Update app.py

app.py

@@ -1,64 +1,63 @@
 import gradio as gr
+from langchain_community.llms import HuggingFaceHub
+from langchain_core.output_parsers import StrOutputParser
+from langchain import hub
+from langchain_community.document_loaders import PyPDFLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain_community.vectorstores import Chroma
+from langchain_huggingface import HuggingFaceEmbeddings
 from huggingface_hub import InferenceClient
+from rerankers import Reranker
+import os
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+loader = PyPDFLoader("Constitucion_española.pdf")
+documents = loader.load()
 
+text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=100)
+docs_split = text_splitter.split_documents(documents)
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
+embedding_function = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
 
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
+vectordb = Chroma.from_documents(docs_split, embedding_function)
 
-    messages.append({"role": "user", "content": message})
+client = InferenceClient("google/flan-t5-base", token=os.getenv("HUGGINGFACEHUB_API_TOKEN"))
 
-    response = ""
+ranker = Reranker("answerdotai/answerai-colbert-small-v1", model_type='colbert')
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+def generate_text(context, query):
+    inputs = f"Context: {context} Question: {query}"
+    response = client.chat_completion(inputs=inputs, task="text2text-generation")
+    return response
 
-        response += token
-        yield response
+def test_rag_reranking(query, ranker):
+    docs = vectordb.similarity_search_with_score(query)
+    context = []
+    for doc, score in docs:
+        if score < 7:  
+            doc_details = doc.to_json()['kwargs']
+            context.append(doc_details['page_content'])
 
+    if len(context) > 0:
+        useful_context = context[0]  
+        generation = generate_text(useful_context, query)
+        return generation
+    else:
+        return "No se encontró información suficiente para responder."
+
+    ranked = ranker.rerank(query=query, documents=raw_contexts, top_k=1)
+
+    best_context = ranked[0]["content"]
+    return generate_text(best_context, query)
+
+def responder_chat(message, history):
+    respuesta = test_rag_reranking(message, ranker)
+    return respuesta
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
 demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
+    fn=responder_chat,
+    title="Chatbot sobre la constitución española",
+    theme="soft"
 )
 
-
 if __name__ == "__main__":
     demo.launch()