Datasets:
Modalities:
Text
Formats:
parquet
Languages:
English
Size:
1M - 10M
Tags:
conversational
code
software-engineering
troubleshooting
technical-discussions
developer-discussions
License:
Publish one-argument RAG access documentation
Browse files
docs/releases/v0.7.0-multisource-merged.md
CHANGED
|
@@ -5,7 +5,17 @@ Release status: published experimental derivative.
|
|
| 5 |
Published to the existing Hugging Face dataset:
|
| 6 |
[`aluha501/community-qa-lab`](https://huggingface.co/datasets/aluha501/community-qa-lab)
|
| 7 |
|
| 8 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
|
| 10 |
## Published configurations
|
| 11 |
|
|
@@ -18,27 +28,28 @@ Hub commit: [`2eb94ed72d6dc0985a85437940ec40423e7f6703`](https://huggingface.co/
|
|
| 18 |
|
| 19 |
### Primary use: community content and RAG
|
| 20 |
|
| 21 |
-
`corpus_rag_multisource_merged_v0.1` is the
|
| 22 |
-
row is a licensed question/post document with its
|
| 23 |
-
metadata. It is intended for ordinary NLP, semantic
|
|
|
|
| 24 |
|
| 25 |
```python
|
| 26 |
from datasets import load_dataset
|
| 27 |
|
| 28 |
-
rag = load_dataset(
|
| 29 |
-
"aluha501/community-qa-lab",
|
| 30 |
-
"corpus_rag_multisource_merged_v0.1",
|
| 31 |
-
)
|
| 32 |
```
|
| 33 |
|
| 34 |
Use `document_text` as the text input. `accepted`, marked-answer status and
|
| 35 |
other interaction fields are not factual ground-truth annotations; they are
|
| 36 |
retained only as optional metadata for filtering and analysis.
|
| 37 |
|
| 38 |
-
The
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
|
|
|
|
|
|
|
|
|
| 42 |
|
| 43 |
The DPO and listwise views use accepted/marked-answer interaction signals as
|
| 44 |
weak labels; they do not represent factual truth. The RAG view contains only
|
|
|
|
| 5 |
Published to the existing Hugging Face dataset:
|
| 6 |
[`aluha501/community-qa-lab`](https://huggingface.co/datasets/aluha501/community-qa-lab)
|
| 7 |
|
| 8 |
+
The root `default` files are the merged RAG corpus, so ordinary users can load
|
| 9 |
+
the release with one dataset identifier and no second config argument:
|
| 10 |
+
|
| 11 |
+
```python
|
| 12 |
+
from datasets import load_dataset
|
| 13 |
+
|
| 14 |
+
rag = load_dataset("aluha501/community-qa-lab")
|
| 15 |
+
```
|
| 16 |
+
|
| 17 |
+
The publication commit is updated whenever the Hub card or root projection is
|
| 18 |
+
refreshed; the final commit is recorded in the release manifest.
|
| 19 |
|
| 20 |
## Published configurations
|
| 21 |
|
|
|
|
| 28 |
|
| 29 |
### Primary use: community content and RAG
|
| 30 |
|
| 31 |
+
`corpus_rag_multisource_merged_v0.1` is the source view used to populate the
|
| 32 |
+
root `default` files. Each row is a licensed question/post document with its
|
| 33 |
+
source and provenance metadata. It is intended for ordinary NLP, semantic
|
| 34 |
+
search and RAG indexing:
|
| 35 |
|
| 36 |
```python
|
| 37 |
from datasets import load_dataset
|
| 38 |
|
| 39 |
+
rag = load_dataset("aluha501/community-qa-lab")
|
|
|
|
|
|
|
|
|
|
| 40 |
```
|
| 41 |
|
| 42 |
Use `document_text` as the text input. `accepted`, marked-answer status and
|
| 43 |
other interaction fields are not factual ground-truth annotations; they are
|
| 44 |
retained only as optional metadata for filtering and analysis.
|
| 45 |
|
| 46 |
+
The derived view files remain namespaced under
|
| 47 |
+
`v0.7.0-multisource-merged/` in the Hub repository for reproducibility. The
|
| 48 |
+
root `default` projection now points to the merged RAG corpus. The historical
|
| 49 |
+
v0.4.0 files remain recoverable under `releases/v0.4.0/`; the named
|
| 50 |
+
`legacy_default_v0.4.0` configuration is provided for explicit compatibility.
|
| 51 |
+
The uploaded Parquet and release artifacts are read back and verified
|
| 52 |
+
byte-for-byte at the final Hub commit.
|
| 53 |
|
| 54 |
The DPO and listwise views use accepted/marked-answer interaction signals as
|
| 55 |
weak labels; they do not represent factual truth. The RAG view contains only
|