Refresh evaluation dataset attribution link

ATTRIBUTIONS.md

@@ -32,4 +32,4 @@ Aparecium extracts short headlines, summaries, and sentences for training. The r
 
 ## Evaluation sources
 
-Evaluation includes held-out texts from the training sources, CNN/DailyMail and XSum summaries, BBC news descriptions, social posts, Q&A titles, and crypto headlines. The independent similarity threshold is calibrated using [STS Benchmark](https://ixa2.si.ehu.eus/stswiki/index.php/STSbenchmark). Aggregate scores and illustrative reconstructions do not change the licenses of those sources.
+Evaluation includes held-out texts from the training sources, CNN/DailyMail and XSum summaries, BBC news descriptions, social posts, Q&A titles, and crypto headlines. The independent similarity threshold is calibrated using [STS Benchmark](https://huggingface.co/datasets/sentence-transformers/stsb). Aggregate scores and illustrative reconstructions do not change the licenses of those sources.

SHA256SUMS

@@ -1,4 +1,4 @@
-225d632d0ba94d83c4a9dc710967044da4bda1e37847bbb36e6527c8c7db3899  ATTRIBUTIONS.md
+c3150eb2b9cb9bf5f8f686d49a807734f6fb38274aba350172fb6b7465ca1e68  ATTRIBUTIONS.md
 5b00bcb8a6dc9942d8256909fe3690361786da5e0dd2d78b9eb6db8d11a08e25  LICENSE
 a59a4fdf880561e2c40c961e0239fd8587d0857d44b6748c0a483ed03bc855f3  README.md
 0a494d16cd47b9ccfd3406926898fb619e6b811815d8713790c2e2fccd9b2028  adapter/adapter_config.json