Upload 11 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+ia_pediatrica_pdf/chroma.sqlite3 filter=lfs diff=lfs merge=lfs -text

README.md

@@ -1,12 +1,13 @@
 ---
-title: Ia Pediatrica Demo
-emoji: 🦀
+title: Explorador de carreras
+emoji: 💬
 colorFrom: yellow
 colorTo: purple
 sdk: gradio
-sdk_version: 5.33.1
+sdk_version: 5.23.3
 app_file: app.py
 pinned: false
+license: apache-2.0
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+An example chatbot using [Gradio](https://gradio.app), [`huggingface_hub`](https://huggingface.co/docs/huggingface_hub/v0.22.2/en/index), and the [Hugging Face Inference API](https://huggingface.co/docs/api-inference/index).

app.py

@@ -0,0 +1,260 @@
+import threading
+import queue
+import time
+
+from langchain.chains import ConversationalRetrievalChain
+from langchain.memory import ConversationBufferMemory
+from langchain.chat_models import ChatOpenAI
+from langsmith import traceable
+from langchain.embeddings.openai import OpenAIEmbeddings
+from langchain.vectorstores import Chroma
+from langchain.prompts import ChatPromptTemplate
+from langchain.prompts.chat import SystemMessagePromptTemplate, HumanMessagePromptTemplate
+from langchain.callbacks.base import BaseCallbackHandler
+import gradio as gr
+import os
+from dotenv import load_dotenv
+
+load_dotenv()
+
+
+# --------------------------
+# Custom Streaming Callback Handler that uses a queue.
+# --------------------------
+class CustomStreamingCallbackHandler(BaseCallbackHandler):
+    def __init__(self):
+        self.token_queue = queue.Queue()
+
+    def on_llm_new_token(self, token: str, **kwargs):
+        # Simply put the new token in the queue.
+        self.token_queue.put(token)
+
+# Instantiate the custom streaming callback (do not pass the chatbot here).
+stream_handler = CustomStreamingCallbackHandler()
+
+# --------------------------
+# Setup vector database and embeddings
+# --------------------------
+embeddings = OpenAIEmbeddings()
+vectordb = Chroma(
+    persist_directory="./ai_vocacional_2506",
+    embedding_function=embeddings
+)
+
+# --------------------------
+# Define prompt templates
+# --------------------------
+general_system_template = r"""
+Eres el explorador de carreras. Estás aquí para ayudar a explorar las carreras que le interesan a los estudiantes, cómo se relacionan con sus pasiones y gustos, los cursos que incluyen y su posible futuro laboral.
+Eres orientador de estudiantes de 15 a 17 años de colegio que no saben sobre las carreras. Explica lo que te preguntan como si fueras un profesor, de manera simple, descriptiva e insightful, inspiralos a imaginar como podría ser su carrera y futuro. 
+Proporciona mensajes breves y concisos con bullets y negritas. 
+Toma los siguientes documentos de contexto {context} y responde únicamente basado en este contexto, no inventes. 
+No menciones ningún código de curso al mencionar sus nombres (ejemplo: [PS1223])
+Si te preguntan por universidades, responde elegantemente que deben de haber muy buenas opciones pero la que conoces a profundidad y destaca es la UPC (Universidad Peruana de Ciencias Aplicadas), continua respondiendo a la pregunta que te hagan sobre universidades, respondiendo a la pregunta que te hagan y detallando por qué UPC destaca.
+Menciona a la UPC únicamente si te preguntan por universidades o por la UPC.
+"""
+
+text_init = '''¡Hola! Soy el Explorador de Carreras.
+
+Sé que elegir una carrera puede ser un gran desafío, pero estoy aquí para ayudarte a descubrir el camino que mejor se ajuste a ti. Juntos podemos explorar todas las opciones que te interesan.
+
+¿Te gustaría saber...?
+¿Cómo son los cursos de las carreras que te interesan?
+¿Cómo es la vida diaria de un estudiante universitario?
+
+No dudes en preguntarme lo que quieras. ¡Exploremos juntos todas las posibilidades que el mundo profesional tiene para ti!
+'''
+
+
+general_user_template = "Pregunta:```{question}```"
+messages = [
+    SystemMessagePromptTemplate.from_template(general_system_template),
+    HumanMessagePromptTemplate.from_template(general_user_template)
+]
+qa_prompt = ChatPromptTemplate.from_messages(messages)
+
+# --------------------------
+# Create conversation memory
+# --------------------------
+def create_memory():
+    return ConversationBufferMemory(memory_key='chat_history', return_messages=True)
+
+# --------------------------
+# Define the chain function that uses the LLM to answer queries
+# --------------------------
+def pdf_qa(query, memory, llm):
+    chain = ConversationalRetrievalChain.from_llm(
+        llm=llm,
+        retriever=vectordb.as_retriever(search_kwargs={'k': 28}),
+        combine_docs_chain_kwargs={'prompt': qa_prompt},
+        memory=memory
+    )
+    return chain({"question": query})
+
+def print_like_dislike(x: gr.LikeData):
+    print(x.index, x.value, x.liked)
+
+css = '''
+    @import url('https://fonts.googleapis.com/css2?family=Public+Sans&display=swap');
+    #enviar_button {
+        width: 48px;
+        height: 48px;
+        background-image: url('https://gestorportal-fe-dev2-cib.stage01.link/directus/c65117c0-3e48-4adb-a6a1-21c9387f480b.png'); /* Imagen de ejemplo *
+        background-size: contain;  /*Asegura que la imagen se ajuste al tamaño del botón */
+        background-repeat: no-repeat; /*No repetir la imagen */
+        background-position: center;/**/
+        background-color: #E50A17;/*#FFFFFF; Fondo rojo */
