Create app.py

app.py

@@ -0,0 +1,88 @@
+import streamlit as st
+from transformers import AutoTokenizer
+from llm_engine import HuggingFaceEndpoint, ChatHuggingFace
+from huggingface_hub import login
+from PyPDF2 import PdfReader
+from docx import Document
+import csv
+import json
+import os
+
+huggingface_token = os.getenv('HUGGINGFACE_TOKEN')
+
+# Realizar el inicio de sesión de Hugging Face solo si el token está disponible
+if huggingface_token:
+    login(token=huggingface_token)
+
+# Configuración del modelo
+@st.cache_resource
+def load_llm():
+    llm = HuggingFaceEndpoint(
+        repo_id="mistralai/Mistral-7B-Instruct-v0.3",
+        task="text-generation"
+    )
+    llm_engine_hf = ChatHuggingFace(llm=llm)
+    tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.3")
+    return llm_engine_hf, tokenizer
+
+llm_engine_hf, tokenizer = load_llm()
+
+st.title("LexAIcon")
+st.write("Puedes conversar con este chatbot basado en Mistral7B-Instruct y subir archivos para que el chatbot los procese.")
+
+if "generated" not in st.session_state:
+    st.session_state["generated"] = []
+if "past" not in st.session_state:
+    st.session_state["past"] = []
+
+def generate_response(prompt):
+    response = llm_engine_hf.invoke(prompt)
+    return response
+
+def handle_uploaded_file(uploaded_file):
+    try:
+        if uploaded_file.name.endswith(".txt"):
+            text = uploaded_file.read().decode("utf-8")
+        elif uploaded_file.name.endswith(".pdf"):
+            reader = PdfReader(uploaded_file)
+            text = ""
+            for page in range(len(reader.pages)):
+                text += reader.pages[page].extract_text()
+        elif uploaded_file.name.endswith(".docx"):
+            doc = Document(uploaded_file)
+            text = "\n".join([para.text for para in doc.paragraphs])
+        elif uploaded_file.name.endswith(".csv"):
+            text = ""
+            content = uploaded_file.read().decode("utf-8").splitlines()
+            reader = csv.reader(content)
+            text = " ".join([" ".join(row) for row in reader])
+        elif uploaded_file.name.endswith(".json"):
+            data = json.load(uploaded_file)
+            text = json.dumps(data, indent=4)
+        else:
+            text = "Tipo de archivo no soportado."
+        return text
+    except Exception as e:
+        return str(e)
+
+# Entrada del usuario
+user_input = st.text_input("Tú: ", "")
+
+# Manejo de archivos subidos
+uploaded_files = st.file_uploader("Sube un archivo", type=["txt", "pdf", "docx", "csv", "json"], accept_multiple_files=True)
+
+if st.button("Enviar"):
+    if user_input:
+        response = generate_response(user_input)
+        st.session_state.generated.append({"user": user_input, "bot": response})
+
+if st.session_state["generated"]:
+    for chat in st.session_state["generated"]:
+        st.write(f"Tú: {chat['user']}")
+        st.write(f"Chatbot: {chat['bot']}")
+
+if uploaded_files:
+    for uploaded_file in uploaded_files:
+        st.write(f"Archivo subido: {uploaded_file.name}")
+        file_content = handle_uploaded_file(uploaded_file)
+        st.write(file_content)