tamnd commited on
Commit
d852a5b
·
verified ·
1 Parent(s): 37de199

Publish shard CC-MAIN-2026-08/02617

Browse files
Files changed (3) hide show
  1. README.md +9 -9
  2. data/CC-MAIN-2026-08/02617.parquet +3 -0
  3. stats.csv +2 -1
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 **15,007,621 documents across 776 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
 
@@ -164,28 +164,28 @@ Empty conversions (pages where trafilatura could not extract meaningful content)
164
 
165
  ### Compression Ratios
166
 
167
- Numbers below are actual measurements summed across all 776 files of CC-MAIN-2026-08 (15,007,621 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
- | Stage | 776 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
- | Raw WARC (.warc.gz, downloaded) | ~629.0 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 1.8 TB | ~295 TB | — |
173
  | Packed markdown WARC (.md.warc.gz) | ~31.8 GB | ~3.7 TB | **-98.3%** vs HTML |
174
  | Final Parquet (Zstd level 19) | 21.5 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 1.8 TB of uncompressed HTML down to 67.7 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: ~629.0 GB of raw gzipped WARCs becomes **21.5 GB of Parquet** — a **96.6% total reduction** — containing 15,007,621 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
- Pipeline timings across 776 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
- Download (raw WARC) ████████████████████████ total 44h 1m 44s avg 3m 24s
186
  Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 1h 50m 33s avg 8s
187
- Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 9h 21m 53s avg 43s
188
- Publish (HuggingFace) ███░░░░░░░░░░░░░░░░░░░░░ total 5h 48m 10s avg 26s
189
  ```
190
 
191
  ### Dataset Charts
 
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 **15,027,006 documents across 777 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 777 files of CC-MAIN-2026-08 (15,027,006 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
+ | Stage | 777 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
+ | Raw WARC (.warc.gz, downloaded) | ~629.8 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 1.8 TB | ~295 TB | — |
173
  | Packed markdown WARC (.md.warc.gz) | ~31.8 GB | ~3.7 TB | **-98.3%** vs HTML |
174
  | Final Parquet (Zstd level 19) | 21.5 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 1.8 TB of uncompressed HTML down to 67.7 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: ~629.8 GB of raw gzipped WARCs becomes **21.5 GB of Parquet** — a **96.6% total reduction** — containing 15,027,006 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
+ Pipeline timings across 777 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
+ Download (raw WARC) ████████████████████████ total 44h 5m 1s avg 3m 24s
186
  Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 1h 50m 33s avg 8s
187
+ Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 9h 22m 21s avg 43s
188
+ Publish (HuggingFace) ███░░░░░░░░░░░░░░░░░░░░░ total 5h 48m 33s avg 26s
189
  ```
190
 
191
  ### Dataset Charts
data/CC-MAIN-2026-08/02617.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9b5e66499daa93ee6c4e958b96821ee8d2ec1e5a69405769dd96f5573af9b412
3
+ size 28106661
stats.csv CHANGED
@@ -695,7 +695,7 @@ CC-MAIN-2026-08,2324,19548,2591124523,95026738,29706594,2026-03-13T15:08:51Z,243
695
  CC-MAIN-2026-08,2325,19795,2595958710,94769924,30381849,2026-03-13T15:21:56Z,340,0,120,4
696
  CC-MAIN-2026-08,2326,19585,2528968070,92409071,30194739,2026-03-13T15:25:30Z,413,0,125,18
697
  CC-MAIN-2026-08,2327,19472,2582899616,94317330,32276911,2026-03-13T15:32:45Z,281,0,163,15
698
- CC-MAIN-2026-08,2354,19403,2561720258,91946981,29788280,2026-03-13T17:24:10Z,945,0,44,0
699
  CC-MAIN-2026-08,2551,5410,611835922,35005098,10258608,2026-03-13T12:28:41Z,32,0,10,0
700
  CC-MAIN-2026-08,2552,19598,2602424114,96235194,30473559,2026-03-13T12:36:26Z,437,0,26,0
701
  CC-MAIN-2026-08,2553,19534,2542432805,90911643,29613848,2026-03-13T12:44:33Z,444,0,41,0
@@ -718,6 +718,7 @@ CC-MAIN-2026-08,2584,19484,2553112375,97091387,30756815,2026-03-13T16:21:41Z,120
718
  CC-MAIN-2026-08,2614,19566,2581388744,91738092,29557124,2026-03-13T17:13:26Z,207,0,47,15
719
  CC-MAIN-2026-08,2615,19805,2626737496,95131439,30207314,2026-03-13T17:17:25Z,208,0,34,54
720
  CC-MAIN-2026-08,2616,19500,2582825107,93822871,29979582,2026-03-13T17:21:11Z,171,0,52,26
 
721
  CC-MAIN-2026-08,2801,19800,2573332995,93935647,29781057,2026-03-13T12:36:24Z,480,0,24,0
722
  CC-MAIN-2026-08,2802,19377,2552826046,96926255,30379332,2026-03-13T12:44:04Z,402,0,56,0
723
  CC-MAIN-2026-08,2810,19479,2567872535,94762671,29742907,2026-03-13T14:03:56Z,0,0,0,4
 
695
  CC-MAIN-2026-08,2325,19795,2595958710,94769924,30381849,2026-03-13T15:21:56Z,340,0,120,4
696
  CC-MAIN-2026-08,2326,19585,2528968070,92409071,30194739,2026-03-13T15:25:30Z,413,0,125,18
697
  CC-MAIN-2026-08,2327,19472,2582899616,94317330,32276911,2026-03-13T15:32:45Z,281,0,163,15
698
+ CC-MAIN-2026-08,2354,19403,2561720258,91946981,29788280,2026-03-13T17:24:10Z,945,0,44,23
699
  CC-MAIN-2026-08,2551,5410,611835922,35005098,10258608,2026-03-13T12:28:41Z,32,0,10,0
700
  CC-MAIN-2026-08,2552,19598,2602424114,96235194,30473559,2026-03-13T12:36:26Z,437,0,26,0
701
  CC-MAIN-2026-08,2553,19534,2542432805,90911643,29613848,2026-03-13T12:44:33Z,444,0,41,0
 
718
  CC-MAIN-2026-08,2614,19566,2581388744,91738092,29557124,2026-03-13T17:13:26Z,207,0,47,15
719
  CC-MAIN-2026-08,2615,19805,2626737496,95131439,30207314,2026-03-13T17:17:25Z,208,0,34,54
720
  CC-MAIN-2026-08,2616,19500,2582825107,93822871,29979582,2026-03-13T17:21:11Z,171,0,52,26
721
+ CC-MAIN-2026-08,2617,19385,2596795747,89963126,28106661,2026-03-13T17:24:55Z,197,0,28,0
722
  CC-MAIN-2026-08,2801,19800,2573332995,93935647,29781057,2026-03-13T12:36:24Z,480,0,24,0
723
  CC-MAIN-2026-08,2802,19377,2552826046,96926255,30379332,2026-03-13T12:44:04Z,402,0,56,0
724
  CC-MAIN-2026-08,2810,19479,2567872535,94762671,29742907,2026-03-13T14:03:56Z,0,0,0,4