1 of 2 TODOs

Noticed 2 TODOs here. I think one of them means to link to https://huggingface.co/spaces/HuggingFaceFW/blogpost-fineweb-v1 ? Not sure if the other does as well.

README.md

@@ -544,7 +544,7 @@ We fine-tuned a Bert-like regression model using these annotations, based on [Sn
 The classifier is available at: [https://huggingface.co/HuggingFaceFW/fineweb-edu-classifier/ ](https://huggingface.co/HuggingFaceFW/fineweb-edu-classifier/)
 
 ### Filtering and results
-**Note**: You can find more details about the ablations and results in the FineWeb blog post (TODO).
+**Note**: You can find more details about the ablations and results in [the FineWeb blog post](https://huggingface.co/spaces/HuggingFaceFW/blogpost-fineweb-v1).
 
 We investigated the impact of using different thresholds for the filtering and found that threshold 3 gave the best overall results. Although using a threshold higher than 3 improves performance on knowledge and reasoning intensive benchmarks, it significantly degrades performance on HellaSwag and PIQA.