml/naver-news-summarization-ko) | 1 | ko | News |
337
+ | 50 | [WiktorS/polish-news](https://huggingface.co/datasets/WiktorS/polish-news) | 1 | pl | News |
338
+ | 51 | [winterkitsune/elka-pl-news](https://huggingface.co/datasets/winterkitsune/elka-pl-news) | 1 | pl | News |
339
+ | 52 | [eduagarcia/pd_books_pt](https://huggingface.co/datasets/eduagarcia/pd_books_pt) | 1 | pt | Books |
340
+ | 53 | [eduagarcia/LegalPT_dedup](https://huggingface.co/datasets/eduagarcia/LegalPT_dedup) | 1 | pt | Legal |
341
+ | 54 | [bigscience-data/roots_pt_ted_talks_iwslt](https://huggingface.co/datasets/bigscience-data/roots_pt_ted_talks_iwslt) | 1 | pt | Live Speech |
342
+ | 55 | [eduagarcia/cc_news_pt_v2](https://huggingface.co/datasets/eduagarcia/cc_news_pt_v2) | 1 | pt | News |
343
+ | 56 | [eduagarcia/scielo_papers](https://huggingface.co/datasets/eduagarcia/scielo_papers) | 1 | pt | Scientific Papers |
344
+ | 57 | [carolina-c4ai/corpus-carolina](https://huggingface.co/datasets/carolina-c4ai/corpus-carolina) | 1 | pt | Social Networks |
345
+ | 58 | [tallesl/quinquilharia](https://huggingface.co/datasets/tallesl/quinquilharia) | 1 | pt | Social Networks |
346
+ | 59 | [bigscience-data/roots_pt_wikipedia](https://huggingface.co/datasets/bigscience-data/roots_pt_wikipedia) | 1 | pt | Wikipedia |
347
+ | 60 | [blinoff/medical_qa_ru_data](https://huggingface.co/datasets/blinoff/medical_qa_ru_data) | 1 | ru | Biomedical |
348
+ | 61 | [AIR-Bench/qa_news_ru](https://huggingface.co/datasets/AIR-Bench/qa_news_ru) | 1 | ru | News |
349
+ | 62 | [IlyaGusev/ru_news](https://huggingface.co/datasets/IlyaGusev/ru_news) | 1 | ru | News |
350
+ | 63 | [mlsa-iai-msu-lab/ru_sci_bench](https://huggingface.co/datasets/mlsa-iai-msu-lab/ru_sci_bench) | 1 | ru | Scientific Papers |
351
+ | 64 | [DavidLanz/medical_pretrain](https://huggingface.co/datasets/DavidLanz/medical_pretrain) | 1 | zh | Biomedical |
352
+ | 65 | [Orion-zhen/dpo-emoji-zh](https://huggingface.co/datasets/Orion-zhen/dpo-emoji-zh) | 1 | zh | Emoji Heavy |
353
+ | 66 | [Orion-zhen/dpo-ruozhiba-emoji](https://huggingface.co/datasets/Orion-zhen/dpo-ruozhiba-emoji) | 1 | zh | Emoji Heavy |
354
+ | 67 | [shareAI/DPO-zh-en-emoji](https://huggingface.co/datasets/shareAI/DPO-zh-en-emoji) | 1 | zh | Emoji Heavy |
355
+ | 68 | [Duxiaoman-DI/FinCorpus](https://huggingface.co/datasets/Duxiaoman-DI/FinCorpus) | 1 | zh | Finance |
356
+ | 69 | [TigerResearch/tigerbot-law-plugin](https://huggingface.co/datasets/TigerResearch/tigerbot-law-plugin) | 1 | zh | Legal |
357
+ | 70 | [bigscience-data/roots_zh_ted_talks_iwslt](https://huggingface.co/datasets/bigscience-data/roots_zh_ted_talks_iwslt) | 1 | zh | Live Speech |
358
+ | 71 | [AIR-Bench/qa_news_zh](https://huggingface.co/datasets/AIR-Bench/qa_news_zh) | 1 | zh | News |
359
+ | 72 | [liswei/news-collection-zhtw](https://huggingface.co/datasets/liswei/news-collection-zhtw) | 1 | zh | News |
360
+ | 73 | [bigscience-data/roots_zh-cn_wikipedia](https://huggingface.co/datasets/bigscience-data/roots_zh-cn_wikipedia) | 1 | zh | Wikipedia |