Update app.py

app.py

@@ -5,16 +5,16 @@ from PIL import Image
 import numpy as np
 import streamlit as st
 
-# Set the device
+# Definir o dispositivo
 device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
 
-# Load the trained model from the Hugging Face Hub
+# Carregar o modelo treinado do Hugging Face Hub
 model = AutoModel.from_pretrained('dhhd255/EfficientNet_ParkinsonsPred')
 
-# Move the model to the device
+# Mover o modelo para o dispositivo
 model = model.to(device)
 
-# Add custom CSS to use the Inter font, define custom classes for healthy and parkinsons results, increase the font size, make the text bold, and define the footer styles
+# Adicionar CSS personalizado para usar a fonte Inter, definir classes personalizadas para resultados de saída e estilos de rodapé
 st.markdown("""
  <style>
  @import url('https://fonts.googleapis.com/css2?family=Inter&display=swap');
@@ -43,29 +43,29 @@ st.markdown("""
  </style>
 """, unsafe_allow_html=True)
 
-st.title("Parkinson's Disease Prediction")
+st.title("Predição de Doença de Parkinson")
 
-uploaded_file = st.file_uploader("Upload your :blue[Spiral] drawing here", type=["png", "jpg", "jpeg"])
+uploaded_file = st.file_uploader("Envie seu desenho em espiral aqui", type=["png", "jpg", "jpeg"])
 st.empty()
 if uploaded_file is not None:
     col1, col2 = st.columns(2)
 
-    # Load and resize the image
+    # Carregar e redimensionar a imagem
     image_size = (224, 224)
     new_image = Image.open(uploaded_file).convert('RGB').resize(image_size)
     col1.image(new_image, width=255)
     new_image = np.array(new_image)
     new_image = torch.from_numpy(new_image).transpose(0, 2).float().unsqueeze(0)
 
-    # Move the data to the device
+    # Mover os dados para o dispositivo
     new_image = new_image.to(device)
 
-    # Make predictions using the trained model
+    # Fazer previsões usando o modelo treinado
     with torch.no_grad():
         predictions = model(new_image)
         logits = predictions.last_hidden_state
         logits = logits.view(logits.shape[0], -1)
-        num_classes=2
+        num_classes = 2
         feature_reducer = nn.Linear(logits.shape[1], num_classes)
 
         logits = logits.to(device)
@@ -74,19 +74,21 @@ if uploaded_file is not None:
         logits = feature_reducer(logits)
         predicted_class = torch.argmax(logits, dim=1).item()
         confidence = torch.softmax(logits, dim=1)[0][predicted_class].item()
-        if(predicted_class == 0):
-            col2.markdown('<span class="result parkinsons">Predicted class: Parkinson\'s</span>', unsafe_allow_html=True)
-            col2.caption(f'{confidence*100:.0f}% sure')
+        
+        # Traduzir as mensagens de saída
+        if predicted_class == 0:
+            col2.markdown('<span class="result parkinsons">Classe prevista: Parkinson</span>', unsafe_allow_html=True)
+            col2.caption(f'{confidence*100:.0f}% de certeza')
         else:
-            col2.markdown('<span class="result healthy">Predicted class: Healthy</span>', unsafe_allow_html=True)
-            col2.caption(f'{confidence*100:.0f}% sure')
+            col2.markdown('<span class="result healthy">Classe prevista: Saudável</span>', unsafe_allow_html=True)
+            col2.caption(f'{confidence*100:.0f}% de certeza')
 
-uploaded_file = st.file_uploader("Upload your :blue[Wave] drawing here", type=["png", "jpg", "jpeg"])
+uploaded_file = st.file_uploader("Envie seu desenho de onda aqui", type=["png", "jpg", "jpeg"])
 st.divider()
 st.markdown("""
     <div class="social-links">
-        <a href="https://twitter.com/your_twitter_handle">Twitter</a>
-        <a href="https://facebook.com/your_facebook_page">Facebook</a>
-        <a href="https://instagram.com/your_instagram_handle">Instagram</a>
+        <a href="https://twitter.com/seu_twitter">Twitter</a>
+        <a href="https://facebook.com/sua_pagina_facebook">Facebook</a>
+        <a href="https://instagram.com/seu_instagram">Instagram</a>
     </div>
 """, unsafe_allow_html=True)