Publish shard CC-MAIN-2026-08/00534
Browse files- README.md +10 -10
- data/CC-MAIN-2026-08/00534.parquet +3 -0
- stats.csv +1 -0
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 **19,
|
| 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
|
| 168 |
|
| 169 |
-
| Stage |
|
| 170 |
|---|---|---|---|
|
| 171 |
-
| Raw WARC (.warc.gz, downloaded) | ~
|
| 172 |
| HTML extracted (uncompressed) | 2.3 TB | ~295 TB | — |
|
| 173 |
| Packed markdown WARC (.md.warc.gz) | ~40.6 GB | ~3.7 TB | **-98.3%** vs HTML |
|
| 174 |
-
| Final Parquet (Zstd level 19) | 27.
|
| 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 86.
|
| 177 |
|
| 178 |
-
End to end: ~
|
| 179 |
|
| 180 |
### Processing Times
|
| 181 |
|
| 182 |
-
Pipeline timings across
|
| 183 |
|
| 184 |
```
|
| 185 |
-
Download (raw WARC) ████████████████████████ total 68h
|
| 186 |
Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 5h 23m 35s avg 19s
|
| 187 |
-
Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 16h
|
| 188 |
Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 7h 0m 47s 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 **19,171,696 documents across 990 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 990 files of CC-MAIN-2026-08 (19,171,696 pages total), projected to the full crawl of 100,000 WARC files.
|
| 168 |
|
| 169 |
+
| Stage | 990 files (measured) | 100,000 files (projected) | Reduction |
|
| 170 |
|---|---|---|---|
|
| 171 |
+
| Raw WARC (.warc.gz, downloaded) | ~802.4 GB | ~83 TB | — |
|
| 172 |
| HTML extracted (uncompressed) | 2.3 TB | ~295 TB | — |
|
| 173 |
| Packed markdown WARC (.md.warc.gz) | ~40.6 GB | ~3.7 TB | **-98.3%** vs HTML |
|
| 174 |
+
| Final Parquet (Zstd level 19) | 27.5 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 2.3 TB of uncompressed HTML down to 86.5 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: ~802.4 GB of raw gzipped WARCs becomes **27.5 GB of Parquet** — a **96.6% total reduction** — containing 19,171,696 clean markdown documents.
|
| 179 |
|
| 180 |
### Processing Times
|
| 181 |
|
| 182 |
+
Pipeline timings across 990 shards of CC-MAIN-2026-08:
|
| 183 |
|
| 184 |
```
|
| 185 |
+
Download (raw WARC) ████████████████████████ total 68h 53m 7s avg 4m 10s
|
| 186 |
Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 5h 23m 35s avg 19s
|
| 187 |
+
Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 16h 25m 29s avg 59s
|
| 188 |
Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 7h 0m 47s avg 25s
|
| 189 |
```
|
| 190 |
|
data/CC-MAIN-2026-08/00534.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c5fb059331f41f9c5c8b7334974dcec2ce3c4c28b5e5422f95808afe01b27f3
|
| 3 |
+
size 29713409
|
stats.csv
CHANGED
|
@@ -394,6 +394,7 @@ CC-MAIN-2026-08,530,19529,2583246308,92897292,30240589,2026-03-13T22:40:41Z,601,
|
|
| 394 |
CC-MAIN-2026-08,531,19460,2553022933,93648476,30171716,2026-03-13T22:53:41Z,609,0,162,8
|
| 395 |
CC-MAIN-2026-08,532,19118,2543492135,93810142,30041771,2026-03-13T23:04:41Z,525,0,136,15
|
| 396 |
CC-MAIN-2026-08,533,19410,2555040866,91016286,29048349,2026-03-13T23:17:18Z,579,0,169,23
|
|
|
|
| 397 |
CC-MAIN-2026-08,551,19551,2609034951,89980913,29167227,2026-03-13T06:42:51Z,280,0,20,176
|
| 398 |
CC-MAIN-2026-08,552,19402,2594817630,94092761,30309254,2026-03-13T06:48:30Z,211,0,46,80
|
| 399 |
CC-MAIN-2026-08,553,19856,2617133708,99106486,30365588,2026-03-13T06:52:51Z,116,0,23,120
|
|
|
|
| 394 |
CC-MAIN-2026-08,531,19460,2553022933,93648476,30171716,2026-03-13T22:53:41Z,609,0,162,8
|
| 395 |
CC-MAIN-2026-08,532,19118,2543492135,93810142,30041771,2026-03-13T23:04:41Z,525,0,136,15
|
| 396 |
CC-MAIN-2026-08,533,19410,2555040866,91016286,29048349,2026-03-13T23:17:18Z,579,0,169,23
|
| 397 |
+
CC-MAIN-2026-08,534,19426,2574915686,94390386,29713409,2026-03-13T23:28:41Z,602,0,93,0
|
| 398 |
CC-MAIN-2026-08,551,19551,2609034951,89980913,29167227,2026-03-13T06:42:51Z,280,0,20,176
|
| 399 |
CC-MAIN-2026-08,552,19402,2594817630,94092761,30309254,2026-03-13T06:48:30Z,211,0,46,80
|
| 400 |
CC-MAIN-2026-08,553,19856,2617133708,99106486,30365588,2026-03-13T06:52:51Z,116,0,23,120
|