Update README.md

README.md

@@ -24,7 +24,7 @@ This dataset repo contains only the queries and relevance judgements. The corpus
 
 
 # Dataset Structure
-To download the files: The queries can be found under `ciral-{lang}/topics` and are in `.tsv` formats with each line in the form:
+1. To download the files: The queries can be found under `ciral-{lang}/topics` and are in `.tsv` formats with each line in the form:
 
    ```
     qid\tquery
@@ -34,7 +34,25 @@ To download the files: The queries can be found under `ciral-{lang}/topics` and
    ```
     qid Q0 docid relevance
    ```
+2. To access the dataset via `datasets`:
+   
+   ```
+   ciral_dataset = load_datasets("ciral/ciral", lang="hausa") #or swahili, somali, yoruba
+
+   for data in ciral_dataset['train']:  # or 'test'
+     query_id = data['query_id']
+     query = data['query']
+     qrels = data['judgements']
+
+     for qrel in qrels:
+       docid = qrel['docid']
+       judgement = qrel['judgement']
+       text = qrel['text']
+   ```
+
+
 
+   
 ## Citation
 
 ...