metadata
license: mit
task_categories:
- text-generation
- question-answering
language:
- fr
- en
tags:
- marketplace
- ecommerce
- products
- digital-marketplace
- french
- jsonl
pretty_name: MarketplaceAI JSONL Dataset
size_categories:
- 10K<n<100K
MarketplaceJSONLDataset - Dataset JSONL
Dataset au format JSON Lines contenant des données de produits pour l'entraînement de modèles d'IA spécialisés dans les marketplaces.
📊 Statistiques du Dataset
- Format : JSON Lines (.jsonl)
- Nombre total d'exemples : 500,000,000
- Taille totale : 566356.09 MB
- Nombre de fichiers : 1
📁 Structure des Fichiers
produits_numeriquesv2.jsonl
- Lignes : 500,000,000
- Taille : 566356.09 MB
- Champs : seo_title_en, seo_description_fr, seo_title_fr, description_fr, updated_at, product_type, comments_fr, created_at, variants, tags_fr, description_en, tags_en, category_id, extend_price, seo_description_en, comments_en, product_id, name_fr, name_en, regular_price, discount_offer
Exemple :
{
"product_id": "998dec5c-f9fb-494a-ad49-996b6bf0c1fd",
"name_en": "La possibilité d'innover à la pointe",
"name_fr": "Le pouvoir d'évoluer sans soucis",
"description_en": "<p>Foi bout présenter mémoire chemin troubler. High-quality digital solution.</p>",
"description_fr": "<p>Lieu nombre deviner voile. Solution numérique de haute qualité.</p>",
"tags_en": "api, digital, software",
"tags_fr": "design, saas, api",
"seo_title_en": "La possibilité d'innover à la pointe - Best saas Resource",
"seo_title_fr": "Le pouvoir d'évoluer sans soucis - Meilleure ressource framework",
"seo_description_en": "Discover La possibilité d'innover à la pointe, the ultimate educational_content solution.",
"seo_description_fr": "Découvrez Le pouvoir d'évoluer sans soucis, la solution ultime visualisation de données interactive.",
"regular_price": 101.31,
"extend_price": 133.87,
"category_id": 6,
"product_type": "tutorial",
"comments_en": "",
"comments_fr": "Danger choix quand. | Ouvrir pourquoi car. | Classe barbe loi découvrir.",
"discount_offer": "",
"variants": "Standard ($101.31) | Extended ($133.87)",
"created_at": "2024-12-24 08:59:32",
"updated_at": "2025-04-23 09:34:30"
}
🔧 Format des Données
Chaque ligne du fichier JSONL contient un objet JSON valide :
{"id": 1, "name": "Produit A", "description": "Description du produit...", "price": 99.99}
{"id": 2, "name": "Produit B", "description": "Autre description...", "price": 149.99}
💻 Utilisation
Avec Python
import json
# Lire un fichier JSONL
def read_jsonl(file_path):
with open(file_path, 'r', encoding='utf-8') as f:
for line in f:
yield json.loads(line.strip())
# Utilisation
for item in read_jsonl('produits_numeriquesv2.jsonl'):
print(item)
Avec Pandas
import pandas as pd
import json
# Charger un JSONL en DataFrame
def jsonl_to_dataframe(file_path):
data = []
with open(file_path, 'r', encoding='utf-8') as f:
for line in f:
data.append(json.loads(line.strip()))
return pd.DataFrame(data)
df = jsonl_to_dataframe('produits_numeriquesv2.jsonl')
print(df.head())
Avec MarketplaceAI
from marketplace_ai_server import MultilingualMarketplaceBot
# Convertir JSONL en JSON pour l'entraînement
def convert_jsonl_to_json(jsonl_path, json_path):
products = []
with open(jsonl_path, 'r', encoding='utf-8') as f:
for line in f:
products.append(json.loads(line.strip()))
with open(json_path, 'w', encoding='utf-8') as f:
json.dump({"products": products}, f, indent=2, ensure_ascii=False)
# Convertir et entraîner
convert_jsonl_to_json('produits_numeriquesv2.jsonl', 'products.json')
bot = MultilingualMarketplaceBot()
bot.train_on_json_file('products.json')
🚀 Intégration avec Hugging Face Datasets
from datasets import load_dataset
# Charger le dataset
dataset = load_dataset('Medyassino/MarketplaceJSONLDataset')
# Accéder aux données
for example in dataset['train']:
print(example)
📚 Citation
Si vous utilisez ce dataset, veuillez citer :
@dataset{marketplace_jsonl_dataset,
title={MarketplaceAI JSONL Dataset},
author={Medyassino},
year={2025},
format={JSONL},
url={https://huggingface.co/datasets/Medyassino/MarketplaceJSONLDataset}
}