Update README.md

README.md

@@ -24,17 +24,16 @@ Asma Ben Abacha, Eugene Agichtein, Yuval Pinter & Dina Demner-Fushman. Overview
 
 ## Medical Training Data
 
-Currently, this part of the dataset is not integrated with huggingface.
-You can download the pre-processed jsonl that I have uploaded in the files tab instead.
-
 The dataset provides 634 question-answer pairs for training: 
 
 	1) TREC-2017-LiveQA-Medical-Train-1.xml => 388 question-answer pairs corresponding to 200 NLM questions. 
 	Each question is divided into one or more subquestion(s). Each subquestion has one or more answer(s). 
-	These question-answer pairs were constructed automatically and validated manually.
+	These question-answer pairs were constructed automatically and validated manually. 
 
 	2) TREC-2017-LiveQA-Medical-Train-2.xml => 246 question-answer pairs corresponding to 246 NLM questions.
-	Answers were retrieved manually by librarians. 
+	Answers were retrieved manually by librarians.
+
+**You can access Train 1 dataset by requesting the `train` split. Train 2 is the `dev` split by just convenience so you can request the `dev` split to download.**
 
 The datasets are not exhaustive with regards to subquestions, i.e., some subquestions might not be annotated. 
 Additional annotations are provided for both (i) the Focus and (ii) the Question Type used to define each subquestion.