Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc
|
| 3 |
+
task_categories:
|
| 4 |
+
- json
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
tags:
|
| 8 |
+
- not-for-all-audiences
|
| 9 |
+
- conversational
|
| 10 |
+
- custom-format
|
| 11 |
+
pretty_name: Danksy Dataset Alpha
|
| 12 |
+
size_categories:
|
| 13 |
+
- 100K<n<200K
|
| 14 |
+
viewer: false
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
# Danksy Dataset - Alpha Version
|
| 18 |
+
|
| 19 |
+
This dataset is a preview of a larger goal - to create a repository of quality responses that can combine both structured json and general humor
|
| 20 |
+
|
| 21 |
+
**- CAUTION: This dataset may contains words and definitions which can be considered "not safe for work" (NSFW) and/or heavily disturbing in nature. -**
|
| 22 |
+
## Dataset Summary
|
| 23 |
+
|
| 24 |
+
|
| 25 |
+
The dataset is represented with a JSONL file, with a singular JSON snippet representing one structured JSON request. Every snippet contains the following pieces of data:
|
| 26 |
+
|
| 27 |
+
- `prompt`: A request to convert a general request to a structured JSON output
|
| 28 |
+
- `expected`: The expected json output
|
| 29 |
+
|
| 30 |
+
## Files
|
| 31 |
+
|
| 32 |
+
- **words.jsonl**: A collection of words and their definition, sourced from urban dictionary
|
| 33 |
+
|
| 34 |
+
If you are using HuggingFace's `datasets` library, you can choose the file you wish to use by specifying the name of it (without extension) as an argument, like so: `dataset = load_dataset("dknoller/danksy-dataset", 'words')`.
|
| 35 |
+
The default value is `words`.
|
| 36 |
+
|
| 37 |
+
|
| 38 |
+
___
|
| 39 |
+
Any relationship between the name of this dataset and any public personas is entirely and totally coincidential.
|