tamnd commited on
Commit
5b9acdb
·
verified ·
1 Parent(s): 364f104

Publish 2 shards CC-MAIN-2026-08/00332–00753

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 **15,552,023 documents across 804 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 804 files of CC-MAIN-2026-08 (15,552,023 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
- | Stage | 804 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
- | Raw WARC (.warc.gz, downloaded) | ~651.7 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 1.9 TB | ~295 TB | — |
173
  | Packed markdown WARC (.md.warc.gz) | ~33.0 GB | ~3.7 TB | **-98.3%** vs HTML |
174
  | Final Parquet (Zstd level 19) | 22.3 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 1.9 TB of uncompressed HTML down to 70.1 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: ~651.7 GB of raw gzipped WARCs becomes **22.3 GB of Parquet** — a **96.6% total reduction** — containing 15,552,023 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
- Pipeline timings across 804 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
- Download (raw WARC) ████████████████████████ total 45h 57m 35s avg 3m 25s
186
- Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 2h 41m 32s avg 12s
187
- Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 10h 0m 12s avg 44s
188
  Publish (HuggingFace) ███░░░░░░░░░░░░░░░░░░░░░ total 5h 56m 44s avg 26s
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 **15,590,933 documents across 806 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 806 files of CC-MAIN-2026-08 (15,590,933 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
+ | Stage | 806 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
+ | Raw WARC (.warc.gz, downloaded) | ~653.3 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 1.9 TB | ~295 TB | — |
173
  | Packed markdown WARC (.md.warc.gz) | ~33.0 GB | ~3.7 TB | **-98.3%** vs HTML |
174
  | Final Parquet (Zstd level 19) | 22.3 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 1.9 TB of uncompressed HTML down to 70.3 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: ~653.3 GB of raw gzipped WARCs becomes **22.3 GB of Parquet** — a **96.6% total reduction** — containing 15,590,933 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
+ Pipeline timings across 806 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
+ Download (raw WARC) ████████████████████████ total 46h 7m 32s avg 3m 26s
186
+ Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 2h 51m 22s avg 12s
187
+ Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 10h 2m 17s avg 44s
188
  Publish (HuggingFace) ███░░░░░░░░░░░░░░░░░░░░░ total 5h 56m 44s avg 26s
189
  ```
190
 
data/CC-MAIN-2026-08/00332.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e590778afa77a0e1fee4d1e12e25ac6309af6d4f7f547302164a2b6ff345d113
3
+ size 30402651
data/CC-MAIN-2026-08/00753.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f7cb5933d14d0cde89503f34074eab20f0fc29b59501ad6cef870f62372f54e
3
+ size 28359904
stats.csv CHANGED
@@ -253,6 +253,7 @@ CC-MAIN-2026-08,328,19493,2599382576,91598938,28956556,2026-03-13T15:22:02Z,196,
253
  CC-MAIN-2026-08,329,19490,2578092970,94257462,29579596,2026-03-13T15:22:02Z,283,0,60,42
254
  CC-MAIN-2026-08,330,19236,2590008930,92805723,30030776,2026-03-13T17:50:47Z,0,401,75,0
255
  CC-MAIN-2026-08,331,19361,2559962643,92469515,29006296,2026-03-13T18:03:29Z,0,471,62,7
 
256
  CC-MAIN-2026-08,401,19441,2586676962,94334130,29618260,2026-03-13T11:41:00Z,0,0,33,0
257
  CC-MAIN-2026-08,402,19446,2552785517,96654791,30560442,2026-03-13T11:49:21Z,0,443,57,0
258
  CC-MAIN-2026-08,403,19320,2579048144,91406292,29341198,2026-03-13T11:56:46Z,0,414,29,0
@@ -397,6 +398,7 @@ CC-MAIN-2026-08,741,19382,2570241665,94898311,30073507,2026-03-13T12:01:53Z,223,
397
  CC-MAIN-2026-08,750,19471,2557384086,90848659,29734352,2026-03-13T17:41:16Z,435,0,70,21
398
  CC-MAIN-2026-08,751,19477,2574917221,91998519,29309299,2026-03-13T17:50:02Z,452,0,71,18
399
  CC-MAIN-2026-08,752,19489,2571204645,97532082,30423490,2026-03-13T18:03:30Z,520,0,60,7
 
400
  CC-MAIN-2026-08,851,19575,2612873485,97000364,30911294,2026-03-13T06:41:39Z,163,0,37,204
401
  CC-MAIN-2026-08,852,19334,2546015208,94959847,30205044,2026-03-13T06:45:52Z,183,0,57,11
402
  CC-MAIN-2026-08,853,19694,2567973975,95901932,30303494,2026-03-13T06:50:12Z,150,0,28,80
 
253
  CC-MAIN-2026-08,329,19490,2578092970,94257462,29579596,2026-03-13T15:22:02Z,283,0,60,42
254
  CC-MAIN-2026-08,330,19236,2590008930,92805723,30030776,2026-03-13T17:50:47Z,0,401,75,0
255
  CC-MAIN-2026-08,331,19361,2559962643,92469515,29006296,2026-03-13T18:03:29Z,0,471,62,7
256
+ CC-MAIN-2026-08,332,19576,2569972796,95254532,30402651,2026-03-13T18:10:41Z,0,590,62,0
257
  CC-MAIN-2026-08,401,19441,2586676962,94334130,29618260,2026-03-13T11:41:00Z,0,0,33,0
258
  CC-MAIN-2026-08,402,19446,2552785517,96654791,30560442,2026-03-13T11:49:21Z,0,443,57,0
259
  CC-MAIN-2026-08,403,19320,2579048144,91406292,29341198,2026-03-13T11:56:46Z,0,414,29,0
 
398
  CC-MAIN-2026-08,750,19471,2557384086,90848659,29734352,2026-03-13T17:41:16Z,435,0,70,21
399
  CC-MAIN-2026-08,751,19477,2574917221,91998519,29309299,2026-03-13T17:50:02Z,452,0,71,18
400
  CC-MAIN-2026-08,752,19489,2571204645,97532082,30423490,2026-03-13T18:03:30Z,520,0,60,7
401
+ CC-MAIN-2026-08,753,19334,2555127733,89627865,28359904,2026-03-13T18:10:41Z,597,0,63,0
402
  CC-MAIN-2026-08,851,19575,2612873485,97000364,30911294,2026-03-13T06:41:39Z,163,0,37,204
403
  CC-MAIN-2026-08,852,19334,2546015208,94959847,30205044,2026-03-13T06:45:52Z,183,0,57,11
404
  CC-MAIN-2026-08,853,19694,2567973975,95901932,30303494,2026-03-13T06:50:12Z,150,0,28,80