tamnd commited on
Commit
168003c
·
verified ·
1 Parent(s): 8eac278

Publish 2 shards CC-MAIN-2026-08/00532–00781

Browse files
README.md CHANGED
@@ -32,7 +32,7 @@ configs:
32
 
33
  Open Index is a large-scale web text dataset built from [Common Crawl](https://commoncrawl.org). Every page goes through a pipeline that extracts the main content from raw HTML, converts it to clean Markdown using [trafilatura](https://github.com/adbar/trafilatura), and packages the result into Parquet files with full WARC metadata preserved.
34
 
35
- The dataset currently includes crawl **CC-MAIN-2026-08** with **18,919,603 documents across 977 shards**. We plan to add more snapshots over time.
36
 
37
  Open Index is released under the **Open Data Commons Attribution License (ODC-By) v1.0**, the same license used by Common Crawl.
38
 
@@ -164,27 +164,27 @@ Empty conversions (pages where trafilatura could not extract meaningful content)
164
 
165
  ### Compression Ratios
166
 
167
- Numbers below are actual measurements summed across all 977 files of CC-MAIN-2026-08 (18,919,603 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
- | Stage | 977 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
- | Raw WARC (.warc.gz, downloaded) | ~791.9 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 2.3 TB | ~295 TB | — |
173
- | Packed markdown WARC (.md.warc.gz) | ~40.1 GB | ~3.7 TB | **-98.3%** vs HTML |
174
- | Final Parquet (Zstd level 19) | 27.1 GB | ~2.9 TB | **-32.4%** vs packed WARC |
175
 
176
- The big win is the HTML → Markdown step: trafilatura strips all tags, scripts, styles, navigation, and ads, keeping only the main content. This cuts 2.3 TB of uncompressed HTML down to 85.3 GB of markdown — a **98.3% reduction** — before any file-level compression is applied. Parquet with Zstd level 19 then compresses the markdown a further 68.2%.
177
 
178
- End to end: ~791.9 GB of raw gzipped WARCs becomes **27.1 GB of Parquet** — a **96.6% total reduction** — containing 18,919,603 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
- Pipeline timings across 977 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
- Download (raw WARC) ████████████████████████ total 67h 0m 46s avg 4m 6s
186
  Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 5h 23m 35s avg 19s
187
- Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 15h 49m 19s avg 58s
188
  Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 6h 57m 0s avg 25s
189
  ```
190
 
 
32
 
33
  Open Index is a large-scale web text dataset built from [Common Crawl](https://commoncrawl.org). Every page goes through a pipeline that extracts the main content from raw HTML, converts it to clean Markdown using [trafilatura](https://github.com/adbar/trafilatura), and packages the result into Parquet files with full WARC metadata preserved.
34
 
35
+ The dataset currently includes crawl **CC-MAIN-2026-08** with **18,958,099 documents across 979 shards**. We plan to add more snapshots over time.
36
 
37
  Open Index is released under the **Open Data Commons Attribution License (ODC-By) v1.0**, the same license used by Common Crawl.
38
 
 
164
 
165
  ### Compression Ratios
166
 
167
+ Numbers below are actual measurements summed across all 979 files of CC-MAIN-2026-08 (18,958,099 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
+ | Stage | 979 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
+ | Raw WARC (.warc.gz, downloaded) | ~793.5 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 2.3 TB | ~295 TB | — |
173
+ | Packed markdown WARC (.md.warc.gz) | ~40.2 GB | ~3.7 TB | **-98.3%** vs HTML |
174
+ | Final Parquet (Zstd level 19) | 27.2 GB | ~2.9 TB | **-32.4%** vs packed WARC |
175
 
176
+ The big win is the HTML → Markdown step: trafilatura strips all tags, scripts, styles, navigation, and ads, keeping only the main content. This cuts 2.3 TB of uncompressed HTML down to 85.5 GB of markdown — a **98.3% reduction** — before any file-level compression is applied. Parquet with Zstd level 19 then compresses the markdown a further 68.2%.
177
 
178
+ End to end: ~793.5 GB of raw gzipped WARCs becomes **27.2 GB of Parquet** — a **96.6% total reduction** — containing 18,958,099 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
+ Pipeline timings across 979 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
+ Download (raw WARC) ████████████████████████ total 67h 18m 12s avg 4m 7s
186
  Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 5h 23m 35s avg 19s
187
+ Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 15h 53m 45s avg 58s
188
  Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 6h 57m 0s avg 25s
189
  ```
190
 
