Spaces:
Running
on
CPU Upgrade
Running
on
CPU Upgrade
File size: 6,420 Bytes
970eef1 c750639 970eef1 c750639 970eef1 c750639 970eef1 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 |
from fastapi import APIRouter, HTTPException
from fastapi.responses import StreamingResponse, JSONResponse
from huggingface_hub import hf_hub_download, snapshot_download
import os
import tempfile
import shutil
import zipfile
import io
import logging
import json
from datasets import load_dataset
router = APIRouter(tags=["download"])
@router.get("/download-dataset/{session_id}")
async def download_dataset(session_id: str):
"""
Télécharge le dataset HuggingFace associé à une session et le renvoie au client
Args:
session_id: Identifiant de la session
Returns:
Fichier ZIP contenant le dataset
"""
try:
# Créer un répertoire temporaire pour stocker les fichiers du dataset
with tempfile.TemporaryDirectory() as temp_dir:
# Identifiant du repo HuggingFace
repo_id = f"yourbench/yourbench_{session_id}"
try:
# Télécharger le snapshot du dataset depuis HuggingFace
logging.info(f"Téléchargement du dataset {repo_id}")
snapshot_path = snapshot_download(
repo_id=repo_id,
repo_type="dataset",
local_dir=temp_dir,
token=os.environ.get("HF_TOKEN")
)
logging.info(f"Dataset téléchargé dans {snapshot_path}")
# Créer un fichier ZIP en mémoire
zip_io = io.BytesIO()
with zipfile.ZipFile(zip_io, 'w', zipfile.ZIP_DEFLATED) as zip_file:
# Parcourir tous les fichiers du dataset et les ajouter au ZIP
for root, _, files in os.walk(snapshot_path):
for file in files:
file_path = os.path.join(root, file)
arc_name = os.path.relpath(file_path, snapshot_path)
zip_file.write(file_path, arcname=arc_name)
# Remettre le curseur au début du stream
zip_io.seek(0)
# Renvoyer le ZIP au client
filename = f"yourbench_{session_id}_dataset.zip"
return StreamingResponse(
zip_io,
media_type="application/zip",
headers={"Content-Disposition": f"attachment; filename={filename}"}
)
except Exception as e:
logging.error(f"Erreur lors du téléchargement du dataset: {str(e)}")
raise HTTPException(
status_code=500,
detail=f"Erreur lors du téléchargement du dataset: {str(e)}"
)
except Exception as e:
logging.error(f"Erreur générale: {str(e)}")
raise HTTPException(
status_code=500,
detail=f"Erreur lors du téléchargement: {str(e)}"
)
@router.get("/download-questions/{session_id}")
async def download_questions(session_id: str):
"""
Télécharge les questions générées pour une session au format JSON
Args:
session_id: Identifiant de la session
Returns:
Fichier JSON contenant les questions générées
"""
try:
# Identifiant du repo HuggingFace
dataset_repo_id = f"yourbench/yourbench_{session_id}"
# Initialize questions list
all_questions = []
# Try to load single-shot questions
try:
single_dataset = load_dataset(dataset_repo_id, 'single_shot_questions')
if single_dataset and len(single_dataset['train']) > 0:
for idx in range(len(single_dataset['train'])):
all_questions.append({
"id": str(idx),
"question": single_dataset['train'][idx].get("question", ""),
"answer": single_dataset['train'][idx].get("self_answer", "No answer available"),
"type": "single_shot"
})
logging.info(f"Loaded {len(all_questions)} single-shot questions")
except Exception as e:
logging.error(f"Error loading single-shot questions: {str(e)}")
# Try to load multi-hop questions
try:
multi_dataset = load_dataset(dataset_repo_id, 'multi_hop_questions')
if multi_dataset and len(multi_dataset['train']) > 0:
start_idx = len(all_questions)
for idx in range(len(multi_dataset['train'])):
all_questions.append({
"id": str(start_idx + idx),
"question": multi_dataset['train'][idx].get("question", ""),
"answer": multi_dataset['train'][idx].get("self_answer", "No answer available"),
"type": "multi_hop"
})
logging.info(f"Loaded {len(multi_dataset['train'])} multi-hop questions")
except Exception as e:
logging.error(f"Error loading multi-hop questions: {str(e)}")
# If we couldn't load any questions, the dataset might not exist
if len(all_questions) == 0:
raise HTTPException(status_code=404, detail="Aucune question trouvée pour cette session")
# Convert questions to JSON
questions_json = json.dumps({
"session_id": session_id,
"questions": all_questions
}, ensure_ascii=False, indent=2)
# Create a BytesIO object with the JSON data
json_bytes = io.BytesIO(questions_json.encode('utf-8'))
json_bytes.seek(0)
# Return the JSON file for download
filename = f"yourbench_{session_id}_questions.json"
return StreamingResponse(
json_bytes,
media_type="application/json",
headers={"Content-Disposition": f"attachment; filename={filename}"}
)
except HTTPException:
# Re-raise HTTP exceptions
raise
except Exception as e:
logging.error(f"Erreur lors de la récupération des questions: {str(e)}")
raise HTTPException(
status_code=500,
detail=f"Erreur lors du téléchargement des questions: {str(e)}"
) |