Datasets:

Modalities:
Text
Formats:
parquet
Libraries:
Datasets
pandas
mlconti commited on
Commit
df93738
·
1 Parent(s): 3abbe73
Files changed (1) hide show
  1. README.md +1 -1
README.md CHANGED
@@ -61,7 +61,7 @@ The dataset is structured into the following columns:
61
  * **`queries`**: Contains query information:
62
  * `"query"`: The text of the query.
63
  * `"answer"`: The answer relevant to the query, from the original dataset.
64
- * `"chunk_ids"`: A list of chunk IDs that are relevant to the query. This is used to link the query to the relevant chunks in the `documents` dataset.
65
 
66
  ## Usage
67
 
 
61
  * **`queries`**: Contains query information:
62
  * `"query"`: The text of the query.
63
  * `"answer"`: The answer relevant to the query, from the original dataset.
64
+ * `"chunk_id"`: The ID of the chunk that the query is related to, of the form `doc-id_chunk-id`, where `doc-id` is the ID of the original document and `chunk-id` is the position of the chunk within that document.
65
 
66
  ## Usage
67