Datasets:
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
1M - 10M
Tags:
conversational
code
software-engineering
troubleshooting
technical-discussions
developer-discussions
License:
docs: clarify v0.7 corpus and metadata semantics
Browse filesMark corpus_rag as the primary content corpus and distinguish interaction metadata from factual labels.
v0.7.0-multisource-merged/README.md
CHANGED
|
@@ -32,25 +32,34 @@ configs:
|
|
| 32 |
path: views/issue_resolution_multisource_merged_v0.1/validation.parquet
|
| 33 |
---
|
| 34 |
|
| 35 |
-
# Community QA Lab v0.7.0 multisource merged —
|
| 36 |
-
|
| 37 |
-
This
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
|
| 42 |
## Merged views
|
| 43 |
|
| 44 |
-
- preference_dpo_multisource_merged_v0.1: 7453 directed pairs.
|
| 45 |
-
|
| 46 |
-
|
| 47 |
- reranking_listwise_multisource_merged_v0.1: 4550 groups.
|
| 48 |
Candidate metadata is retained for evaluation but excluded from the
|
| 49 |
model projection.
|
| 50 |
- corpus_rag_multisource_merged_v0.1: 39215 licensed documents.
|
| 51 |
document_text contains question context and post text only.
|
| 52 |
-
- issue_resolution_multisource_merged_v0.1: 1823 observed
|
| 53 |
-
issue/PR lifecycle records; these are not preference targets.
|
|
|
|
|
|
|
|
|
|
|
|
|
| 54 |
|
| 55 |
## Split counts
|
| 56 |
|
|
@@ -71,8 +80,8 @@ GitHub/GitLab lifecycle evidence. It has not been uploaded or pushed.
|
|
| 71 |
no public-text rows.
|
| 72 |
- Accepted, vote/reaction, timestamps, URLs and provenance fields
|
| 73 |
remain outside model-input projections.
|
| 74 |
-
- This is an experimental
|
| 75 |
-
contamination-free factuality benchmark.
|
| 76 |
|
| 77 |
The exact source manifest hashes, split adjustments, deduplication
|
| 78 |
counts and Parquet hashes are in manifest.json.
|
|
|
|
| 32 |
path: views/issue_resolution_multisource_merged_v0.1/validation.parquet
|
| 33 |
---
|
| 34 |
|
| 35 |
+
# Community QA Lab v0.7.0 multisource merged — corpus-first experimental release
|
| 36 |
+
|
| 37 |
+
This bundle combines the immutable v0.6 Stack Exchange derivative with the
|
| 38 |
+
v0.7 Stack Exchange expansion, GitHub Discussions, GitLab retrieval rows and
|
| 39 |
+
GitHub/GitLab lifecycle evidence. The primary artifact is the licensed
|
| 40 |
+
community-content corpus `corpus_rag_multisource_merged_v0.1`.
|
| 41 |
+
|
| 42 |
+
The corpus contains text and provenance for NLP/RAG. It is not a factuality
|
| 43 |
+
benchmark. Platform interaction fields such as accepted/marked-answer status
|
| 44 |
+
are observations about how a community or author interacted with a post; they
|
| 45 |
+
are not ground-truth correctness labels.
|
| 46 |
|
| 47 |
## Merged views
|
| 48 |
|
| 49 |
+
- preference_dpo_multisource_merged_v0.1: 7453 directed pairs.
|
| 50 |
+
Optional derived view for preference experiments. Chosen/rejected encode
|
| 51 |
+
author/platform interaction, not factual truth.
|
| 52 |
- reranking_listwise_multisource_merged_v0.1: 4550 groups.
|
| 53 |
Candidate metadata is retained for evaluation but excluded from the
|
| 54 |
model projection.
|
| 55 |
- corpus_rag_multisource_merged_v0.1: 39215 licensed documents.
|
| 56 |
document_text contains question context and post text only.
|
| 57 |
+
- issue_resolution_multisource_merged_v0.1: 1823 observed
|
| 58 |
+
issue/PR lifecycle records; these are not preference targets.
|
| 59 |
+
|
| 60 |
+
For ordinary NLP/RAG use, load only `corpus_rag_multisource_merged_v0.1` and
|
| 61 |
+
use `document_text` as the text input. The remaining columns are IDs,
|
| 62 |
+
provenance, timestamps or optional interaction metadata.
|
| 63 |
|
| 64 |
## Split counts
|
| 65 |
|
|
|
|
| 80 |
no public-text rows.
|
| 81 |
- Accepted, vote/reaction, timestamps, URLs and provenance fields
|
| 82 |
remain outside model-input projections.
|
| 83 |
+
- This is an experimental community-content/RAG corpus, not a
|
| 84 |
+
contamination-free factuality benchmark.
|
| 85 |
|
| 86 |
The exact source manifest hashes, split adjustments, deduplication
|
| 87 |
counts and Parquet hashes are in manifest.json.
|