Hugging Face's logo

Datasets:
open-index
/
open-markdown

Tasks:
Text Generation
Feature Extraction
Modalities:
Text
Languages:
English
Size:
1M<n<10M
Tags:
common-crawl
web-crawl
markdown
text
License:
Dataset card Files
xet
 Community
1
tamnd commited on
Commit
c766df5
·
verified ·
1 Parent(s): 258b1fb

Publish 19 shards CC-MAIN-2026-21/26207–26226

Browse files
Files changed (23) hide show
  1. README.md +15 -15
  2. charts/timing_chart.png +2 -2
  3. charts/totals_chart.png +2 -2
  4. data/CC-MAIN-2026-21/02/62/026207.parquet +3 -0
  5. data/CC-MAIN-2026-21/02/62/026208.parquet +3 -0
  6. data/CC-MAIN-2026-21/02/62/026209.parquet +3 -0
  7. data/CC-MAIN-2026-21/02/62/026210.parquet +3 -0
  8. data/CC-MAIN-2026-21/02/62/026212.parquet +3 -0
  9. data/CC-MAIN-2026-21/02/62/026213.parquet +3 -0
  10. data/CC-MAIN-2026-21/02/62/026214.parquet +3 -0
  11. data/CC-MAIN-2026-21/02/62/026215.parquet +3 -0
  12. data/CC-MAIN-2026-21/02/62/026216.parquet +3 -0
  13. data/CC-MAIN-2026-21/02/62/026217.parquet +3 -0
  14. data/CC-MAIN-2026-21/02/62/026218.parquet +3 -0
  15. data/CC-MAIN-2026-21/02/62/026219.parquet +3 -0
  16. data/CC-MAIN-2026-21/02/62/026220.parquet +3 -0
  17. data/CC-MAIN-2026-21/02/62/026221.parquet +3 -0
  18. data/CC-MAIN-2026-21/02/62/026222.parquet +3 -0
  19. data/CC-MAIN-2026-21/02/62/026223.parquet +3 -0
  20. data/CC-MAIN-2026-21/02/62/026224.parquet +3 -0
  21. data/CC-MAIN-2026-21/02/62/026225.parquet +3 -0
  22. data/CC-MAIN-2026-21/02/62/026226.parquet +3 -0
  23. stats.csv +23 -4
README.md CHANGED
@@ -32,19 +32,19 @@ configs:
32
 
33
  **Open Markdown** is a large-scale web text dataset built from [Common Crawl](https://commoncrawl.org). Common Crawl is a non-profit that crawls the web and freely provides its archives and datasets to the public — see [their latest crawl announcement](https://commoncrawl.org/blog/march-2026-crawl-archive-now-available) for details on the source data. Every page goes through a pipeline that extracts the main content from raw HTML, converts it to clean Markdown, and packages the result into Parquet files with useful WARC metadata for traceability.
34
 
35
- The dataset currently includes crawl **CC-MAIN-2026-21** with **45,356,295 documents across 2352 shards**. Processed 6.1 TB of raw HTML into 364.1 GB of clean Markdown — a **94.2% reduction**. We plan to add more snapshots over time.
36
 
37
  ### Live Progress
38
 
39
- Processing at **36.9 shards/hour** — 2,352 of 100,000 done (**2.35%**)
40
 
41
- Estimated completion: **September 22, 2026** (110 days)
42
 
43
- **Current server:** 6 CPU cores, 12 GB RAM (8.4 GB available), 24 GB disk free
44
 
45
  **Memory per session:** avg 607 MB, peak 667 MB (measured via VmRSS)
46
 
47
- **With 10 identical servers:** 369 shards/hour → June 15, 2026 (11 days)
48
 
49
  **Open Markdown** is released under the **Open Data Commons Attribution License (ODC-By) v1.0**, the same license used by Common Crawl.
50
 
@@ -182,27 +182,27 @@ No intermediate files are created — the pipeline streams from compressed WARC
182
 
183
  ### Compression Ratios
184
 
185
- Numbers below are actual measurements summed across all 2352 files of CC-MAIN-2026-21 (45,356,295 pages total), projected to the full crawl of 100,000 WARC files.
186
 
187
- | Stage | 2352 files (measured) | 100,000 files (projected) | Reduction |
188
  |---|---|---|---|
189
  | Raw WARC (.warc.gz, downloaded) | ~1.9 TB | ~79.2 TB | — |
190
  | HTML extracted (uncompressed) | 6.1 TB | ~259.7 TB | — |
191
- | Markdown (clean text) | 364.1 GB | ~15.1 TB | **-94.2%** vs HTML |
192
- | Final Parquet (Zstd) | 115.2 GB | ~4.8 TB | **-68.4%** vs markdown |
193
 
