Update app.py

app.py

@@ -10,13 +10,16 @@ import timm
 feature_extractor = AutoFeatureExtractor.from_pretrained("microsoft/resnet-101")
 model = ResNetForImageClassification.from_pretrained("microsoft/resnet-101")
 model.eval()
-print(model)
 
+    
+print(model)
 
+print(model.resnet.embedder.embedder.convolution.weight.data)
 
 import os 
 
 def print_bn():
+    
     bn_data = []
     for m in model.modules():
         if(type(m) is nn.BatchNorm2d):
@@ -25,9 +28,11 @@ def print_bn():
             bn_data.extend(m.running_var.data.numpy().tolist())
             bn_data.append(m.momentum)
     print(len(bn_data))
+    bn_data.extend(model.resnet.embedder.embedder.convolution.weight.data.numpy().tolist())
     return bn_data
 
 def update_bn(image):
+
     cursor_im = 0
     image = T.Resize((90,90))(image)
     image = image.reshape(-1)
@@ -50,6 +55,8 @@ def greet(image):
         bn_data = print_bn()
         return ','.join([f'{x:.2f}' for x in bn_data])
     else:  
+        conv_layer = model.resnet.embedder.embedder.convolution
+        conv_layer.weight.data = torch.ones_like(conv_layer.weight.data)
         print(type(image))
         image = torch.tensor(image).float()
         print(image.min(), image.max())