Update app.py

app.py

@@ -24,11 +24,11 @@ def load_documents_from_jsonl(embeddings_model, jsonl_path, createEmbeddings=Fal
     return df
        
 def generate_embeddings(tokenizer, model, text):
-#    with torch.no_grad():
-#        embeddings = model.encode(text, convert_to_tensor=True)
-    encoded_input = tokenizer(text, padding=True, truncation=True, return_tensors='pt')
     with torch.no_grad():
-        embeddings = model(**encoded_input)
+        embeddings = model.encode(text, convert_to_tensor=True)
+#    encoded_input = tokenizer(text, padding=True, truncation=True, return_tensors='pt')
+#    with torch.no_grad():
+#        embeddings = model(**encoded_input)
     return embeddings.cpu().numpy()
 
 def save_to_faiss(df):
@@ -55,11 +55,11 @@ def main():
     # Заголовок приложения
     st.title("Demo for LLAMA-2 RAG with CPU only")
 
-    #model = SentenceTransformer('sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2')
-    tokenizer = AutoTokenizer.from_pretrained('sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2')
-    model = AutoModel.from_pretrained('sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2')
+    model = SentenceTransformer('sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2')
+    #tokenizer = AutoTokenizer.from_pretrained('sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2')
+    #model = AutoModel.from_pretrained('sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2')
 
-    df_qa = load_documents_from_jsonl('ExportForAI1.jsonl', model, False)     
+    df_qa = load_documents_from_jsonl(model, 'ExportForAI1.jsonl', False)     
     save_to_faiss(df_qa)
     
     # Текстовое поле для ввода вопроса