194
- The big win is HTML → Markdown conversion: the tokenizer strips all tags, scripts, styles, navigation, and ads, keeping only the main content. This cuts 6.1 TB of uncompressed HTML down to 364.1 GB of markdown — a **94.2% reduction**. Parquet with Zstd then compresses the markdown a further 68.4%.
195
 
196
- End to end: ~1.9 TB of raw gzipped WARCs becomes **115.2 GB of Parquet** — a **94.0% total reduction** — containing 45,356,295 clean markdown documents.
197
 
198
  ### Processing Times
199
 
200
- Pipeline timings across 2352 shards of CC-MAIN-2026-21:
201
 
202
  ```
203
- Download (raw WARC) ███████░░░░░░░░░░░░░░░░░ 27h 22m 54s
204
- Convert (HTML → Markdown → Parquet) ████████████████████████ 92h 33m 46s
205
- Publish (HuggingFace) ████████████████░░░░░░░░ 64h 45m 52s
206
  ```
207
 
208
  ### Dataset Charts
 
32
 
33
  **Open Markdown** is a large-scale web text dataset built from [Common Crawl](https://commoncrawl.org). Common Crawl is a non-profit that crawls the web and freely provides its archives and datasets to the public — see [their latest crawl announcement](https://commoncrawl.org/blog/march-2026-crawl-archive-now-available) for details on the source data. Every page goes through a pipeline that extracts the main content from raw HTML, converts it to clean Markdown, and packages the result into Parquet files with useful WARC metadata for traceability.
34
 
35
+ The dataset currently includes crawl **CC-MAIN-2026-21** with **45,433,873 documents across 2356 shards**. Processed 6.1 TB of raw HTML into 364.7 GB of clean Markdown — a **94.2% reduction**. We plan to add more snapshots over time.
36
 
37
  ### Live Progress
38
 
39
+ Processing at **35.1 shards/hour** — 2,356 of 100,000 done (**2.36%**)
40
 
41
+ Estimated completion: **September 28, 2026** (116 days)
42
 
43
+ **Current server:** 6 CPU cores, 12 GB RAM (8.1 GB available), 21 GB disk free
44
 
45
  **Memory per session:** avg 607 MB, peak 667 MB (measured via VmRSS)
46
 
47
+ **With 10 identical servers:** 351 shards/hour → June 15, 2026 (12 days)
48
 
49
  **Open Markdown** is released under the **Open Data Commons Attribution License (ODC-By) v1.0**, the same license used by Common Crawl.
50
 
 
182
 
183
  ### Compression Ratios
184
 
185
+ Numbers below are actual measurements summed across all 2356 files of CC-MAIN-2026-21 (45,433,873 pages total), projected to the full crawl of 100,000 WARC files.
186
 
187
+ | Stage | 2356 files (measured) | 100,000 files (projected) | Reduction |
188
  |---|---|---|---|
189
  | Raw WARC (.warc.gz, downloaded) | ~1.9 TB | ~79.2 TB | — |
190
  | HTML extracted (uncompressed) | 6.1 TB | ~259.7 TB | — |
191
+ | Markdown (clean text) | 364.7 GB | ~15.1 TB | **-94.2%** vs HTML |
192
+ | Final Parquet (Zstd) | 115.4 GB | ~4.8 TB | **-68.4%** vs markdown |
193
 
194
+ The big win is HTML → Markdown conversion: the tokenizer strips all tags, scripts, styles, navigation, and ads, keeping only the main content. This cuts 6.1 TB of uncompressed HTML down to 364.7 GB of markdown — a **94.2% reduction**. Parquet with Zstd then compresses the markdown a further 68.4%.
195
 
196
+ End to end: ~1.9 TB of raw gzipped WARCs becomes **115.4 GB of Parquet** — a **94.0% total reduction** — containing 45,433,873 clean markdown documents.
197
 
198
  ### Processing Times
199
 
200
+ Pipeline timings across 2356 shards of CC-MAIN-2026-21:
201
 
202
  ```
203
+ Download (raw WARC) ███████░░░░░░░░░░░░░░░░░ 27h 44m 6s
204
+ Convert (HTML → Markdown → Parquet) ████████████████████████ 93h 20m 52s
205
+ Publish (HuggingFace) ████████████████░░░░░░░░ 65h 36m 0s
206
  ```
207
 
208
  ### Dataset Charts
charts/timing_chart.png CHANGED

Git LFS Details

  • SHA256: d306f26b48cce2f551ebabb6d54f700d9caa7e7352bb06c82644b51866b47151
  • Pointer size: 130 Bytes
  • Size of remote file: 82.2 kB

Git LFS Details

  • SHA256: 035dbc6cb2846bface4f697ac6e0820c60df383ec67e8f53ce580d96aa32a1cd
  • Pointer size: 130 Bytes
  • Size of remote file: 82 kB
charts/totals_chart.png CHANGED

