Spaces:
Sleeping
Sleeping
Upload 2 files
Browse files- app.py +17 -16
- game_engine.py +73 -230
app.py
CHANGED
|
@@ -1,17 +1,15 @@
|
|
| 1 |
# ==========================================
|
| 2 |
-
# app.py - Calcul OCR v3.0 - Version
|
| 3 |
# ==========================================
|
| 4 |
|
| 5 |
"""
|
| 6 |
-
Application principale
|
|
|
|
| 7 |
"""
|
| 8 |
|
| 9 |
import gradio as gr
|
| 10 |
import warnings
|
| 11 |
import os
|
| 12 |
-
import gc
|
| 13 |
-
import numpy as np
|
| 14 |
-
from PIL import Image
|
| 15 |
|
| 16 |
warnings.filterwarnings("ignore")
|
| 17 |
|
|
@@ -19,7 +17,7 @@ warnings.filterwarnings("ignore")
|
|
| 19 |
from image_processing_gpu import init_ocr_model, create_white_canvas, cleanup_memory
|
| 20 |
from game_engine import MathGame, export_to_clean_dataset
|
| 21 |
|
| 22 |
-
print("🚀 Initialisation Calcul OCR v3.0 (
|
| 23 |
print("🔄 Chargement modèle TrOCR...")
|
| 24 |
init_ocr_model()
|
| 25 |
print("✅ Modèle TrOCR prêt")
|
|
@@ -29,24 +27,24 @@ def create_new_game_session():
|
|
| 29 |
return MathGame()
|
| 30 |
|
| 31 |
def start_game_wrapper(duration: str, operation: str, difficulty: str, game_state) -> tuple:
|
| 32 |
-
"""
|
| 33 |
if game_state is None:
|
| 34 |
game_state = create_new_game_session()
|
| 35 |
|
| 36 |
cleanup_memory()
|
| 37 |
result = game_state.start_game(duration, operation, difficulty)
|
| 38 |
-
return result + (game_state, "")
|
| 39 |
|
| 40 |
def next_question_wrapper(image_data, game_state) -> tuple:
|
| 41 |
-
"""
|
| 42 |
if game_state is None:
|
| 43 |
game_state = create_new_game_session()
|
| 44 |
|
| 45 |
result = game_state.next_question(image_data)
|
| 46 |
-
return result + (game_state,)
|
| 47 |
|
| 48 |
def export_current_session(game_state) -> str:
|
| 49 |
-
"""Export
|
| 50 |
if game_state is None:
|
| 51 |
return "❌ Aucune session active"
|
| 52 |
|
|
@@ -81,7 +79,7 @@ def export_current_session(game_state) -> str:
|
|
| 81 |
game_state.export_status = "not_exported"
|
| 82 |
return f"❌ Erreur export: {str(e)}"
|
| 83 |
|
| 84 |
-
# Interface Gradio
|
| 85 |
with gr.Blocks(
|
| 86 |
title="🧮 Calcul OCR - Entraînement mathématiques",
|
| 87 |
theme=gr.themes.Soft(),
|
|
@@ -111,7 +109,7 @@ with gr.Blocks(
|
|
| 111 |
head="<meta name='viewport' content='width=device-width, initial-scale=1.0'>"
|
| 112 |
) as demo:
|
| 113 |
|
| 114 |
-
# State management -
|
| 115 |
game_state = gr.State(value=None)
|
| 116 |
|
| 117 |
gr.Markdown(
|
|
@@ -125,6 +123,8 @@ with gr.Blocks(
|
|
| 125 |
4. Cliquez sur **➡️ NEXT !** pour la question suivante
|
| 126 |
|
| 127 |
À la fin, vous pourrez consulter vos résultats et contribuer au dataset ouvert !
