Publish shard CC-MAIN-2026-08/00795
Browse files- README.md +9 -9
- data/CC-MAIN-2026-08/00795.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 **22,
|
| 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) | 2.7 TB | ~295 TB | — |
|
| 173 |
| Packed markdown WARC (.md.warc.gz) | ~47.7 GB | ~3.7 TB | **-98.3%** vs HTML |
|
| 174 |
| Final Parquet (Zstd level 19) | 32.2 GB | ~2.9 TB | **-32.5%** 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 2.7 TB of uncompressed HTML down to 101.
|
| 177 |
|
| 178 |
-
End to end: ~
|
| 179 |
|
| 180 |
### Processing Times
|
| 181 |
|
| 182 |
-
Pipeline timings across
|
| 183 |
|
| 184 |
```
|
| 185 |
-
Download (raw WARC) ████████████████████████ total 88h
|
| 186 |
Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 5h 55m 4s avg 18s
|
| 187 |
-
Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 21h
|
| 188 |
Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 7h 31m 45s avg 23s
|
| 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 **22,504,900 documents across 1165 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 1165 files of CC-MAIN-2026-08 (22,504,900 pages total), projected to the full crawl of 100,000 WARC files.
|
| 168 |
|
| 169 |
+
| Stage | 1165 files (measured) | 100,000 files (projected) | Reduction |
|
| 170 |
|---|---|---|---|
|
| 171 |
+
| Raw WARC (.warc.gz, downloaded) | ~944.3 GB | ~83 TB | — |
|
| 172 |
| HTML extracted (uncompressed) | 2.7 TB | ~295 TB | — |
|
| 173 |
| Packed markdown WARC (.md.warc.gz) | ~47.7 GB | ~3.7 TB | **-98.3%** vs HTML |
|
| 174 |
| Final Parquet (Zstd level 19) | 32.2 GB | ~2.9 TB | **-32.5%** 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 2.7 TB of uncompressed HTML down to 101.5 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 68.3%.
|
| 177 |
|
| 178 |
+
End to end: ~944.3 GB of raw gzipped WARCs becomes **32.2 GB of Parquet** — a **96.6% total reduction** — containing 22,504,900 clean markdown documents.
|
| 179 |
|
| 180 |
### Processing Times
|
| 181 |
|
| 182 |
+
Pipeline timings across 1165 shards of CC-MAIN-2026-08:
|
| 183 |
|
| 184 |
```
|
| 185 |
+
Download (raw WARC) ████████████████████████ total 88h 12m 44s avg 4m 32s
|
| 186 |
Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 5h 55m 4s avg 18s
|
| 187 |
+
Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 21h 18m 37s avg 1m 5s
|
| 188 |
Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 7h 31m 45s avg 23s
|
| 189 |
```
|
| 190 |
|
data/CC-MAIN-2026-08/00795.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02296d809aa235fecef25e9b8520ad19ce9cded2fba7735414dc8e45d5590b47
|
| 3 |
+
size 30888077
|
stats.csv
CHANGED
|
@@ -569,6 +569,7 @@ CC-MAIN-2026-08,791,19549,2595677953,94842727,29594346,2026-03-14T01:12:35Z,654,
|
|
| 569 |
CC-MAIN-2026-08,792,10720,1240933387,57370777,17474176,2026-03-14T01:23:01Z,0,0,23,13
|
| 570 |
CC-MAIN-2026-08,793,19518,2529274504,93017510,29938743,2026-03-14T01:30:07Z,346,0,88,15
|
| 571 |
CC-MAIN-2026-08,794,19336,2556055535,89907616,28944770,2026-03-14T01:41:55Z,395,0,220,16
|
|
|
|
| 572 |
CC-MAIN-2026-08,851,19575,2612873485,97000364,30911294,2026-03-13T06:41:39Z,163,0,37,204
|
| 573 |
CC-MAIN-2026-08,852,19334,2546015208,94959847,30205044,2026-03-13T06:45:52Z,183,0,57,11
|
| 574 |
CC-MAIN-2026-08,853,19694,2567973975,95901932,30303494,2026-03-13T06:50:12Z,150,0,28,80
|
|
|
|
| 569 |
CC-MAIN-2026-08,792,10720,1240933387,57370777,17474176,2026-03-14T01:23:01Z,0,0,23,13
|
| 570 |
CC-MAIN-2026-08,793,19518,2529274504,93017510,29938743,2026-03-14T01:30:07Z,346,0,88,15
|
| 571 |
CC-MAIN-2026-08,794,19336,2556055535,89907616,28944770,2026-03-14T01:41:55Z,395,0,220,16
|
| 572 |
+
CC-MAIN-2026-08,795,19307,2538738629,96815330,30888077,2026-03-14T01:50:50Z,444,0,175,0
|
| 573 |
CC-MAIN-2026-08,851,19575,2612873485,97000364,30911294,2026-03-13T06:41:39Z,163,0,37,204
|
| 574 |
CC-MAIN-2026-08,852,19334,2546015208,94959847,30205044,2026-03-13T06:45:52Z,183,0,57,11
|
| 575 |
CC-MAIN-2026-08,853,19694,2567973975,95901932,30303494,2026-03-13T06:50:12Z,150,0,28,80
|