Publish shard CC-MAIN-2026-08/01029
Browse files- README.md +10 -10
- data/CC-MAIN-2026-08/01029.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 **17,
|
| 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) | ~736.
|
| 172 |
| HTML extracted (uncompressed) | 2.1 TB | ~295 TB | — |
|
| 173 |
-
| Packed markdown WARC (.md.warc.gz) | ~37.
|
| 174 |
| Final Parquet (Zstd level 19) | 25.2 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 2.1 TB of uncompressed HTML down to 79.
|
| 177 |
|
| 178 |
-
End to end: ~736.
|
| 179 |
|
| 180 |
### Processing Times
|
| 181 |
|
| 182 |
-
Pipeline timings across
|
| 183 |
|
| 184 |
```
|
| 185 |
-
Download (raw WARC) ████████████████████████ total 57h
|
| 186 |
Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 4h 30m 54s avg 17s
|
| 187 |
-
Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 13h
|
| 188 |
Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 6h 33m 7s 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 **17,595,283 documents across 909 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 909 files of CC-MAIN-2026-08 (17,595,283 pages total), projected to the full crawl of 100,000 WARC files.
|
| 168 |
|
| 169 |
+
| Stage | 909 files (measured) | 100,000 files (projected) | Reduction |
|
| 170 |
|---|---|---|---|
|
| 171 |
+
| Raw WARC (.warc.gz, downloaded) | ~736.8 GB | ~83 TB | — |
|
| 172 |
| HTML extracted (uncompressed) | 2.1 TB | ~295 TB | — |
|
| 173 |
+
| Packed markdown WARC (.md.warc.gz) | ~37.3 GB | ~3.7 TB | **-98.3%** vs HTML |
|
| 174 |
| Final Parquet (Zstd level 19) | 25.2 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 2.1 TB of uncompressed HTML down to 79.3 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: ~736.8 GB of raw gzipped WARCs becomes **25.2 GB of Parquet** — a **96.6% total reduction** — containing 17,595,283 clean markdown documents.
|
| 179 |
|
| 180 |
### Processing Times
|
| 181 |
|
| 182 |
+
Pipeline timings across 909 shards of CC-MAIN-2026-08:
|
| 183 |
|
| 184 |
```
|
| 185 |
+
Download (raw WARC) ████████████████████████ total 57h 28m 57s avg 3m 47s
|
| 186 |
Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 4h 30m 54s avg 17s
|
| 187 |
+
Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 13h 5m 40s avg 51s
|
| 188 |
Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 6h 33m 7s avg 25s
|
| 189 |
```
|
| 190 |
|
data/CC-MAIN-2026-08/01029.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b8e74f636b49d0dfd84558d3545616b2430001fcac51c06306355b420f37938
|
| 3 |
+
size 30396446
|
stats.csv
CHANGED
|
@@ -561,6 +561,7 @@ CC-MAIN-2026-08,1025,19227,2537502059,92746222,29545444,2026-03-13T20:10:04Z,478
|
|
| 561 |
CC-MAIN-2026-08,1026,19623,2563586507,92866472,29527517,2026-03-13T20:19:09Z,499,0,82,15
|
| 562 |
CC-MAIN-2026-08,1027,19469,2558334483,99999457,31018064,2026-03-13T20:31:26Z,589,0,157,21
|
| 563 |
CC-MAIN-2026-08,1028,19557,2589891377,92034631,28857038,2026-03-13T20:42:40Z,429,0,213,13
|
|
|
|
| 564 |
CC-MAIN-2026-08,1101,19721,2641197765,95719568,30240239,2026-03-13T09:59:33Z,564,0,296,0
|
| 565 |
CC-MAIN-2026-08,1102,19562,2608129691,99280759,31748974,2026-03-13T10:28:07Z,1518,0,191,0
|
| 566 |
CC-MAIN-2026-08,1103,19270,2517580674,89253461,28361470,2026-03-13T10:55:39Z,1380,0,268,0
|
|
|
|
| 561 |
CC-MAIN-2026-08,1026,19623,2563586507,92866472,29527517,2026-03-13T20:19:09Z,499,0,82,15
|
| 562 |
CC-MAIN-2026-08,1027,19469,2558334483,99999457,31018064,2026-03-13T20:31:26Z,589,0,157,21
|
| 563 |
CC-MAIN-2026-08,1028,19557,2589891377,92034631,28857038,2026-03-13T20:42:40Z,429,0,213,13
|
| 564 |
+
CC-MAIN-2026-08,1029,19460,2551984186,97459661,30396446,2026-03-13T20:54:20Z,466,0,257,0
|
| 565 |
CC-MAIN-2026-08,1101,19721,2641197765,95719568,30240239,2026-03-13T09:59:33Z,564,0,296,0
|
| 566 |
CC-MAIN-2026-08,1102,19562,2608129691,99280759,31748974,2026-03-13T10:28:07Z,1518,0,191,0
|
| 567 |
CC-MAIN-2026-08,1103,19270,2517580674,89253461,28361470,2026-03-13T10:55:39Z,1380,0,268,0
|