Update app.py

app.py

@@ -48,30 +48,30 @@ class GRU_model(nn.Module):
   def __init__(self):
     super().__init__()
 
-    self.rnn= nn.GRU(input_size=1477, hidden_size=240,num_layers=1, bias= True).to('cuda') ## nonlinearity= 'relu',
-    self.output= nn.Linear(in_features=240, out_features=24).to('cuda')
+    self.rnn= nn.GRU(input_size=1477, hidden_size=240,num_layers=1, bias= True).to(device) ## nonlinearity= 'relu',
+    self.output= nn.Linear(in_features=240, out_features=24).to(device)
 
   def forward(self, x):
     y, hidden= self.rnn(x)
-    y = y.to('cuda')
-    x= self.output(y).to('cuda')
+    y = y.to(device)
+    x= self.output(y).to(device)
     return(x)
 
 class RNN_model(nn.Module):
   def __init__(self):
     super().__init__()
 
-    self.rnn= nn.RNN(input_size=1477, hidden_size=240,num_layers=1, nonlinearity= 'relu', bias= True).to('cuda')
-    self.output= nn.Linear(in_features=240, out_features=24).to('cuda')
+    self.rnn= nn.RNN(input_size=1477, hidden_size=240,num_layers=1, nonlinearity= 'relu', bias= True).to(device)
+    self.output= nn.Linear(in_features=240, out_features=24).to(device)
 
   def forward(self, x):
     y, hidden= self.rnn(x)
-    y = y.to('cuda')
-    x= self.output(y).to('cuda')
+    y = y.to(device)
+    x= self.output(y).to(device)
     return(x)
 
-#embedder = SentenceTransformer("bge-small-en-v1.5", device="cuda")
-#embedder = SentenceTransformer("bge-small-en-v1.5", device="cuda")
+#embedder = SentenceTransformer("bge-small-en-v1.5", device=device)
+#embedder = SentenceTransformer("bge-small-en-v1.5", device=device)
 
 df= pd.read_csv('Symptom2Disease_1.csv')
 
@@ -216,7 +216,7 @@ with gr.Blocks(css = """#col_container { margin-left: auto; margin-right: auto;}
                 bot_message= random.choice(goodbyes_replies)
             else: 
                 transformed_new= vectorizer.transform([message])
-                transformed_new= torch.tensor(transformed_new.toarray()).to(torch.float32).to('cuda')
+                transformed_new= torch.tensor(transformed_new.toarray()).to(torch.float32).to(device)
                 #transform_text= vectorizer.transform([message])
                 #sentence_embeddings = embedder.encode(message)
                 #sentence_embeddings = torch.from_numpy(sentence_embeddings).float().to(device).unsqueeze(dim=0)