tamnd commited on
Commit
bd346e8
·
verified ·
1 Parent(s): 2736b07

Publish 2 shards CC-MAIN-2026-08/00196–00357

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 **18,530,580 documents across 957 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 957 files of CC-MAIN-2026-08 (18,530,580 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
- | Stage | 957 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
- | Raw WARC (.warc.gz, downloaded) | ~775.7 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 2.2 TB | ~295 TB | — |
173
- | Packed markdown WARC (.md.warc.gz) | ~39.3 GB | ~3.7 TB | **-98.3%** vs HTML |
174
- | Final Parquet (Zstd level 19) | 26.5 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 2.2 TB of uncompressed HTML down to 83.6 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: ~775.7 GB of raw gzipped WARCs becomes **26.5 GB of Parquet** — a **96.6% total reduction** — containing 18,530,580 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
- Pipeline timings across 957 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
- Download (raw WARC) ████████████████████████ total 63h 53m 6s avg 4m 0s
186
  Convert (HTML → MD) ██░░░░░░░░░░░░░░░░░░░░░░ total 5h 23m 35s avg 20s
187
- Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 14h 54m 41s avg 56s
188
  Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 6h 52m 30s 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 **18,569,427 documents across 959 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 959 files of CC-MAIN-2026-08 (18,569,427 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
+ | Stage | 959 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
+ | Raw WARC (.warc.gz, downloaded) | ~777.3 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 2.2 TB | ~295 TB | — |
173
+ | Packed markdown WARC (.md.warc.gz) | ~39.4 GB | ~3.7 TB | **-98.3%** vs HTML |
174
+ | Final Parquet (Zstd level 19) | 26.6 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 2.2 TB of uncompressed HTML down to 83.7 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: ~777.3 GB of raw gzipped WARCs becomes **26.6 GB of Parquet** — a **96.6% total reduction** — containing 18,569,427 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
+ Pipeline timings across 959 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
+ Download (raw WARC) ████████████████████████ total 64h 13m 7s avg 4m 1s
186
  Convert (HTML → MD) ██░░░░░░░░░░░░░░░░░░░░░░ total 5h 23m 35s avg 20s
187
+ Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 15h 0m 15s avg 56s
188
  Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 6h 52m 30s avg 25s
189
  ```
190
 
data/CC-MAIN-2026-08/00196.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e3f718604fed0d7a4bb47a364166b0861772889d243ffc749088d01233c0a27
3
+ size 30336090
data/CC-MAIN-2026-08/00357.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c78e687c0faa832fedfb2bbdfb524a86d99c982dc368aa33286ca9194b63bd2
3
+ size 29465451
stats.csv CHANGED
@@ -196,6 +196,7 @@ CC-MAIN-2026-08,192,19505,2557456629,92443840,29228248,2026-03-13T21:39:11Z,507,
196
  CC-MAIN-2026-08,193,19419,2582221477,93138448,30243524,2026-03-13T21:53:56Z,699,0,184,15
197
  CC-MAIN-2026-08,194,19501,2557747129,95458846,30399109,2026-03-13T22:02:44Z,419,0,111,27
198
  CC-MAIN-2026-08,195,19461,2590605003,94963046,29990690,2026-03-13T22:13:26Z,512,0,126,17
 
199
  CC-MAIN-2026-08,250,19200,2541559671,94959737,30214413,2026-03-13T17:42:47Z,511,0,85,19
200
  CC-MAIN-2026-08,251,19610,2594898345,97558492,31917087,2026-03-13T06:42:12Z,277,0,23,136
201
  CC-MAIN-2026-08,252,19240,2525928701,98163046,31104034,2026-03-13T06:47:10Z,238,0,49,10
@@ -303,6 +304,7 @@ CC-MAIN-2026-08,353,19470,2597274403,94820809,30335173,2026-03-13T21:39:11Z,0,51
303
  CC-MAIN-2026-08,354,19427,2562359871,96600307,30529507,2026-03-13T21:53:56Z,0,701,184,15
304
  CC-MAIN-2026-08,355,19508,2528378333,92581345,29133792,2026-03-13T22:02:26Z,0,385,115,13
305
  CC-MAIN-2026-08,356,19423,2558791293,93777087,30021761,2026-03-13T22:13:26Z,0,544,121,17
 
306
  CC-MAIN-2026-08,401,19441,2586676962,94334130,29618260,2026-03-13T11:41:00Z,0,0,33,0
307
  CC-MAIN-2026-08,402,19446,2552785517,96654791,30560442,2026-03-13T11:49:21Z,0,443,57,0
308
  CC-MAIN-2026-08,403,19320,2579048144,91406292,29341198,2026-03-13T11:56:46Z,0,414,29,0
 
196
  CC-MAIN-2026-08,193,19419,2582221477,93138448,30243524,2026-03-13T21:53:56Z,699,0,184,15
197
  CC-MAIN-2026-08,194,19501,2557747129,95458846,30399109,2026-03-13T22:02:44Z,419,0,111,27
198
  CC-MAIN-2026-08,195,19461,2590605003,94963046,29990690,2026-03-13T22:13:26Z,512,0,126,17
199
+ CC-MAIN-2026-08,196,19473,2586001479,97105191,30336090,2026-03-13T22:26:11Z,600,0,169,0
200
  CC-MAIN-2026-08,250,19200,2541559671,94959737,30214413,2026-03-13T17:42:47Z,511,0,85,19
201
  CC-MAIN-2026-08,251,19610,2594898345,97558492,31917087,2026-03-13T06:42:12Z,277,0,23,136
202
  CC-MAIN-2026-08,252,19240,2525928701,98163046,31104034,2026-03-13T06:47:10Z,238,0,49,10
 
304
  CC-MAIN-2026-08,354,19427,2562359871,96600307,30529507,2026-03-13T21:53:56Z,0,701,184,15
305
  CC-MAIN-2026-08,355,19508,2528378333,92581345,29133792,2026-03-13T22:02:26Z,0,385,115,13
306
  CC-MAIN-2026-08,356,19423,2558791293,93777087,30021761,2026-03-13T22:13:26Z,0,544,121,17
307
+ CC-MAIN-2026-08,357,19374,2574238367,93222357,29465451,2026-03-13T22:26:11Z,601,0,165,0
308
  CC-MAIN-2026-08,401,19441,2586676962,94334130,29618260,2026-03-13T11:41:00Z,0,0,33,0
309
  CC-MAIN-2026-08,402,19446,2552785517,96654791,30560442,2026-03-13T11:49:21Z,0,443,57,0
310
  CC-MAIN-2026-08,403,19320,2579048144,91406292,29341198,2026-03-13T11:56:46Z,0,414,29,0