Update README.md
Browse files
README.md
CHANGED
|
@@ -48,6 +48,4 @@ Added **2,345** posts (**2,345** images) from [the public Telegram channel](http
|
|
| 48 |
|
| 49 |
The existing Parquet files and examples are preserved. Schema remains `id`, `description`, `photos` (a large list of embedded images). Descriptions are channel annotations, with HTML formatting removed and line breaks preserved. New images are Telegram public web previews; their resolution may differ from original Telegram files. Posts require both a nonempty description and at least one valid photo. Deduplication uses Telegram post IDs; reposted images with distinct annotations are retained.
|
| 50 |
|
| 51 |
-
Excluded 15 posts without text/photos; deferred 4 posts with unavailable media ([retry inventory](deferred.jsonl)). The crawl reached the previous maximum message ID 6445. Deleted/unavailable posts, older gaps, and future posts are outside this update. See [audit](audit.json) for exclusions, counts, hashes and coverage, and [provenance](provenance.jsonl) for source URLs, timestamps and image hashes. The scraper is maintained locally, outside this dataset repository.
|
| 52 |
-
|
| 53 |
-
The repository's existing Apache-2.0 metadata is retained. This does not establish ownership or permissions for third-party meme images and underlying works.
|
|
|
|
| 48 |
|
| 49 |
The existing Parquet files and examples are preserved. Schema remains `id`, `description`, `photos` (a large list of embedded images). Descriptions are channel annotations, with HTML formatting removed and line breaks preserved. New images are Telegram public web previews; their resolution may differ from original Telegram files. Posts require both a nonempty description and at least one valid photo. Deduplication uses Telegram post IDs; reposted images with distinct annotations are retained.
|
| 50 |
|
| 51 |
+
Excluded 15 posts without text/photos; deferred 4 posts with unavailable media ([retry inventory](deferred.jsonl)). The crawl reached the previous maximum message ID 6445. Deleted/unavailable posts, older gaps, and future posts are outside this update. See [audit](audit.json) for exclusions, counts, hashes and coverage, and [provenance](provenance.jsonl) for source URLs, timestamps and image hashes. The scraper is maintained locally, outside this dataset repository.
|
|
|
|
|
|