v3.2, add subtitles for all langs and social for pt

ar/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7aba37d0ada1129d4e41ca58f9e540294a921076d4543861f71f5c69cf280779
+size 4725115

dataset_meta.yaml

@@ -1,9 +1,9 @@
 ar:
   name: ar
   pretty_name: Arabic
-  n_docs: 64917
-  word_count: 7821175
-  char_count: 40023302
+  n_docs: 261134
+  word_count: 8911565
+  char_count: 45023313
   sources:
   - filepath: ar/biomedical.jsonl.gz
     name: biomedical
@@ -53,6 +53,14 @@ ar:
     char_count: 5014133
     hf_datasets:
     - bigscience-data/roots_ar_ted_talks_iwslt
+  - filepath: ar/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 196217
+    word_count: 1090390
+    char_count: 5000011
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: ar/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -238,9 +246,9 @@ code:
 de:
   name: de
   pretty_name: German
-  n_docs: 10301
-  word_count: 4759253
-  char_count: 30064541
+  n_docs: 170757
+  word_count: 5876322
+  char_count: 35064573
   sources:
   - filepath: de/biomedical.jsonl.gz
     name: biomedical
@@ -274,6 +282,14 @@ de:
     char_count: 5035408
     hf_datasets:
     - almanach/HALvest
+  - filepath: de/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 160456
+    word_count: 1117069
+    char_count: 5000032
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: de/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -307,9 +323,9 @@ de:
 en:
   name: en
   pretty_name: English
-  n_docs: 78465
-  word_count: 11172683
-  char_count: 56817762
+  n_docs: 220206
+  word_count: 12361906
+  char_count: 61817794
   sources:
   - filepath: en/biomedical.jsonl.gz
     name: biomedical
@@ -373,7 +389,7 @@ en:
     - orionweller/arxiv_mds_incremental
   - filepath: en/social.jsonl.gz
     name: social
-    pretty_name: Social Media
+    pretty_name: Social Networks
     n_docs: 4882
     word_count: 1049919
     char_count: 5000428
@@ -387,6 +403,14 @@ en:
     char_count: 5014876
     hf_datasets:
     - bigscience-data/roots_en_ted_talks_iwslt
+  - filepath: en/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 141741
+    word_count: 1189223
+    char_count: 5000032
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: en/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -407,9 +431,9 @@ en:
 es:
   name: es
   pretty_name: Spanish
-  n_docs: 15852
-  word_count: 7319632
-  char_count: 40045175
+  n_docs: 148323
+  word_count: 8420656
+  char_count: 45045204
   sources:
   - filepath: es/biomedical.jsonl.gz
     name: biomedical
@@ -460,6 +484,14 @@ es:
     char_count: 5009865
     hf_datasets:
     - bigscience-data/roots_es_ted_talks_iwslt
+  - filepath: es/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 132471
+    word_count: 1101024
+    char_count: 5000029
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: es/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -493,9 +525,9 @@ es:
 fr:
   name: fr
   pretty_name: French
-  n_docs: 17250
-  word_count: 7465342
-  char_count: 40040646
+  n_docs: 172179
+  word_count: 8594454
+  char_count: 45040658
   sources:
   - filepath: fr/biomedical.jsonl.gz
     name: biomedical
@@ -546,6 +578,14 @@ fr:
     char_count: 5006652
     hf_datasets:
     - bigscience-data/roots_fr_ted_talks_iwslt
+  - filepath: fr/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 154929
+    word_count: 1129112
+    char_count: 5000012
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: fr/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -579,9 +619,9 @@ fr:
 it:
   name: it
   pretty_name: Italian
-  n_docs: 19500
-  word_count: 4360165
-  char_count: 25021504
+  n_docs: 168201
+  word_count: 5462293
+  char_count: 30021509
   sources:
   - filepath: it/biomedical.jsonl.gz
     name: biomedical
@@ -608,6 +648,14 @@ it:
     char_count: 5002831
     hf_datasets:
     - almanach/HALvest
+  - filepath: it/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 148701
+    word_count: 1102128
+    char_count: 5000005
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: it/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -641,9 +689,9 @@ it:
 jp:
   name: jp
   pretty_name: Japanese
-  n_docs: 80861
-  word_count: 17014203
-  char_count: 30514472
+  n_docs: 466676
+  word_count: 19903904
+  char_count: 35514477
   sources:
   - filepath: jp/books.jsonl.gz
     name: books
@@ -685,6 +733,14 @@ jp:
     char_count: 5002167
     hf_datasets:
     - kunishou/J-ResearchCorpus
+  - filepath: jp/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 385815
+    word_count: 2889701
+    char_count: 5000005
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: jp/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -718,9 +774,9 @@ jp:
 ko:
   name: ko
   pretty_name: Korean
-  n_docs: 34207
-  word_count: 6799214
-  char_count: 25023091
+  n_docs: 382647
+  word_count: 8362948
+  char_count: 30023105
   sources:
   - filepath: ko/finance.jsonl.gz
     name: finance
@@ -748,6 +804,14 @@ ko:
     hf_datasets:
     - AIR-Bench/qa_news_ko
     - daekeun-ml/naver-news-summarization-ko
+  - filepath: ko/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 348440
+    word_count: 1563734
+    char_count: 5000014
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: ko/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -781,9 +845,9 @@ ko:
 pl:
   name: pl
   pretty_name: Polish
-  n_docs: 6005
-  word_count: 3077830
-  char_count: 20012011
+  n_docs: 171889
+  word_count: 4131439
+  char_count: 25012135
   sources:
   - filepath: pl/legal.jsonl.gz
     name: legal
@@ -802,6 +866,14 @@ pl:
     hf_datasets:
     - WiktorS/polish-news
     - winterkitsune/elka-pl-news
+  - filepath: pl/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 165884
+    word_count: 1053609
+    char_count: 5000124
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: pl/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -835,9 +907,9 @@ pl:
 pt:
   name: pt
   pretty_name: Portuguese
-  n_docs: 13410
-  word_count: 7417205
-  char_count: 40031524
+  n_docs: 188503
+  word_count: 9603267
+  char_count: 50037914
   sources:
   - filepath: pt/biomedical.jsonl.gz
     name: biomedical
@@ -879,6 +951,15 @@ pt:
     char_count: 5008389
     hf_datasets:
     - eduagarcia/scielo_papers
+  - filepath: pt/social.jsonl.gz
+    name: social
+    pretty_name: Social Networks
+    n_docs: 7244
+    word_count: 1034347
+    char_count: 5006373
+    hf_datasets:
+    - carolina-c4ai/corpus-carolina
+    - tallesl/quinquilharia
   - filepath: pt/speech_text.jsonl.gz
     name: speech_text
     pretty_name: Live Speech
@@ -887,6 +968,15 @@ pt:
     char_count: 5005175
     hf_datasets:
     - bigscience-data/roots_pt_ted_talks_iwslt
+  - filepath: pt/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 167849
+    word_count: 1151715
+    char_count: 5000017
+    hf_datasets:
+    - badranx/opus_raw
+    - badranx/opus_raw
   - filepath: pt/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -920,9 +1010,9 @@ pt:
 ru:
   name: ru
   pretty_name: Russian
-  n_docs: 14595
-  word_count: 4114208
-  char_count: 25195216
+  n_docs: 189804
+  word_count: 5250785
+  char_count: 30195271
   sources:
   - filepath: ru/biomedical.jsonl.gz
     name: biomedical
@@ -951,6 +1041,14 @@ ru:
     hf_datasets:
     - almanach/HALvest
     - mlsa-iai-msu-lab/ru_sci_bench
+  - filepath: ru/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 175209
+    word_count: 1136577
+    char_count: 5000055
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: ru/web.jsonl.gz
     name: web
     pretty_name: Web
@@ -984,9 +1082,9 @@ ru:
 zh:
   name: zh
   pretty_name: Chinese
-  n_docs: 64921
-  word_count: 21382011
-  char_count: 37788317
+  n_docs: 468496
+  word_count: 24276657
+  char_count: 42788317
   sources:
   - filepath: zh/biomedical.jsonl.gz
     name: biomedical
@@ -1040,6 +1138,14 @@ zh:
     char_count: 5002517
     hf_datasets:
     - bigscience-data/roots_zh_ted_talks_iwslt
+  - filepath: zh/subtitles.jsonl.gz
+    name: subtitles
+    pretty_name: Subtitles
+    n_docs: 403575
+    word_count: 2894646
+    char_count: 5000000
+    hf_datasets:
+    - badranx/opus_raw
   - filepath: zh/web.jsonl.gz
     name: web
     pretty_name: Web

