Update README.md

README.md

@@ -74,11 +74,11 @@ Given than training models of size 7 Billion parameters require lot more compute
 
 <img src="fig12.jpg" alt="fig12.jpg" style="width:1000px;"/>
 
-**Figure7:**  Comparison of Average Eval Scores on High Signal and Extended Eval Tasks at 7B model size.  Scores are averaged over 3 random seeds used for data sampling and are reported along with standard deviations.
+**Figure 5:**  Comparison of Average Eval Scores on High Signal and Extended Eval Tasks at 7B model size.  Scores are averaged over 3 random seeds used for data sampling and are reported along with standard deviations.
 
 <img src="Fig18.png" alt="Fig18.png" style="width:1000px;"/>
 
-**Figure 8:** Average evaluation score on High-Signal tasks versus the number of tokens at 7 Billion model size for 350 Billion tokens. The model trained on GneissWeb consistently outperforms the one trained on FineWeb.V1.1 throughout the training.
+**Figure 6:** Average evaluation score on High-Signal tasks versus the number of tokens at 7 Billion model size for 350 Billion tokens. The model trained on GneissWeb consistently outperforms the one trained on FineWeb.V1.1 throughout the training.
 
 **Dataset Summary**