--- a
+++ b/model/resnet.py
@@ -0,0 +1,22 @@
+import torch.nn as nn
+
+from torchvision import models
+
+
+class ResnetModel(nn.Module):
+
+    def __init__(self, classes):
+        """
+        Arguments
+        ---------
+        classes: número de clases (tipos de hemorragias)
+        """
+        super(ResnetModel, self).__init__()
+        self.backbone = models.resnet50(pretrained=False)
+        n_filters = self.backbone.fc.in_features
+        self.backbone.fc = nn.Linear(n_filters, classes)
+
+    def forward(self, x):
+        x = self.backbone(x)
+
+        return x