Update README.md
Browse files
README.md
CHANGED
|
@@ -22,3 +22,36 @@ configs:
|
|
| 22 |
- split: test
|
| 23 |
path: data/test-*
|
| 24 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 22 |
- split: test
|
| 23 |
path: data/test-*
|
| 24 |
---
|
| 25 |
+
|
| 26 |
+
## Amazon Pricer Dataset Lite Version
|
| 27 |
+
|
| 28 |
+
### Dataset Description
|
| 29 |
+
This dataset contains Amazon product text paired with ground-truth prices, derived from the McAuley-Lab/Amazon-Reviews-2023 dataset.
|
| 30 |
+
|
| 31 |
+
This dataset is a lite version of MightyOctopus/amazon-pricer-dataset-v2-0, having only one category(Electronics) and shrunk size (3,600 examples, including Test dataset)
|
| 32 |
+
|
| 33 |
+
The dataset is designed for training and evaluating language models on text-to-number regression tasks such as product price prediction.
|
| 34 |
+
|
| 35 |
+
### Data Fields
|
| 36 |
+
|
| 37 |
+
- text: Structured natural-language prompt containing product description and category
|
| 38 |
+
|
| 39 |
+
- price: Ground-truth product price (USD)
|
| 40 |
+
|
| 41 |
+
- Source: McAuley-Lab/Amazon-Reviews-2023
|
| 42 |
+
|
| 43 |
+
### Intended Use
|
| 44 |
+
|
| 45 |
+
- LLM-based price prediction
|
| 46 |
+
|
| 47 |
+
- Benchmarking against classical ML models
|
| 48 |
+
|
| 49 |
+
- Research and educational purposes
|
| 50 |
+
|
| 51 |
+
- Out-of-Scope Use
|
| 52 |
+
|
| 53 |
+
- Real-time pricing systems
|
| 54 |
+
|
| 55 |
+
- Financial or commercial decision-making
|
| 56 |
+
|
| 57 |
+
- License: MIT
|