Upload folder using huggingface_hub

README.md

@@ -85,8 +85,8 @@ valuable for pre-training of medical LLMS.
 #### Initial Data Collection and Normalization
 
 Langchain's RecursiveCharacterTextSplitter is used for chunking and the most commonly-appearing non-ASCII characters 
-are replaced with readable equivalents.   The textbooks are then broken into separate subsets, indicated below along with 
-the textbooks they comprise: 
+are replaced with readable equivalents.   Chunks comprising less than 90% ASCII characters were excluded.  The textbooks 
+were then broken into separate subsets, indicated below along with the textbook source(s) they comprise: 
 
 - Core Clinical Medicine (_*core_clinical*_)
   - Anatomy_Gray.txt, First_Aid_Step1.txt, First_Aid_Step2.txt, Immunology_Janeway.txt, InternalMed_Harrison.txt, Neurology_Adams.txt, Obstentrics_Williams.txt, Pathoma_Husain.txt, Pediatrics_Nelson.txt, and Surgery_Schwartz.txt