Update README.md

README.md

@@ -242,9 +242,9 @@ configs:
 
 **Highly multilingual, large-scale paired speech corpus assembled from primary public archives such as national parliaments, public broadcasters, and public-domain audiobooks across 107 language-country configs, with strong non-European and low-resource coverage.**
 
-Every segment ships with a verified *human* transcript (not only ASR pseudo-labels), a Whisper ASR transcript, the CER between them, a WADA-SNR estimate, and four DNSMOS-P.835 quality scores, so you can filter, rank, or stratify training data without rerunning a single signal-quality model.
+Every segment ships with a verified *human* transcript (not ASR pseudo-labels), an ASR transcript, the CER between them, a WADA-SNR estimate, and four DNSMOS-P.835 quality scores, so you can filter, rank, or stratify training data without rerunning a single signal-quality model.
 
-Audio is OGG Opus 24 kHz mono ~64 kbps VBR. Splits are 95/5 with seed 42; full provenance (`source`, `source_url`, `session_date`, `source_start_s`/`source_end_s`, `segment_id`) is preserved on every row, so you can trace any clip back to its original parliamentary record or broadcast.
+Audio is OGG Opus 24 kHz mono ~64 kbps VBR. Splits are 95/5, full provenance (`source`, `source_url`, `session_date`, `source_start_s`/`source_end_s`, `segment_id`) is preserved on every row, so you can trace any clip back to its original parliamentary record or broadcast.
 
 ## Loading