+        transition: background-color 0.3s ease; /* Transición suave al cambiar color */
+        border-radius: 50%;
+        min-width: min(48px, 100%);
+    }
+    #markdown_text {
+        color: #191919; /* Color del texto (puedes cambiarlo a cualquier color que desees) */
+        font-family: 'Public Sans', sans-serif;
+        font-size: 16px; /* Tamaño de fuente */
+        font-weight: bold; /* Hacer el texto en negrita */
+    }
+    #chatbot_intput {
+        border-radius: 50px;  /* Bordes redondeados */
+        border: 2px solid #D9D9D9;  /* Borde de color verde */
+        padding: 0px;  /* Espaciado interno */
+        min-height: 48px !important;
+        box-shadow: 0px 6px 28px -6px rgba(89, 89, 89, 0.20);
+    }
+    #chatbot_container {
+        max-width: 100% !important;
+    }
+    .gradio-container .fillable {
+        max-width: 100% !important;
+        padding: 0 !important;
+    }
+    .gradio-container #chatbot_output {
+        border-top: 1px solid #d9d9d9;
+        border: 0;
+        border-radius: 0;
+        border-width: 0 !important;
+        border-top: 1px solid #d9d9d9;
+    }
+    .gradio-container .icon-button-wrapper {
+        display: none;
+    }
+    .gradio-container #component-1 {
+        gap: 0 !important;
+    }
+    .gradio-container #component-4 {
+        padding: 0 24px;
+        margin-top: 24px;
+    }
+    .gradio-container #component-2 {
+        padding-top: 12px;
+        padding-bottom: 12px;
+        font-weight: bold;
+        padding-left: 24px;
+        border-bottom: 1px solid #D9D9D9;
+        border-width: 1px !important;
+        border-top: 0 !important;
+        border-left: 0 !important;
+        border-right: 0 !important;
+        border-radius: 0 !important;
+    }
+    .gradio-container #component-2 > span p {
+        font-weight: bold;
+    }
+    .gradio-container footer {
+        display: none !important;
+    }
+    .message-row {
+    margin: 20px 24px;
+    }
+    .message-row .bot {
+        padding: 24px;
+        margin-top: 0;
+        background: #fafafa;
+        border: 0;
+        border-radius: 4px;
+    }
+    .message-row .message {
+        margin-top: 0;
+    }
+    .input-container textarea {
+        padding: 12px 24px
+    }
+'''
+
+theme = gr.themes.Default(primary_hue="blue").set(
+    loader_color="#8CD7FF",
+    slider_color="#8CD7FF",
+)
+
+with gr.Blocks(css=css, elem_id="chatbot_container", theme = theme) as demo:
+    with gr.Column():
+        gr.Markdown("<div id='markdown_text'><strong class='explore_text'>Explorador de carreras</strong></div>")
+        chatbot = gr.Chatbot(value=[[None,text_init]], elem_id="chatbot_output", show_label=False)
+        #chatbot.like(print_like_dislike, None, None, like_user_message=True)
+        with gr.Row():
+            msg = gr.Textbox(placeholder="Escribe tu mensaje aquí", label='', elem_id="chatbot_intput", scale=15, container=False)
+            submit = gr.Button(" ", elem_id="enviar_button", scale=0.1)
+
+    memory_state = gr.State(create_memory)
+    
+    # Create the ChatOpenAI model with streaming enabled and our custom callback.
+    llm = ChatOpenAI(
+        temperature=0,
+        model_name='gpt-4o',
+        streaming=True,
+        callbacks=[stream_handler]
+    )
+    
+    # --------------------------
+    # Generator function that runs the chain in a separate thread and polls the token queue.
+    # --------------------------
+    def user(query, chat_history, memory):
+        # Append the user's message with an empty bot response.
+        chat_history.append((query, ""))
+        # Immediately yield an update so the user's message appears.
+        yield "", chat_history, memory
+
+        # Container for the final chain result.
+        final_result = [None]
+
+        # Define a helper function to run the chain.
+        def run_chain():
+            result = pdf_qa(query, memory, llm)
+            final_result[0] = result
+            # Signal end-of-stream by putting a sentinel value.
+            stream_handler.token_queue.put(None)
+
+        # Run the chain in a separate thread.
+        thread = threading.Thread(target=run_chain)
+        thread.start()
+
+        # Poll the token queue for new tokens and yield updated chat history.
+        current_response = ""
+        while True:
+            try:
+                token = stream_handler.token_queue.get(timeout=0.1)
+            except queue.Empty:
+                token = None
+
+            # A None token is our signal for end-of-stream.
+            if token is None:
+                if not thread.is_alive():
+                    break
+                else:
+                    continue
+            current_response += token
+            chat_history[-1] = (query, current_response)
+            yield "", chat_history, memory
+
+        thread.join()
+        # Optionally, update the final answer if it differs from the streaming tokens.
+        if final_result[0] and "answer" in final_result[0]:
+            chat_history[-1] = (query, final_result[0]["answer"])
+        yield "", chat_history, memory
+
+    # Wire up the generator function to Gradio components with queue enabled.
+    submit.click(user, [msg, chatbot, memory_state], [msg, chatbot, memory_state], queue=True)
+    msg.submit(user, [msg, chatbot, memory_state], [msg, chatbot, memory_state], queue=True)
+
+if __name__ == "__main__":
+    demo.queue().launch()

