ajout de quelques agents / fonctions

agent.py

@@ -24,12 +24,17 @@ llm_4o = ChatOpenAI(
     model_name="gpt-4o",
     openai_api_key=api_open_ai_agent_key,  # ou variable d’environnement
 )
+llm_4_1 = ChatOpenAI(
+    model_name='gpt-4.1',
+    openai_api_key=api_open_ai_agent_key,  # ou variable d’environnement
+)
 llm_reasoning = ChatOpenAI(
     model_name = "o3-2025-04-16",
     openai_api_key=api_open_ai_agent_key,  # ou variable d’environnement
 )
 
 
+
 llm_reasoning_small = ChatOpenAI(
     model_name = "o4-mini-2025-04-16",
     openai_api_key=api_open_ai_agent_key,  # ou variable d’environnement
@@ -133,18 +138,17 @@ def divide(a: float, b: float):
     return a / b
 
 
-
 @tool
-def create_agent_and_answer(input: ToolInput) -> str:
+def create_agent_and_answer(dict_data, question) -> str:
     """ From a dataframe, can anwser any question
-    The input should be like : input_data = ToolInput(
-    question="Quel est l'âge moyen ?",
-    data= 'ouput of the open_xlsx_doc tool'
+    The input should be like :
+    - dict_data= a dict
+    - question= a str. Exemple : "Quel est l'âge moyen ?",
+
     """
-    df = pd.DataFrame(input.data)
-    question = input.question
-    agent_excel_inter = create_pandas_dataframe_agent(llm_4o, df, verbose=True, allow_dangerous_code=True)
-    text = agent_excel_inter.run(question)
+    df = pd.DataFrame(dict_data)
+    agent_excel = create_pandas_dataframe_agent(llm_4_1, df, verbose=True, allow_dangerous_code=True)
+    text = agent_excel.run(question)
 
     return text
 
@@ -199,24 +203,9 @@ math_agent = create_react_agent(
 
     name="math_agent")
 
-agent_excel = create_pandas_dataframe_agent(
-    model=llm_4o,
-
-    tools=[create_agent_and_answer],
-    prompt=(
-        "You are an agent specialized with Excel files.\n\n"
-        "INSTRUCTIONS:\n"
-        "- Assist ONLY when an Excel file is mentionned \n"
-        "- You will receive the path of a dataframe. '\n"
-        "- The output of the first tool  is a part of the input of the second tool\n"
-
-        "- Once you have got a response from the 'create_agent_and_answer_tool', transmit it to your supervisor \n"
-    ),
-    name="agent_excel",
-)
 
 reflexion_agent = create_react_agent(
-    model=llm_reasoning,
+    model=llm_reasoning_small,
 
     tools=[],
     prompt=(
@@ -229,22 +218,37 @@ reflexion_agent = create_react_agent(
     ),
     name="reflexion_agent",
 )
+agent_excel_new = create_react_agent(
+    model=llm_4o,
 
+    tools=[create_agent_and_answer],
+    prompt=(
+        "You are an agent psecialized with Excel files.\n\n"
+        "INSTRUCTIONS:\n"
+        "- Assist ONLY when an Excel file is mentionned \n"
+        "- First, you will receive a dict that you can give to the associated file. If you don't have one, ask it to the supervisor.\n"
+        "- Then you use the 'create_agent_and_answer' tool to answer the question.   \n"
+        "- Once you have got a response from the 'create_agent_and_answer_tool', transmit it to your supervisor \n"
+    ),
+    name="agent_excel_new",
+)
 supervisor = create_supervisor(
     model=init_chat_model("openai:gpt-4.1", api_key = api_open_ai_agent_key),
-    agents=[research_agent, web_search_openai_agent, agent_excel, reflexion_agent],
+    agents=[research_agent, web_search_openai_agent, agent_excel_new, reflexion_agent],
     prompt=(
-        "You are a supervisor managing four agents:\n"
+        "You are a supervisor managing three agents:\n"
         "- research_agent: Specialised in ArXiv and Wikipedia. Assign research-related tasks to this agent.\n"
+        "- reflexion: Called when the supervisor need a reflexion, not general knowledge. Can handle  math-related tasks such as solving equations, performing calculations or working on abstract maths subject such as matrix or demonstrating subjects.\n"
         "- web_search_openai_agent: Can browse the web to find up-to-date and relevant information. Assign web-related tasks to this agent.\n"
-        "- agent_excel: Can exploit Excel file. You have to give him an Excel path file with a question, and wait for his response.\n"
-        "- reflexion_agent: This agent uses a powerful model. It is your most intelligent agent. Useful for reflexion tasks. It can solve math problems also.\n"
+        "- agent_excel_new: Can understand tabular data. If an excel file is mentioned, call this agent. \n"
         "Assign work to one agent at a time. Do not call agents in parallel.\n"
-        "When a new question arises, always first consult the research_agent — it may provide useful information.\n"
+        "The reflexion agent is your best weapon when the is a complex question. Call him only one time maximum by question.\n"
+        " If there is an attached file, it will already loaded. Juste give the information to the agent. \n"
+        "When a new question arises, if it is about an information that you can find on Wikipedia,  first consult the research_agent — it may provide useful information.\n"
         "If research_agent yields no results, then delegate the task to web_search_openai_agent.\n"
-        "Each time you receive information from an agent, you have to analyze, process it then decide what to do (call an agent or give your final answer)."
-        "When you consider it pertinent, give your thought to the reflexion agent. He will advise you. "
-        "Report your thoughts, and finish your answer with the following template: FINAL ANSWER: [YOUR FINAL ANSWER]. YOUR FINAL ANSWER should be a number OR as few words as possible OR a comma separated list of numbers and/or strings. If you are asked for a number, don't use comma to write your number neither use units such as $ or percent sign unless specified otherwise. If you are asked for a string, don't use articles, neither abbreviations (e.g. for cities), and write the digits in plain text unless specified otherwise. If you are asked for a comma separated list, apply the above rules depending of whether the element to be put in the list is a number or a string. If no punctuation is precised, don't add any. Respect the requested format"
+        "Each time you receive information from an agent, you have to analyze, process it then decide what to do (call an agent or give your final answer).\n"
+        "As soon as you get a question, you have to analyze it and determine which agent is the most competent. Call at least one for each question.\n"
+        "IMPORTANT : Report your thoughts, and finish your answer with the following template: FINAL ANSWER: [YOUR FINAL ANSWER]. YOUR FINAL ANSWER should be a number OR as few words as possible OR a comma separated list of numbers and/or strings. If you are asked for a number, don't use comma to write your number NEITHER use units such as $, percent sign, or the currency unless specified otherwise. If you are asked for a string, don't use articles, neither abbreviations (e.g. for cities), and write the digits in plain text unless specified otherwise. If you are asked for a comma separated list, apply the above rules depending of whether the element to be put in the list is a number or a string. If no punctuation is precised, don't add any. If you are asked for a price, don't precise the format, only the number. Respect the requested format"
 
