Update app.py

app.py

@@ -1,22 +1,27 @@
 import streamlit as st
-import torch
 from transformers import AutoTokenizer, AutoModel
+import torch
+
+# carregar o modelo e o tokenizer
+model = AutoModel.from_pretrained("neuralmind/bert-large-portuguese-cased")
+tokenizer = AutoTokenizer.from_pretrained("neuralmind/bert-large-portuguese-cased")
 
-# load the pre-trained model
-model = SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')
+st.title("Comparador de Similaridade de Textos em Português")
+st.write("Este aplicativo usa o modelo BERT-large treinado para o português para calcular a similaridade entre dois textos.")
 
-# set the app title
-st.title("Brazilian Portuguese Sentence Similarity Checker")
+# definir as caixas de entrada para as duas frases
+frase1 = st.text_input("Insira a primeira frase:")
+frase2 = st.text_input("Insira a segunda frase:")
 
-# get the input sentences from the user
-sentence1 = st.text_input("Enter the first sentence:")
-sentence2 = st.text_input("Enter the second sentence:")
+# verificar se as frases foram inseridas e calcular a similaridade se sim
+if frase1 and frase2:
+    # tokenizar as frases e obter os embeddings
+    input_ids = tokenizer.encode_plus(frase1, frase2, padding='max_length', truncation=True, return_tensors='pt')
+    with torch.no_grad():
+        embeddings = model(input_ids['input_ids'], attention_mask=input_ids['attention_mask'])[0]
 
-# check if both sentences are not empty
-if sentence1 and sentence2:
-    embedding_1= model.encode(sentence1, convert_to_tensor=True)
-    embedding_2 = model.encode(sentence2, convert_to_tensor=True)
+    # calcular a similaridade entre os embeddings
+    cos_sim = torch.nn.functional.cosine_similarity(embeddings[0], embeddings[1], dim=0)
 
-    similarity = util.pytorch_cos_sim(embedding_1, embedding_2)
-    
-    st.write("Similarity score between the sentences:", similarity)
+    # imprimir o resultado
+    st.write("A similaridade entre as frases é:", cos_sim.item())