Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
@@ -1,11 +1,7 @@
|
|
1 |
-
---
|
2 |
-
datasets:
|
3 |
-
- leonvanbokhorst/topic-drift
|
4 |
-
---
|
5 |
|
6 |
# Topic Drift Detector Model
|
7 |
|
8 |
-
## Version:
|
9 |
|
10 |
This model detects topic drift in conversations using an enhanced attention-based architecture.
|
11 |
|
@@ -29,16 +25,16 @@ R²: 0.8373
|
|
29 |
```
|
30 |
|
31 |
## Training Curves
|
32 |
-

|
44 |
-
```
|
|
|
|
|
|
|
|
|
|
|
1 |
|
2 |
# Topic Drift Detector Model
|
3 |
|
4 |
+
## Version: v20241225_090318
|
5 |
|
6 |
This model detects topic drift in conversations using an enhanced attention-based architecture.
|
7 |
|
|
|
25 |
```
|
26 |
|
27 |
## Training Curves
|
28 |
+

|
29 |
|
30 |
## Usage
|
31 |
```python
|
32 |
import torch
|
33 |
|
34 |
# Load model
|
35 |
+
model = torch.load('models/v20241225_090318/topic_drift_model.pt')
|
36 |
|
37 |
# Use model for inference
|
38 |
# Input shape: [batch_size, sequence_length * embedding_dim]
|
39 |
# Output shape: [batch_size, 1] (drift score between 0 and 1)
|
40 |
+
```
|