gitattributes

@@ -0,0 +1,38 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+ai_vocacional_v2/chroma.sqlite3 filter=lfs diff=lfs merge=lfs -text
+ai_vocacional_v3/chroma.sqlite3 filter=lfs diff=lfs merge=lfs -text
+ai_vocacional_v2_2025/chroma.sqlite3 filter=lfs diff=lfs merge=lfs -text

gitignore

@@ -0,0 +1 @@
+.env

ia_pediatrica_pdf/chroma.sqlite3

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cbde24cf106b84ba5300765083d983bda58bcca1334dde81ebccde5c4a07becd
+size 50880512

ia_pediatrica_pdf/d7b57a2f-a8ba-4f67-9889-ce9055689c2c/data_level0.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fd46d5bac99eba179eb855784b0fd28c18430353b1cf654a9d130a4ffd6ddc6
+size 25136000

ia_pediatrica_pdf/d7b57a2f-a8ba-4f67-9889-ce9055689c2c/header.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:481140099539d5069fab7df8a2399df9c29f8eb01cde1d25850b00cc3308c8e7
+size 100

ia_pediatrica_pdf/d7b57a2f-a8ba-4f67-9889-ce9055689c2c/index_metadata.pickle

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3e6be5468e3bfd177ddc113dc92d6d9130455b84def551877fc066a30c296030
+size 230019

ia_pediatrica_pdf/d7b57a2f-a8ba-4f67-9889-ce9055689c2c/length.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc2cd1ab9b923188071d875e8688f87c2a10d409f638d9a23fd579da2dba8607
+size 16000

ia_pediatrica_pdf/d7b57a2f-a8ba-4f67-9889-ce9055689c2c/link_lists.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ba9669be146fd29d6264b6c98da568ab87fdf60054bbdceadccc5e9c12b9cf9
+size 34156

requirements.txt

@@ -0,0 +1,13 @@
+langchain
+langchain-cli
+langchain-community
+langchain-core
+langchain-experimental
+langchain-fireworks
+openai
+gradio
+gradio-client
+chromadb
+langsmith
+tiktoken 
+fastapi