Git LFS Details

  • SHA256: 2ca7a95a4283f63e1d63bba810ed4e6fb463704b6dd3c6f3a8c06a3721438988
  • Pointer size: 130 Bytes
  • Size of remote file: 78.2 kB

Git LFS Details

  • SHA256: 101b34d3313a2073c2a35a7236e948bc275966f181ce2f78e88d0b6917f6fad2
  • Pointer size: 130 Bytes
  • Size of remote file: 78.7 kB
data/CC-MAIN-2026-21/02/62/026207.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fbef6d7455bb9d8a6caf8ffb56bf7e6ede2b25591f25ba155776c75e48b0a878
3
+ size 54852990
data/CC-MAIN-2026-21/02/62/026208.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:595996793bd0c30255cdb446d4bb869853ae6fc290fb7e445c3414cf3fcf7ec8
3
+ size 52438431
data/CC-MAIN-2026-21/02/62/026209.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4802802a3da68d21dfff67374bf002f5d1d05f0782e62e3af6ff7684e540825
3
+ size 52561940
data/CC-MAIN-2026-21/02/62/026210.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2d2960b432a7852726c3ba7be8d92feaa7ecefa36da45a9f71d62e563f9e0c8
3
+ size 53229764
data/CC-MAIN-2026-21/02/62/026212.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb485b75d4ad0328c6bcd4dd849b19ac26d4bf07d396f8a282368620be01778f
3
+ size 53745473
data/CC-MAIN-2026-21/02/62/026213.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a01f1ce7c018812a3d0a812dd49225d3628fa8b230a45de4db567fa4b663d3d3
3
+ size 53213328
data/CC-MAIN-2026-21/02/62/026214.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1fee9ffe21195d12dc4d30666a8d2271bd06937aae948f5fb1f5de613e1b194b
3
+ size 51572352
data/CC-MAIN-2026-21/02/62/026215.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e6a6842f086c126957ed61edd1d21357a43a6be7a3719b60f33a741f68b2138
3
+ size 55790928
data/CC-MAIN-2026-21/02/62/026216.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fdc8bacc7c2690706154ed81b20b39cbbfd2491ae6813cdecf591b41355391ee
3
+ size 52601609
data/CC-MAIN-2026-21/02/62/026217.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a9a9c36621ad482eba996921c50e13373cd98bbd7d097ae52fcb650502a52dc
3
+ size 52208690
data/CC-MAIN-2026-21/02/62/026218.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8c1f84d1a7051e2bb7240c09390bc69c9472dc0d550d854a477cb301ddbca21
3
+ size 53390746
data/CC-MAIN-2026-21/02/62/026219.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:61ca9b471c2f0d5fc174834f30e63b487130806c23fe10caa8d35f134a163b31
3
+ size 53309850
data/CC-MAIN-2026-21/02/62/026220.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:548f2d8fc14d5a35b8fe607c3d59fee7bac02be36b3005ea39c5dcb8e200eaff
3
+ size 53663174
data/CC-MAIN-2026-21/02/62/026221.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b027fca07e04a733ff4bf6f57bb1b415fe9854f46f2624c82f4ac270a8de087
3
+ size 51788947
data/CC-MAIN-2026-21/02/62/026222.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aefd4d9ef7d7c879e13b4fa13c5b4f413c19cb5cce758a5b9c6784f6ea654b48
3
+ size 50662874
data/CC-MAIN-2026-21/02/62/026223.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa8df05c452f02e8567c502dbd3c54b8b586d08c0dd9b3348247128036aba80f
3
+ size 52855828
data/CC-MAIN-2026-21/02/62/026224.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7fc0b1d4e8ea2ce50fb5b74b513a1f536dda11f44fe9220ab9066f94942efd0d
3
+ size 52770237
data/CC-MAIN-2026-21/02/62/026225.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f721fc254fed084306ad29bf7349e79a5aa2f303e9a76fefc070419b7abb734
3
+ size 52209157
data/CC-MAIN-2026-21/02/62/026226.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e2c091ae88117d7eba2d053eb2a267ce26337ccdb8fe4c68323d115fbf58465
3
+ size 54534367