data/CC-MAIN-2026-08/00532.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05e5887ab891d814cff174d7512cbff55fe8c87f6b36062810cf9ea3b98748a5
3
+ size 30041771
data/CC-MAIN-2026-08/00781.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:665c15621b94af82932bd39c513b85edf679c7a1c262f37a4bc0e015157c6d20
3
+ size 30579446
stats.csv CHANGED
@@ -388,6 +388,7 @@ CC-MAIN-2026-08,528,19705,2604246714,96317981,30640287,2026-03-13T22:17:41Z,528,
388
  CC-MAIN-2026-08,529,19635,2618160914,92856775,30314956,2026-03-13T22:28:41Z,515,0,124,32
389
  CC-MAIN-2026-08,530,19529,2583246308,92897292,30240589,2026-03-13T22:40:41Z,601,0,143,12
390
  CC-MAIN-2026-08,531,19460,2553022933,93648476,30171716,2026-03-13T22:53:41Z,609,0,162,8
 
391
  CC-MAIN-2026-08,551,19551,2609034951,89980913,29167227,2026-03-13T06:42:51Z,280,0,20,176
392
  CC-MAIN-2026-08,552,19402,2594817630,94092761,30309254,2026-03-13T06:48:30Z,211,0,46,80
393
  CC-MAIN-2026-08,553,19856,2617133708,99106486,30365588,2026-03-13T06:52:51Z,116,0,23,120
@@ -511,6 +512,7 @@ CC-MAIN-2026-08,777,19589,2585482711,97636638,31875194,2026-03-13T22:16:47Z,538,
511
  CC-MAIN-2026-08,778,19374,2563110936,89906933,28709645,2026-03-13T22:28:11Z,579,0,125,26
512
  CC-MAIN-2026-08,779,19381,2581538675,94470928,30923647,2026-03-13T22:40:41Z,612,0,140,12
513
  CC-MAIN-2026-08,780,19412,2530827471,89106010,28268620,2026-03-13T22:53:41Z,619,0,161,8
 
514
  CC-MAIN-2026-08,851,19575,2612873485,97000364,30911294,2026-03-13T06:41:39Z,163,0,37,204
515
  CC-MAIN-2026-08,852,19334,2546015208,94959847,30205044,2026-03-13T06:45:52Z,183,0,57,11
516
  CC-MAIN-2026-08,853,19694,2567973975,95901932,30303494,2026-03-13T06:50:12Z,150,0,28,80
 
388
  CC-MAIN-2026-08,529,19635,2618160914,92856775,30314956,2026-03-13T22:28:41Z,515,0,124,32
389
  CC-MAIN-2026-08,530,19529,2583246308,92897292,30240589,2026-03-13T22:40:41Z,601,0,143,12
390
  CC-MAIN-2026-08,531,19460,2553022933,93648476,30171716,2026-03-13T22:53:41Z,609,0,162,8
391
+ CC-MAIN-2026-08,532,19118,2543492135,93810142,30041771,2026-03-13T23:04:41Z,525,0,136,0
392
  CC-MAIN-2026-08,551,19551,2609034951,89980913,29167227,2026-03-13T06:42:51Z,280,0,20,176
393
  CC-MAIN-2026-08,552,19402,2594817630,94092761,30309254,2026-03-13T06:48:30Z,211,0,46,80
394
  CC-MAIN-2026-08,553,19856,2617133708,99106486,30365588,2026-03-13T06:52:51Z,116,0,23,120
 
512
  CC-MAIN-2026-08,778,19374,2563110936,89906933,28709645,2026-03-13T22:28:11Z,579,0,125,26
513
  CC-MAIN-2026-08,779,19381,2581538675,94470928,30923647,2026-03-13T22:40:41Z,612,0,140,12
514
  CC-MAIN-2026-08,780,19412,2530827471,89106010,28268620,2026-03-13T22:53:41Z,619,0,161,8
515
+ CC-MAIN-2026-08,781,19378,2573128052,96502694,30579446,2026-03-13T23:04:41Z,521,0,130,0
516
  CC-MAIN-2026-08,851,19575,2612873485,97000364,30911294,2026-03-13T06:41:39Z,163,0,37,204
517
  CC-MAIN-2026-08,852,19334,2546015208,94959847,30205044,2026-03-13T06:45:52Z,183,0,57,11
518
  CC-MAIN-2026-08,853,19694,2567973975,95901932,30303494,2026-03-13T06:50:12Z,150,0,28,80