readme

README.md

@@ -72,13 +72,12 @@ The format is the following:
 - "top1k_offsets": Passage ID (int) when the numpy matrices are loaded sequentially and vertically stacked
 - "top1k_passage_ids": Passage ID (string) as they appear in the dataset 
 - "top1k_cossim": Cosine similarities
-- "qrels": Relevance annotations for the 215 annotated queries by NIST
+- "qrels": Relevance annotations for the 215 annotated queries by NIST. The **document relevance** scores are provided. You can get the doc_id for a passage via `row['_id'].split("#")[0]` 
 
 ### Queries - JSONL
 The folder `queries_jsonl/` contains the queries in a `.jsonl.gz` format.
 
-Load this .jsonl.gz format is significantly faster than the parquet file.
-
+Note: qrels are provided here as a dictionary lookup, while in the parquet format as a list in the format `[doc_id, score]` due to the limited support for dictionaries in parquet.
 
 ### Queries - Parquet