Essogbe

import dspy
from typing import List, Dict, Optional
import json
import os

# Configuration du modèle Gemini Flash
lm = dspy.LM("gemini/gemini-2.0-flash", api_key="YOUR_GEMINI_API_KEY")
dspy.configure(lm=lm)

# Signature pour la validation d'affiches avec scoring

Essogbe

import os
import requests

api_key = os.environ.get("MISTRAL_API_KEY")  # Remplacez par votre clé réelle

def transcription_from_audio(audio_file_path: str) -> str:
    """
    Envoie un fichier audio à l'API Mistral pour transcription.
    Retourne le texte transcrit.
    """

Essogbe

"""
Audio Preprocessing Mini-Pipeline 

This module implements a comprehensive audio preprocessing pipeline designed to optimize audio samples. The preprocessing steps ensure consistent
audio quality and format while removing noise and silent segments that could interfere 

Libraries Used:
- librosa: Advanced audio analysis and feature extraction library
- numpy: Numerical computing for array operations and mathematical transformations
- noisereduce: Spectral gating noise reduction using signal processing techniques

Essogbe

import dspy
import json

# Configuration du LLM

lm = dspy.LM('openai/mistral-small-latest', api_key='CLE-API', api_base='https://api.mistral.ai/v1')
dspy.configure(lm=lm)

# === MODULE 1: Générateur de requêtes utilisateur ===
class UserRequestSignature(dspy.Signature):

Essogbe

Essogbe

#libraries
import math
import pandas as pd
import numpy as np
import random

def distance(a,b):
    return math.sqrt(((a[0]-b[0])**2)+(a[1]-b[1])**2)

def dist_from_centroid(points,c):

Essogbe

def get_something(elements,chaine):
    elements_temp=[]
    for  i in elements:
        elements_temp.extend(list(i))
    elements=elements_temp
    elements =[re.escape(elem) for elem in elements]
    resultat = "|".join([f"{elem}" for elem in elements])
    text=chaine
    matches = re.findall(r'!\[([^\]]+)\]\(([^)]+)\)', text)
    to_find=[]