de/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:875e689ad4ecdb13d33dd52027651308f04c5321211b5235425e70a5a18c486b
+size 3170428

en/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:22f447be0c239e49e5332ae7de5166482f2c084ec2f162db8d74adbbee6e5f15
+size 3270084

es/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0e703f531cee9810e53ebc272ffe8802ffce471e06a6bafc531004bbfed5105e
+size 3284307

fr/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fcdb9b59367f005a35bb9dbdc66de11bc86b0989303db00a1de56d65a2da9e9f
+size 3199457

it/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a98b7034eb646f7bbe638b16e1a36dd8c732c6e76361c84678e1a9a2e546e23
+size 3200864

jp/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ef00743329b29f9db007a3285e87c4c7f7c2ba8ee43e5af8af1236af74de90b
+size 8694047

ko/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6c4d53e919964bad9858cbdd41458ab5a6efeb627ecd66e50279742e7d24847
+size 7620926

pl/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:770e5b7904fc6ce037a6649bf827acf7c698bc3cb6b134c6d5c50b1aa17f0165
+size 3556127

pt/social.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dd210398d0e3346dc3cae08b0eb8940a191058f489e792a060f5c54bcd8b4478
+size 2237882

pt/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:657eca02b5c446f3cad4d3068dda2e0001f17145a0f26dac13dd10a8f8f4ef45
+size 3332816

ru/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:560467dca8b14147ee43eee1d7d86e028df2bce10128d893478537828f019ba9
+size 4331627

zh/subtitles.jsonl.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:34fabcd61fb1f1fb40dbe1a92ccabe2845d6443d6922db936a3b9d8c7c6f247c
+size 9598235