Add abt_Latn shard 0: 100,437 sentences, 4,098 paragraphs, 17,810 words, 577,835 ngrams

README.md

@@ -888,6 +888,22 @@ configs:
     data_files:
       - split: train
         path: data/ngrams/tlh_Latn/*.parquet
+  - config_name: sentences-abt_Latn
+    data_files:
+      - split: train
+        path: data/sentences/abt_Latn/*.parquet
+  - config_name: paragraphs-abt_Latn
+    data_files:
+      - split: train
+        path: data/paragraphs/abt_Latn/*.parquet
+  - config_name: words-abt_Latn
+    data_files:
+      - split: train
+        path: data/words/abt_Latn/*.parquet
+  - config_name: ngrams-abt_Latn
+    data_files:
+      - split: train
+        path: data/ngrams/abt_Latn/*.parquet
   - config_name: sentences-gil_Latn
     data_files:
       - split: train
@@ -984,27 +1000,11 @@ configs:
     data_files:
       - split: train
         path: data/ngrams/tly_Latn/*.parquet
-  - config_name: sentences-iso_Latn
-    data_files:
-      - split: train
-        path: data/sentences/iso_Latn/*.parquet
-  - config_name: paragraphs-iso_Latn
-    data_files:
-      - split: train
-        path: data/paragraphs/iso_Latn/*.parquet
-  - config_name: words-iso_Latn
-    data_files:
-      - split: train
-        path: data/words/iso_Latn/*.parquet
-  - config_name: ngrams-iso_Latn
-    data_files:
-      - split: train
-        path: data/ngrams/iso_Latn/*.parquet
 ---
 
 # FineWeb-2 NLP
 
-**31,549,455 sentences** and **492,639,524 word tokens** across **1049 languages**, extracted from **701,380 source documents** (1.1 GB source data) in [FineWeb-2](https://huggingface.co/datasets/HuggingFaceFW/fineweb-2). Every sentence, paragraph, word frequency, and n-gram frequency, split with language-aware segmentation and continuously updated.
+**31,649,892 sentences** and **493,687,501 word tokens** across **1050 languages**, extracted from **705,478 source documents** (1.1 GB source data) in [FineWeb-2](https://huggingface.co/datasets/HuggingFaceFW/fineweb-2). Every sentence, paragraph, word frequency, and n-gram frequency, split with language-aware segmentation and continuously updated.
 
 ## Table of Contents
 
@@ -1041,10 +1041,10 @@ linguistically useful units:
 
 | Type | Rows | What you get |
 |------|------|-------------|
-| **sentences** | 31,549,455 | One row per sentence, with source document ID, URL, and position index |
-| **paragraphs** | 767,677 | One row per paragraph, with sentence count per paragraph |
-| **words** | 26,576,817 | Per-shard word frequency and document frequency tables |
-| **ngrams** | 508,804,795 | Per-shard bigram through 5-gram frequency tables |
+| **sentences** | 31,649,892 | One row per sentence, with source document ID, URL, and position index |
+| **paragraphs** | 771,775 | One row per paragraph, with sentence count per paragraph |
+| **words** | 26,594,627 | Per-shard word frequency and document frequency tables |
+| **ngrams** | 509,382,630 | Per-shard bigram through 5-gram frequency tables |
 
