Publish shard CC-MAIN-2026-08/00214
Browse files- README.md +9 -9
- data/CC-MAIN-2026-08/00214.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) | ~47.8 GB | ~3.7 TB | **-98.3%** vs HTML |
|
| 174 |
| Final Parquet (Zstd level 19) | 32.3 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 32m 26s 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 **22,562,981 documents across 1168 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 1168 files of CC-MAIN-2026-08 (22,562,981 pages total), projected to the full crawl of 100,000 WARC files.
|
| 168 |
|
| 169 |
+
| Stage | 1168 files (measured) | 100,000 files (projected) | Reduction |
|
| 170 |
|---|---|---|---|
|
| 171 |
+
| Raw WARC (.warc.gz, downloaded) | ~946.7 GB | ~83 TB | — |
|
| 172 |
| HTML extracted (uncompressed) | 2.7 TB | ~295 TB | — |
|
| 173 |
| Packed markdown WARC (.md.warc.gz) | ~47.8 GB | ~3.7 TB | **-98.3%** vs HTML |
|
| 174 |
| Final Parquet (Zstd level 19) | 32.3 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 101.8 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: ~946.7 GB of raw gzipped WARCs becomes **32.3 GB of Parquet** — a **96.6% total reduction** — containing 22,562,981 clean markdown documents.
|
| 179 |
|
| 180 |
### Processing Times
|
| 181 |
|
| 182 |
+
Pipeline timings across 1168 shards of CC-MAIN-2026-08:
|
| 183 |
|
| 184 |
```
|
| 185 |
+
Download (raw WARC) ████████████████████████ total 88h 35m 14s avg 4m 33s
|
| 186 |
Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 5h 55m 4s avg 18s
|
| 187 |
+
Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 21h 23m 50s avg 1m 5s
|
| 188 |
Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 7h 32m 26s avg 23s
|
| 189 |
```
|
| 190 |
|
data/CC-MAIN-2026-08/00214.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fa10d8588b7a3af7e1b38ed2688f72c2253fd4bed1c12a34e7add41940c0580c
|
| 3 |
+
size 31239634
|
stats.csv
CHANGED
|
@@ -214,6 +214,7 @@ CC-MAIN-2026-08,210,19330,2559492399,92707630,28965419,2026-03-14T01:12:35Z,662,
|
|
| 214 |
CC-MAIN-2026-08,211,12344,1513255589,64363121,20146574,2026-03-14T01:23:01Z,0,0,22,13
|
| 215 |
CC-MAIN-2026-08,212,19279,2520135948,92429416,30474894,2026-03-14T01:31:35Z,431,0,99,13
|
| 216 |
CC-MAIN-2026-08,213,19264,2564809177,94197421,29007844,2026-03-14T01:44:35Z,551,0,185,14
|
|
|
|
| 217 |
CC-MAIN-2026-08,250,19200,2541559671,94959737,30214413,2026-03-13T17:42:47Z,511,0,85,19
|
| 218 |
CC-MAIN-2026-08,251,19610,2594898345,97558492,31917087,2026-03-13T06:42:12Z,277,0,23,136
|
| 219 |
CC-MAIN-2026-08,252,19240,2525928701,98163046,31104034,2026-03-13T06:47:10Z,238,0,49,10
|
|
|
|
| 214 |
CC-MAIN-2026-08,211,12344,1513255589,64363121,20146574,2026-03-14T01:23:01Z,0,0,22,13
|
| 215 |
CC-MAIN-2026-08,212,19279,2520135948,92429416,30474894,2026-03-14T01:31:35Z,431,0,99,13
|
| 216 |
CC-MAIN-2026-08,213,19264,2564809177,94197421,29007844,2026-03-14T01:44:35Z,551,0,185,14
|
| 217 |
+
CC-MAIN-2026-08,214,19510,2565062520,97415953,31239634,2026-03-14T01:54:46Z,451,0,157,0
|
| 218 |
CC-MAIN-2026-08,250,19200,2541559671,94959737,30214413,2026-03-13T17:42:47Z,511,0,85,19
|
| 219 |
CC-MAIN-2026-08,251,19610,2594898345,97558492,31917087,2026-03-13T06:42:12Z,277,0,23,136
|
| 220 |
CC-MAIN-2026-08,252,19240,2525928701,98163046,31104034,2026-03-13T06:47:10Z,238,0,49,10
|