Update README.md

README.md

@@ -225,8 +225,7 @@ Every mid-training config has the same columns:
 | Total | 514.30B | 100% | 302.0B | |
 
 
-Carbon-X trained on 302B of these 514.30B tokens. Long-DNA biology text was read about 7 times and
-instruction data about 5; of English, prokaryote and mRNA-splice the run read only a part.
+Carbon-X trained on 302B of these 514.30B tokens. 
 
 | config | documents | tokens | share of training (weight) | tokens read by Carbon-X | passes |
 |---|---:|---:|---:|---:|---:|