     ),
     add_handoff_back_messages=True,

app.py

@@ -3,7 +3,7 @@ import gradio as gr
 import requests
 import pandas as pd
 from agent import response_from_agent, supervisor
-
+from fonctions import load_data
 # (Keep Constants as is)
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
@@ -76,9 +76,12 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
             continue
         try:
             print(f'la question est {question_text}')
-            if filename :
-                print('fichier attaché : ', filename)
-            submitted_answer = response_from_agent(supervisor,question_text + 'The file_name (path) is : ' + filename)
+
+            data = load_data(item)
+            print("data:", data)
+            #submitted_answer = response_from_agent(supervisor,question_text + 'The file_name (path) is : ' + filename)
+            submitted_answer = response_from_agent(f'the question is {question_text}. The attached file is {data}. If it is a .mp3, you have the transcripted text in the attached file. If it is an excel file, you have a dictionnary. If it is a png file, you will have the path to the file.')
+
             print('submitted_answer:', submitted_answer)
             answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer})
             results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": submitted_answer})

fonctions.py

@@ -1,9 +1,11 @@
 from io import BytesIO
-from pydub import AudioSegment
 import requests
 import re
 import pandas as pd
+from openai import OpenAI
 
+api_open_ai_agent_key=os.environ["OPENAI_API_KEY"]
+client = OpenAI(api_key=api_open_ai_agent_key)
 def clean_response(response):
     match = re.search(r'FINAL ANSWER:\s*(.+)', response['supervisor']['messages'][-1].content)
     answer = match.group(1).strip() if match else None
@@ -21,9 +23,11 @@ def response_from_agent(supervisor, question):
 
     response = clean_response(response)
     return response
+
+
 def load_data(question):
-    task_id = question['task_id']
-    file_name = question['file_name']
+    task_id = question.get('task_id')
+    file_name = question.get('file_name')
 
     if file_name == "":
         return 'There is no attached file'
@@ -33,17 +37,34 @@ def load_data(question):
         return 'Le lien ne fonctionne pas'
     if file_name.endswith('.xlsx'):
         excel_data = BytesIO(files_response.content)
-        df = pd.read_excel(excel_data, engine='openpyxl')
-        return df
+        df = pd.read_excel(excel_data)
+        data_dict = df.to_dict(orient="list")
+        return data_dict
 
     elif file_name.endswith('.png'):
-        filename = f"fichier_{task_id}.png"
-        with open(filename, "wb") as f:
-            f.write(files_response.content)
-        return filename
+        response = client.responses.create(
+            model="gpt-4.1-mini",
+            input=[{
+                "role": "user",
+                "content": [
+                    {"type": "input_text", "text": "what's in this image? Please give as much details as possible"},
+                    {
+                        "type": "input_image",
+                        "image_url": f"https://agents-course-unit4-scoring.hf.space/files/{task_id}",
+                    },
+                ],
+            }],
+        )
+
+        return response.output_text
 
     elif file_name.endswith('.mp3'):
         audio_bytes = BytesIO(files_response.content)
-        audio = AudioSegment.from_file(audio_bytes, format="mp3")
+        audio_bytes.name = "audio.mp3"
+
+        transcription = client.audio.transcriptions.create(
+            model="whisper-1",  # ou "whisper-1", mais "gpt-4o" est aussi correct
+            file=audio_bytes
+        )
 
-        return audio
+        return transcription.text