Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
datasets:
|
| 4 |
+
- McAuley-Lab/Amazon-Reviews-2023
|
| 5 |
+
language:
|
| 6 |
+
- en
|
| 7 |
+
metrics:
|
| 8 |
+
- accuracy
|
| 9 |
+
base_model:
|
| 10 |
+
- microsoft/deberta-v3-base
|
| 11 |
+
pipeline_tag: text-classification
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
### Training Details
|
| 15 |
+
The model was trained on the McAuley-Lab/Amazon-Reviews-2023 dataset. This dataset contains labeled customer reviews from Amazon, focusing on two primary categories: Positive and Negative.
|
| 16 |
+
|
| 17 |
+
### Training Hyperparameters
|
| 18 |
+
* Model: microsoft/deberta-v3-base
|
| 19 |
+
* Learning Rate: 3e-5
|
| 20 |
+
* Epochs: 6
|
| 21 |
+
* Train Batch Size: 16
|
| 22 |
+
* Gradient Accumulation Steps: 2
|
| 23 |
+
* Weight Decay: 0.015
|
| 24 |
+
* Warm-up Ratio: 0.1
|
| 25 |
+
|
| 26 |
+
### Evaluation
|
| 27 |
+
The model was evaluated using a subset of the Amazon reviews dataset, focusing on the binary classification of text as either positive or negative.
|
| 28 |
+
|
| 29 |
+
### Metrics
|
| 30 |
+
Accuracy: 0.98
|
| 31 |
+
|
| 32 |
+
Precision: 0.98
|
| 33 |
+
|
| 34 |
+
Recall: 0.99
|
| 35 |
+
|
| 36 |
+
F1-Score: 0.98
|