Spaces:
Sleeping
Sleeping
Update README.md
Browse files
README.md
CHANGED
|
@@ -73,6 +73,21 @@ conda activate tox21_snn_cls
|
|
| 73 |
pip install -r requirements.txt
|
| 74 |
```
|
| 75 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 76 |
# Inference
|
| 77 |
|
| 78 |
For inference, you only need `predict.py`.
|
|
|
|
| 73 |
pip install -r requirements.txt
|
| 74 |
```
|
| 75 |
|
| 76 |
+
# Training
|
| 77 |
+
|
| 78 |
+
To train the self-normalizing network (SNN) from scratch, run:
|
| 79 |
+
|
| 80 |
+
```bash
|
| 81 |
+
python preprocess.py
|
| 82 |
+
python train.py
|
| 83 |
+
```
|
| 84 |
+
|
| 85 |
+
These commands will:
|
| 86 |
+
1. Load and preprocess the Tox21 training dataset
|
| 87 |
+
2. Train an SNN classifier
|
| 88 |
+
3. Store the resulting model in the `checkpoints/` directory.
|
| 89 |
+
|
| 90 |
+
|
| 91 |
# Inference
|
| 92 |
|
| 93 |
For inference, you only need `predict.py`.
|