# , get_pinecone_vectorstore, find_similar_vectors from climateqa.engine.vectorstore import build_vectores_stores, get_PDF_Names_from_GCP, get_categories_files from climateqa.engine.text_retriever import ClimateQARetriever from climateqa.engine.rag import make_rag_chain from climateqa.engine.llm import get_llm from utils import create_user_id from datetime import datetime import json import re import gradio as gr from sentence_transformers import CrossEncoder reranker = CrossEncoder("mixedbread-ai/mxbai-rerank-xsmall-v1") # Load environment variables in local mode try: from dotenv import load_dotenv load_dotenv() except Exception as e: pass # Set up Gradio Theme theme = gr.themes.Soft( primary_hue="yellow", secondary_hue="orange", font=[gr.themes.GoogleFont("Poppins"), "ui-sans-serif", "system-ui", "sans-serif"], ) init_prompt = "" system_template = { "role": "system", "content": init_prompt, } user_id = create_user_id() list_categorie = get_categories_files() categories=list_categorie["AllCat"] def parse_output_llm_with_sources(output): # Split the content into a list of text and "[Doc X]" references content_parts = re.split(r'\[(Doc\s?\d+(?:,\s?Doc\s?\d+)*)\]', output) parts = [] for part in content_parts: if part.startswith("Doc"): subparts = part.split(",") subparts = [subpart.lower().replace("doc", "").strip() for subpart in subparts] subparts = [f"""({ subpart})""" for subpart in subparts] parts.append("".join(subparts)) else: parts.append(part) content_parts = "".join(parts) return content_parts def serialize_docs(docs): new_docs = [] for doc in docs: new_doc = {} new_doc["page_content"] = doc.page_content new_doc["metadata"] = doc.metadata new_docs.append(new_doc) return new_docs # Create vectorstore and retriever vectorstore = build_vectores_stores("./sources") llm = get_llm(provider="openai", max_tokens=1024, temperature=0.0) async def chat(query, history, categories, src_nb_max, src_pertinence): """taking a query and a message history, use a pipeline (reformulation, retriever, answering) to yield a tuple of: (messages in gradio format, messages in langchain format, source documents)""" print(f">> NEW QUESTION : {query} -> sources max:{src_nb_max} - pertience: {src_pertinence}") filter = None if len(categories): filter={ "$or" : [] } for cat in categories: for fich in list_categorie[cat]: filter["$or"].append({"ax_name": fich}) print( ">> Filter :" + str(filter) ) print( ">> nb sources :" + str(src_nb_max) ) print( ">> pertinence :" + str(src_pertinence) ) retriever = ClimateQARetriever( vectorstore=vectorstore, sources=["Custom"], reports=[], threshold=src_pertinence, k_total=src_nb_max, filter=filter ) rag_chain = make_rag_chain(retriever, llm) inputs = {"query": query, "audience": None} result = rag_chain.astream_log(inputs) path_reformulation = "/logs/reformulation/final_output" path_keywords = "/logs/keywords/final_output" path_retriever = "/logs/find_documents/final_output" path_answer = "/logs/answer/streamed_output_str/-" docs_html = "" output_query = "" output_language = "" output_keywords = "" gallery = [] try: async for op in result: op = op.ops[0] if op['path'] == path_reformulation: # reforulated question try: output_language = op['value']["language"] # str output_query = op["value"]["question"] except Exception as e: raise gr.Error(f"ClimateQ&A Error: {e} - The error has been noted, try another question and if the error remains, you can contact us :)") if op["path"] == path_keywords: try: output_keywords = op['value']["keywords"] # str output_keywords = " AND ".join(output_keywords) except Exception as e: pass elif op['path'] == path_retriever: # documents try: docs = op['value']['docs'] # List[Document] docs_html = [] for i, d in enumerate(docs, 1): docs_html.append(make_html_source(d, i)) docs_html = "".join(docs_html) except TypeError: print("No documents found") print("op: ", op) continue elif op['path'] == path_answer: # final answer new_token = op['value'] # str # time.sleep(0.01) previous_answer = history[-1][1] previous_answer = previous_answer if previous_answer is not None else "" answer_yet = previous_answer + new_token answer_yet = parse_output_llm_with_sources(answer_yet) history[-1] = (query, answer_yet) else: continue history = [tuple(x) for x in history] yield history, docs_html, output_query, output_language, gallery, output_query, output_keywords except Exception as e: raise gr.Error(f"{e}") timestamp = str(datetime.now().timestamp()) log_file = "logs/" + timestamp + ".json" prompt = history[-1][0] logs = { "user_id": str(user_id), "prompt": prompt, "query": prompt, "question": output_query, "sources": ["Custom"], "docs": serialize_docs(docs), "answer": history[-1][1], "time": timestamp, } #log_locally(log_file, logs) yield history, docs_html, output_query, output_language, gallery, output_query, output_keywords def make_html_source(source, i): # Prépare le contenu HTML pour un fichier texte text_content = source.page_content.strip() meta = source.metadata # Nom de la source name = f"Document {i}" # Contenu HTML de la carte card = f"""