stats.csv CHANGED
@@ -2307,7 +2307,26 @@ CC-MAIN-2026-21,26199,19472,2878607506,165993984,53605739,2026-06-04T00:34:33Z,6
2307
  CC-MAIN-2026-21,26200,19300,2810049943,162353174,51216676,2026-06-04T00:45:54Z,81,158,1,97,591
2308
  CC-MAIN-2026-21,26201,19268,2844050500,167248834,52188192,2026-06-04T00:45:54Z,69,139,1,97,623
2309
  CC-MAIN-2026-21,26202,19323,2883594340,171861419,54826499,2026-06-04T00:45:54Z,81,145,1,97,637
2310
- CC-MAIN-2026-21,26203,19401,2820877893,167716554,52240377,2026-06-04T00:55:33Z,58,147,3,0,628
2311
- CC-MAIN-2026-21,26204,19316,2866041397,167207216,53448491,2026-06-04T00:55:33Z,80,129,1,0,627
2312
- CC-MAIN-2026-21,26205,19355,2859915952,170790533,53592243,2026-06-04T00:55:33Z,70,134,1,0,622
2313
- CC-MAIN-2026-21,26206,19175,2820298273,166881892,54385524,2026-06-04T00:55:33Z,71,115,2,0,633
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2307
  CC-MAIN-2026-21,26200,19300,2810049943,162353174,51216676,2026-06-04T00:45:54Z,81,158,1,97,591
2308
  CC-MAIN-2026-21,26201,19268,2844050500,167248834,52188192,2026-06-04T00:45:54Z,69,139,1,97,623
2309
  CC-MAIN-2026-21,26202,19323,2883594340,171861419,54826499,2026-06-04T00:45:54Z,81,145,1,97,637
2310
+ CC-MAIN-2026-21,26203,19401,2820877893,167716554,52240377,2026-06-04T00:55:33Z,58,147,3,752,628
2311
+ CC-MAIN-2026-21,26204,19316,2866041397,167207216,53448491,2026-06-04T00:55:33Z,80,129,1,752,627
2312
+ CC-MAIN-2026-21,26205,19355,2859915952,170790533,53592243,2026-06-04T00:55:33Z,70,134,1,752,622
2313
+ CC-MAIN-2026-21,26206,19175,2820298273,166881892,54385524,2026-06-04T00:55:33Z,71,115,2,752,633
2314
+ CC-MAIN-2026-21,26207,19350,2900644536,167693166,54852990,2026-06-04T01:45:43Z,0,125,1,0,644
2315
+ CC-MAIN-2026-21,26208,19513,2898657102,166379964,52438431,2026-06-04T01:45:43Z,79,115,1,0,641
2316
+ CC-MAIN-2026-21,26209,19184,2893700054,165395720,52561940,2026-06-04T01:45:43Z,63,158,1,0,640
2317
+ CC-MAIN-2026-21,26210,19323,2872861483,165171485,53229764,2026-06-04T01:45:43Z,79,149,1,0,642
2318
+ CC-MAIN-2026-21,26212,19219,2840220974,170128037,53745473,2026-06-04T01:45:43Z,73,181,2,0,602
2319
+ CC-MAIN-2026-21,26213,19219,2861343370,167362185,53213328,2026-06-04T01:45:43Z,169,242,1,0,605
2320
+ CC-MAIN-2026-21,26214,19247,2869010085,164508396,51572352,2026-06-04T01:45:43Z,91,251,3,0,641
2321
+ CC-MAIN-2026-21,26215,19160,2828170608,168642778,55790928,2026-06-04T01:45:43Z,74,196,1,0,625
2322
+ CC-MAIN-2026-21,26216,19421,2878724771,165738755,52601609,2026-06-04T01:45:43Z,65,111,0,0,616
2323
+ CC-MAIN-2026-21,26217,19367,2869363742,167826857,52208690,2026-06-04T01:45:43Z,0,126,1,0,608
2324
+ CC-MAIN-2026-21,26218,19517,2899314084,169394970,53390746,2026-06-04T01:45:43Z,107,137,1,0,621
2325
+ CC-MAIN-2026-21,26219,19368,2904319773,167993909,53309850,2026-06-04T01:45:43Z,59,171,2,0,643
2326
+ CC-MAIN-2026-21,26220,19264,2857441597,165131635,53663174,2026-06-04T01:45:43Z,59,133,2,0,616
2327
+ CC-MAIN-2026-21,26221,19136,2848739309,165532279,51788947,2026-06-04T01:45:43Z,67,108,1,0,612
2328
+ CC-MAIN-2026-21,26222,19153,2816594821,161310000,50662874,2026-06-04T01:45:43Z,50,130,0,0,609
2329
+ CC-MAIN-2026-21,26223,19152,2828596246,166032296,52855828,2026-06-04T01:45:43Z,53,106,1,0,641
2330
+ CC-MAIN-2026-21,26224,19274,2845037762,167421188,52770237,2026-06-04T01:45:43Z,46,134,1,0,625
2331
+ CC-MAIN-2026-21,26225,19328,2871465764,165451926,52209157,2026-06-04T01:45:43Z,85,101,2,0,625
2332
+ CC-MAIN-2026-21,26226,19640,2910241971,166889375,54534367,2026-06-04T01:45:43Z,53,129,1,0,630