tamnd commited on
Commit
977eb80
·
verified ·
1 Parent(s): c9e08c9

Publish shard CC-MAIN-2026-08/01295

Browse files
Files changed (3) hide show
  1. README.md +11 -11
  2. data/CC-MAIN-2026-08/01295.parquet +3 -0
  3. 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 **20,185,472 documents across 1045 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
 
@@ -164,27 +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 1045 files of CC-MAIN-2026-08 (20,185,472 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
- | Stage | 1045 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
- | Raw WARC (.warc.gz, downloaded) | ~847.0 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 2.4 TB | ~295 TB | — |
173
- | Packed markdown WARC (.md.warc.gz) | ~42.8 GB | ~3.7 TB | **-98.3%** vs HTML |
174
- | Final Parquet (Zstd level 19) | 28.9 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 2.4 TB of uncompressed HTML down to 91.1 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: ~847.0 GB of raw gzipped WARCs becomes **28.9 GB of Parquet** — a **96.6% total reduction** — containing 20,185,472 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
- Pipeline timings across 1045 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
- Download (raw WARC) ████████████████████████ total 77h 15m 3s avg 4m 26s
186
  Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 5h 55m 4s avg 20s
187
- Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 18h 26m 46s avg 1m 3s
188
  Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 7h 6m 34s avg 24s
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 **20,205,093 documents across 1046 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 1046 files of CC-MAIN-2026-08 (20,205,093 pages total), projected to the full crawl of 100,000 WARC files.
168
 
169
+ | Stage | 1046 files (measured) | 100,000 files (projected) | Reduction |
170
  |---|---|---|---|
171
+ | Raw WARC (.warc.gz, downloaded) | ~847.8 GB | ~83 TB | — |
172
  | HTML extracted (uncompressed) | 2.4 TB | ~295 TB | — |
173
+ | Packed markdown WARC (.md.warc.gz) | ~42.9 GB | ~3.7 TB | **-98.3%** vs HTML |
174
+ | Final Parquet (Zstd level 19) | 29.0 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 2.4 TB of uncompressed HTML down to 91.2 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: ~847.8 GB of raw gzipped WARCs becomes **29.0 GB of Parquet** — a **96.6% total reduction** — containing 20,205,093 clean markdown documents.
179
 
180
  ### Processing Times
181
 
182
+ Pipeline timings across 1046 shards of CC-MAIN-2026-08:
183
 
184
  ```
185
+ Download (raw WARC) ████████████████████████ total 77h 22m 0s avg 4m 26s
186
  Convert (HTML → MD) █░░░░░░░░░░░░░░░░░░░░░░░ total 5h 55m 4s avg 20s
187
+ Export (Parquet) █████░░░░░░░░░░░░░░░░░░░ total 18h 30m 52s avg 1m 3s
188
  Publish (HuggingFace) ██░░░░░░░░░░░░░░░░░░░░░░ total 7h 6m 34s avg 24s
189
  ```
190
 
data/CC-MAIN-2026-08/01295.parquet ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b6d2a5b1d5df2eb9027795f5f2f9acb52f0f4617f79eddb8f87ebd0d9f9ea7af
3
+ size 30555936
stats.csv CHANGED
@@ -778,6 +778,7 @@ CC-MAIN-2026-08,1291,19269,2541012693,92877966,29210276,2026-03-14T01:06:46Z,803
778
  CC-MAIN-2026-08,1292,19443,2585178785,93531562,29428217,2026-03-14T01:06:46Z,729,0,84,5
779
  CC-MAIN-2026-08,1293,494,37594543,4081674,1141373,2026-03-14T01:06:46Z,280,0,3,5
780
  CC-MAIN-2026-08,1294,2146,186927156,18961900,5265842,2026-03-14T01:11:50Z,265,0,119,5
 
781
  CC-MAIN-2026-08,1301,19417,2601765043,93758124,29790521,2026-03-13T11:40:27Z,0,0,0,0
782
  CC-MAIN-2026-08,1302,19471,2592470194,95488689,29747494,2026-03-13T11:40:27Z,0,0,0,0
783
  CC-MAIN-2026-08,1303,19681,2609544680,95027524,30250671,2026-03-13T11:40:27Z,0,0,0,0
 
778
  CC-MAIN-2026-08,1292,19443,2585178785,93531562,29428217,2026-03-14T01:06:46Z,729,0,84,5
779
  CC-MAIN-2026-08,1293,494,37594543,4081674,1141373,2026-03-14T01:06:46Z,280,0,3,5
780
  CC-MAIN-2026-08,1294,2146,186927156,18961900,5265842,2026-03-14T01:11:50Z,265,0,119,5
781
+ CC-MAIN-2026-08,1295,19621,2593972556,98457838,30555936,2026-03-14T01:20:01Z,417,0,246,0
782
  CC-MAIN-2026-08,1301,19417,2601765043,93758124,29790521,2026-03-13T11:40:27Z,0,0,0,0
783
  CC-MAIN-2026-08,1302,19471,2592470194,95488689,29747494,2026-03-13T11:40:27Z,0,0,0,0
784
  CC-MAIN-2026-08,1303,19681,2609544680,95027524,30250671,2026-03-13T11:40:27Z,0,0,0,0