Extrait {i} (Score:{float(meta['similarity_score'])})

{meta['ax_name']} - Page {int(meta['ax_page'])}

{text_content}

""" return card def log_locally(file, logs): # Convertit les logs en format JSON logs_json = json.dumps(logs) # Écrit les logs dans un fichier local with open(file, 'w') as f: f.write(logs_json) # -------------------------------------------------------------------- # Gradio # -------------------------------------------------------------------- init_prompt = """ Hello, I am Clara, an AI Assistant created by Axionable. My purpose is to answer your questions using the provided extracted passages, context, and guidelines. ❓ How to interact with Clara Ask your question: You can ask me anything you want to know. I'll provide an answer based on the extracted passages and other relevant sources. Response structure: I aim to provide clear and structured answers using the given data. Guidelines: I follow specific guidelines to ensure that my responses are accurate and useful. ⚠️ Limitations Though I do my best to help, there might be times when my responses are incorrect or incomplete. If that happens, please feel free to ask for more information or provide feedback to help improve my performance. What would you like to know today? """ with gr.Blocks(title="CLARA", css="style.css", theme=theme, elem_id="main-component", elem_classes="ax_background") as demo: gr.HTML(""" """, elem_classes="logo-axio ") # TAB Clara with gr.Tab("CLARA"): with gr.Row(elem_id="chatbot-row"): with gr.Column(scale=2): chatbot = gr.Chatbot( value=[(None, init_prompt)], show_copy_button=True, show_label=False, elem_id="chatbot", layout="panel", avatar_images=(None, "assets/logo4.png")) with gr.Row(elem_id="input-message"): textbox = gr.Textbox(placeholder="Posez votre question", show_label=False, scale=7, lines=1, interactive=True, elem_id="input-textbox") with gr.Column(scale=1, variant="panel", elem_id="right-panel"): # with gr.Column(scale=1, elem_id="tab-citations"): # gr.HTML("

Sources

") # slider = gr.Slider(1, 10, value=src_nb_max, step=1, label="nb max", interactive=True, elem_id="source-nb-max") # slider_p = gr.Slider(0.0, 1.0, value=src_pertinence, step=0.01, label="pertinence", interactive=True, elem_id="source-pertinence") # sources_textbox = gr.HTML( # show_label=False, elem_id="sources-textbox") # docs_textbox = gr.State("") # l'object tabs est necessaire actuellement # J'ai l'impression qu'il est utiliser pour freezre les contenu des tabs # pendant que l'ia gènère une reponse .. with gr.Tabs() as tabs: # None with gr.Tab("sources"): sources_textbox = gr.HTML( show_label=False, elem_id="sources-textbox") docs_textbox = gr.State("") with gr.Tab("filtres"): cat_sel = gr.CheckboxGroup(categories,label="Catégories") slider = gr.Slider(1, 10, value=7, step=1, label="nb max", interactive=True, elem_id="source-nb-max") slider_p = gr.Slider(0.0, 1.0, value=0.5, step=0.01, label="pertinence", interactive=True, elem_id="source-pertinence") # TAB A propos with gr.Tab("À propos", elem_classes="max-height other-tabs"): with gr.Row(): with gr.Column(scale=1): gr.Markdown( ("CLARA (Climate LLM for Adaptation & Risks Answers) by [Axionable](https://www.axionable.com/)" "– Fork de [ClimateQ&A](https://huggingface.co/spaces/Ekimetrics/climate-question-answering/tree/main)"), elem_classes="a-propos") # # TAB Configuration # with gr.Tab("Configuration"): # # with gr.Row(elem_id="config-row"): # with gr.Column(scale=1): # # for pdfName in get_PDF_Names_from_GCP(): # gr.Markdown( pdfName, elem_classes="a-propos") def start_chat(query, history): history = history + [(query, None)] history = [tuple(x) for x in history] return (gr.update(interactive=False), gr.update(selected=1), history) def finish_chat(): return (gr.update(interactive=True, value="")) (textbox .submit(start_chat, [textbox, chatbot], [textbox, tabs, chatbot], queue=False, api_name="start_chat_textbox") .then(chat, [textbox, chatbot, cat_sel, slider, slider_p], [chatbot, sources_textbox], concurrency_limit=8, api_name="chat_textbox") .then(finish_chat, None, [textbox], api_name="finish_chat_textbox") ) demo.queue() demo.launch(allowed_paths=["assets/download.png", "assets/logo4.png", "assets/axionable.svg"],favicon_path="assets/logo4.png")