Upload asdklasjkladkladskladskjladsklj/AgeClassificationModel with huggingface_hub
Browse files
asdklasjkladkladskladskjladsklj/AgeClassificationModel
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from .base_feature_classification_model import BaseFeatureClassificationModel
|
| 2 |
+
import torch.nn as nn
|
| 3 |
+
import torch.nn.functional as F
|
| 4 |
+
|
| 5 |
+
class ComplexNN(nn.Module):
|
| 6 |
+
def __init__(self,n_classes,input_dim):
|
| 7 |
+
super(ComplexNN, self).__init__()
|
| 8 |
+
self.fc1 = nn.Linear(input_dim, 8)
|
| 9 |
+
self.fc2 = nn.Linear(8, n_classes)
|
| 10 |
+
|
| 11 |
+
def forward(self, x):
|
| 12 |
+
x = F.relu(self.fc1(x))
|
| 13 |
+
x = self.fc2(x)
|
| 14 |
+
return x
|
| 15 |
+
|
| 16 |
+
class AgeClassificationModel(BaseFeatureClassificationModel):
|
| 17 |
+
def __init__(self, ckpt, batch_size=64, val_batch_size=128,
|
| 18 |
+
learning_rate=1e-4, epochs=500, es_paiteince=5, num_workers=8, train_ratio=0.85,device='gpu',gpu_ids=[0]):
|
| 19 |
+
super().__init__(ckpt, batch_size, val_batch_size,
|
| 20 |
+
learning_rate, epochs, es_paiteince, num_workers, train_ratio,device,gpu_ids)
|
| 21 |
+
|
| 22 |
+
def input_dataset(self, dataset):
|
| 23 |
+
model = ComplexNN(n_classes=dataset.n_classes, input_dim=dataset.num_features)
|
| 24 |
+
super().input_dataset(dataset,model)
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
|