 Every row traces back to its source document through `doc_id` and `doc_url` fields, making
 it possible to navigate from any sentence or word back to the original web page. This
@@ -1204,6 +1204,7 @@ smn_Latn       ▓▓▓▓▓▓▓▓▓ 4,518
 cor_Latn       ▓▓▓▓▓▓▓▓ 4,164
 quy_Latn       ▓▓▓▓▓▓▓▓ 4,140
 yua_Latn       ▓▓▓▓▓▓▓▓ 4,139
+abt_Latn       ▓▓▓▓▓▓▓▓ 4,098
 ina_Latn       ▓▓▓▓▓▓▓▓ 4,071
 kua_Latn       ▓▓▓▓▓▓▓▓ 4,031
 gaa_Latn       ▓▓▓▓▓▓▓▓ 3,985
@@ -1212,7 +1213,6 @@ lbe_Cyrl       ▓▓▓▓▓▓▓▓ 3,939
 ven_Latn       ▓▓▓▓▓▓▓▓ 3,890
 crs_Latn       ▓▓▓▓▓▓▓▓ 3,836
 loz_Latn       ▓▓▓▓▓▓▓ 3,707
-tet_Latn       ▓▓▓▓▓▓▓ 3,631
 ```
 
 <details>
@@ -1373,20 +1373,20 @@ print(words_df)
 
 | Metric | Value |
 |--------|-------|
-| **Total sentences** | **31,549,455** |
-| **Total paragraphs** | **767,677** |
-| **Total word tokens** | **492,639,524** |
-| **Unique word entries** (per-shard) | 26,576,817 |
-| **Total n-gram entries** (per-shard) | 508,804,795 |
-| **Languages processed** | **1049** |
-| **Source documents** | **701,380** |
+| **Total sentences** | **31,649,892** |
+| **Total paragraphs** | **771,775** |
+| **Total word tokens** | **493,687,501** |
+| **Unique word entries** (per-shard) | 26,594,627 |
+| **Total n-gram entries** (per-shard) | 509,382,630 |
+| **Languages processed** | **1050** |
+| **Source documents** | **705,478** |
 | **Source data processed** | **1.1 GB** |
 | **Output parquet size** | **7.1 GB** |
-| Avg sentence length | 103.8 chars |
-| Avg paragraph length | 4306.3 chars |
-| Avg sentences per document | 45.0 |
+| Avg sentence length | 103.7 chars |
+| Avg paragraph length | 4292.8 chars |
+| Avg sentences per document | 44.9 |
 | Avg paragraphs per document | 1.1 |
-| Avg sentences per paragraph | 41.1 |
+| Avg sentences per paragraph | 41.0 |
 
 ### Per-language breakdown
 
@@ -1445,13 +1445,13 @@ print(words_df)
 | 51 | nzi_Latn (`nzi_Latn`) | 103,176 | 2,493 | 1,828,257 | 104.6 | 4371.2 | 2,493 | 1 | 3.7 MB | 16.3 MB |
 | 52 | lfn_Latn (`lfn_Latn`) | 101,599 | 3,551 | 1,837,456 | 95.7 | 2766.8 | 3,467 | 1 | 4.0 MB | 12.6 MB |
 | 53 | tlh_Latn (`tlh_Latn`) | 101,194 | 3,293 | 974,575 | 62.4 | 1947.9 | 2,959 | 1 | 2.4 MB | 11.6 MB |
-| 54 | gil_Latn (`gil_Latn`) | 100,201 | 3,289 | 2,139,671 | 108.4 | 3330.5 | 3,241 | 1 | 3.7 MB | 14.7 MB |
-| 55 | sma_Latn (`sma_Latn`) | 99,277 | 3,427 | 1,029,315 | 90.5 | 2650.2 | 3,372 | 1 | 3.4 MB | 10.7 MB |
-| 56 | pam_Latn (`pam_Latn`) | 98,608 | 2,162 | 1,140,731 | 69.7 | 3221.8 | 2,005 | 1 | 2.9 MB | 15.0 MB |
-| 57 | dak_Latn (`dak_Latn`) | 95,906 | 1,328 | 1,259,965 | 81.2 | 5938.5 | 412 | 1 | 1.9 MB | 19.6 MB |
-| 58 | btx_Latn (`btx_Latn`) | 95,824 | 2,305 | 1,384,785 | 91.1 | 3828.6 | 2,294 | 1 | 3.3 MB | 10.4 MB |
-| 59 | tly_Latn (`tly_Latn`) | 95,679 | 3,127 | 866,239 | 68.7 | 2131.3 | 2,821 | 1 | 2.8 MB | 11.3 MB |
-| 60 | iso_Latn (`iso_Latn`) | 95,104 | 2,196 | 1,890,498 | 106.6 | 4658.4 | 2,186 | 1 | 3.4 MB | 11.1 MB |
+| 54 | abt_Latn (`abt_Latn`) | 100,437 | 4,098 | 1,047,977 | 70.7 | 1755.4 | 4,098 | 1 | 2.1 MB | 9.6 MB |
+| 55 | gil_Latn (`gil_Latn`) | 100,201 | 3,289 | 2,139,671 | 108.4 | 3330.5 | 3,241 | 1 | 3.7 MB | 14.7 MB |
+| 56 | sma_Latn (`sma_Latn`) | 99,277 | 3,427 | 1,029,315 | 90.5 | 2650.2 | 3,372 | 1 | 3.4 MB | 10.7 MB |
+| 57 | pam_Latn (`pam_Latn`) | 98,608 | 2,162 | 1,140,731 | 69.7 | 3221.8 | 2,005 | 1 | 2.9 MB | 15.0 MB |
+| 58 | dak_Latn (`dak_Latn`) | 95,906 | 1,328 | 1,259,965 | 81.2 | 5938.5 | 412 | 1 | 1.9 MB | 19.6 MB |
+| 59 | btx_Latn (`btx_Latn`) | 95,824 | 2,305 | 1,384,785 | 91.1 | 3828.6 | 2,294 | 1 | 3.3 MB | 10.4 MB |
+| 60 | tly_Latn (`tly_Latn`) | 95,679 | 3,127 | 866,239 | 68.7 | 2131.3 | 2,821 | 1 | 2.8 MB | 11.3 MB |
 
 ## How it works
 
@@ -1722,4 +1722,4 @@ Created by **Duc-Tam Nguyen** ([tamnd](https://huggingface.co/tamnd)) as part of
 
 ---
 
-*Last updated: 2026-04-15 04:55 UTC*
+*Last updated: 2026-04-15 04:56 UTC*

data/ngrams/abt_Latn/0000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a979752661ed659d5e90f3d65bd0dc02f84d30ffcf79a3881e69779c47bac0f2
+size 5536234

data/paragraphs/abt_Latn/0000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e8184a404c405380555213a106aacbd31e749b30207745e337a4c66aae44b725
+size 2198250

data/sentences/abt_Latn/0000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d18731f9842f2d82e1ff10473802e1bc9f2692aac4ed708608a9167811ccf274
+size 2246332

data/words/abt_Latn/0000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd97f2a9885196f7c9123558ed4823fbb64ed2080d698cb5b6f3581fd2a3fc7a
+size 93882

stats.csv

@@ -1048,3 +1048,4 @@ kua_Latn,kua,0,4006,143986,4031,86238,401829,16003109,2026-04-15T04:49:01Z,56765
 smn_Latn,smn,0,4034,127853,4518,139414,390150,15565650,2026-04-15T04:49:40Z,5001967,1270652,91.0,2601.1
 fro_Latn,fro,0,4072,867115,6618,362971,1959020,60542135,2026-04-15T04:54:37Z,19892008,8441954,52.8,7051.4
 xmf_Geor,xmf,0,4079,93782,6205,193412,202495,16995986,2026-04-15T04:55:55Z,5269240,1079492,199.2,3024.2
+abt_Latn,abt,0,4098,100437,4098,17810,577835,10074698,2026-04-15T04:56:45Z,2172204,1047977,70.7,1755.4