Datasets:

Modalities:
Text
Formats:
parquet
ArXiv:
Libraries:
Datasets
pandas
License:
benjamsz commited on
Commit
01b8500
·
verified ·
1 Parent(s): 592509a

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -16
README.md CHANGED
@@ -58,21 +58,11 @@ dataset_info:
58
 
59
  ## Overview
60
 
61
- **watsonxDocsQA** is a new open-source dataset and benchmark derived from enterprise product documentation. Designed specifically for end-to-end Retrieval-Augmented Generation (RAG) evaluation, it consists of two key components:
62
-
63
- - **Documents**: A corpus of 1,144 text and markdown files generated by crawling enterprise documentation (link to the documentation source : [https://dataplatform.cloud.ibm.com/docs/content/wsj/getting-started/welcome-main.html](#)).
64
- - **Benchmark**: A set of 75 question-answer (QA) pairs with gold document labels and answers.
65
-
66
- The QA pairs are crafted as follows:
67
- - **25 questions**: Human-generated by two subject matter experts.
68
- - **50 questions**: Synthetically generated using the `tiiuae/falcon-180b` model, then manually filtered and reviewed for quality. The methodology is detailed in [this paper](https://arxiv.org/pdf/2401.14367).
69
-
70
- ### Key Details
71
- - **Number of Samples**:
72
- - Corpus: 1,144 documents
73
- - QA Set: 75 questions
74
- - **Format**: CSV
75
- - **License**: [Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)
76
 
77
  ---
78
 
@@ -104,7 +94,6 @@ The QA dataset includes these fields:
104
 
105
  ## Samples
106
 
107
- ---
108
 
109
  Below is an example from the `question_answers` dataset:
110
 
 
58
 
59
  ## Overview
60
 
61
+ **watsonxDocsQA** is a new open-source dataset and benchmark contributed by IBM. The dataset is derived from enterprise product documentation and designed specifically for end-to-end Retrieval-Augmented Generation (RAG) evaluation. The dataset consists of two components:
62
+ - **Documents**: A corpus of 1,144 text and markdown files generated by crawling enterprise documentation ([main page](https://dataplatform.cloud.ibm.com/docs/content/wsj/getting-started/welcome-main.html)).
63
+ - **Benchmark**: A set of 75 question-answer (QA) pairs with gold document labels and answers.The QA pairs are crafted as follows:
64
+ - **25 questions**: Human-generated by two subject matter experts.
65
+ - **50 questions**: Synthetically generated using the `tiiuae/falcon-180b` model, then manually filtered and reviewed for quality. The methodology is detailed in [Yehudai et al. 2024](https://arxiv.org/pdf/2401.14367).
 
 
 
 
 
 
 
 
 
 
66
 
67
  ---
68
 
 
94
 
95
  ## Samples
96
 
 
97
 
98
  Below is an example from the `question_answers` dataset:
99