|
|
|
|
|
|
|
| 128 |
"""
|
| 129 |
)
|
| 130 |
|
|
@@ -180,11 +180,11 @@ with gr.Blocks(
|
|
| 180 |
# Résultats
|
| 181 |
results_display = gr.HTML("")
|
| 182 |
|
| 183 |
-
# Export vers dataset
|
| 184 |
export_button = gr.Button("📤 Exporter la série vers le dataset", variant="primary", size="lg")
|
| 185 |
export_status = gr.Markdown("")
|
| 186 |
|
| 187 |
-
# Événements
|
| 188 |
go_button.click(
|
| 189 |
fn=start_game_wrapper,
|
| 190 |
inputs=[duration_choice, operation_choice, difficulty_choice, game_state],
|
|
@@ -204,8 +204,9 @@ with gr.Blocks(
|
|
| 204 |
)
|
| 205 |
|
| 206 |
if __name__ == "__main__":
|
| 207 |
-
print("🚀 Lancement Calcul OCR v3.0 -
|
| 208 |
print("🎯 ZeroGPU HuggingFace Spaces")
|
|
|
|
| 209 |
print("🎯 Dataset: calcul_ocr_dataset")
|
| 210 |
print("📊 Opérations: ×, +, -, ÷, Aléatoire")
|
| 211 |
print("⚙️ Durées: 30s, 60s")
|
|
|
|
| 1 |
# ==========================================
|
| 2 |
+
# app.py - Calcul OCR v3.0 - Version ultra-simplifiée
|
| 3 |
# ==========================================
|
| 4 |
|
| 5 |
"""
|
| 6 |
+
Application principale ultra-simplifiée pour ZeroGPU HuggingFace Spaces
|
| 7 |
+
OCR en fin de session uniquement - Performance optimale
|
| 8 |
"""
|
| 9 |
|
| 10 |
import gradio as gr
|
| 11 |
import warnings
|
| 12 |
import os
|
|
|
|
|
|
|
|
|
|
| 13 |
|
| 14 |
warnings.filterwarnings("ignore")
|
| 15 |
|
|
|
|
| 17 |
from image_processing_gpu import init_ocr_model, create_white_canvas, cleanup_memory
|
| 18 |
from game_engine import MathGame, export_to_clean_dataset
|
| 19 |
|
| 20 |
+
print("🚀 Initialisation Calcul OCR v3.0 (Ultra-simplifié)...")
|
| 21 |
print("🔄 Chargement modèle TrOCR...")
|
| 22 |
init_ocr_model()
|
| 23 |
print("✅ Modèle TrOCR prêt")
|
|
|
|
| 27 |
return MathGame()
|
| 28 |
|
| 29 |
def start_game_wrapper(duration: str, operation: str, difficulty: str, game_state) -> tuple:
|
| 30 |
+
"""Démarre une nouvelle session de jeu"""
|
| 31 |
if game_state is None:
|
| 32 |
game_state = create_new_game_session()
|
| 33 |
|
| 34 |
cleanup_memory()
|
| 35 |
result = game_state.start_game(duration, operation, difficulty)
|
| 36 |
+
return result + (game_state, "")
|
| 37 |
|
| 38 |
def next_question_wrapper(image_data, game_state) -> tuple:
|
| 39 |
+
"""Passe à la question suivante - Stockage simple, pas d'OCR"""
|
| 40 |
if game_state is None:
|
| 41 |
game_state = create_new_game_session()
|
| 42 |
|
| 43 |
result = game_state.next_question(image_data)
|
| 44 |
+
return result + (game_state,)
|
| 45 |
|
| 46 |
def export_current_session(game_state) -> str:
|
| 47 |
+
"""Export de la session vers le dataset"""
|
| 48 |
if game_state is None:
|
| 49 |
return "❌ Aucune session active"
|
| 50 |
|
|
|
|
| 79 |
game_state.export_status = "not_exported"
|
| 80 |
return f"❌ Erreur export: {str(e)}"
|
| 81 |
|
| 82 |
+
# Interface Gradio
|
| 83 |
with gr.Blocks(
|
| 84 |
title="🧮 Calcul OCR - Entraînement mathématiques",
|
| 85 |
theme=gr.themes.Soft(),
|
|
|
|
| 109 |
head="<meta name='viewport' content='width=device-width, initial-scale=1.0'>"
|
| 110 |
) as demo:
|
| 111 |
|
| 112 |
+
# State management - Isolation par utilisateur
|
| 113 |
game_state = gr.State(value=None)
|
| 114 |
|
| 115 |
gr.Markdown(
|
|
|
|
| 123 |
4. Cliquez sur **➡️ NEXT !** pour la question suivante
|
| 124 |
|
| 125 |
À la fin, vous pourrez consulter vos résultats et contribuer au dataset ouvert !
|
| 126 |
+
|
| 127 |
+
🚀 **Version ultra-optimisée** : OCR en fin de session pour une fluidité maximale !
|
| 128 |
"""
|
| 129 |
)
|
| 130 |
|
|
|
|
| 180 |
# Résultats
|
| 181 |
results_display = gr.HTML("")
|
| 182 |
|
| 183 |
+
# Export vers dataset
|
| 184 |
export_button = gr.Button("📤 Exporter la série vers le dataset", variant="primary", size="lg")
|
| 185 |
export_status = gr.Markdown("")
|
| 186 |
|
| 187 |
+
# Événements
|
| 188 |
go_button.click(
|
| 189 |
fn=start_game_wrapper,
|
| 190 |
inputs=[duration_choice, operation_choice, difficulty_choice, game_state],
|
|
|
|
| 204 |
)
|
| 205 |
|
| 206 |
if __name__ == "__main__":
|
| 207 |
+
print("🚀 Lancement Calcul OCR v3.0 - Ultra-simplifié")
|
| 208 |
print("🎯 ZeroGPU HuggingFace Spaces")
|
| 209 |
+
print("⚡ Performance optimale : OCR en fin de session")
|
| 210 |
print("🎯 Dataset: calcul_ocr_dataset")
|
| 211 |
print("📊 Opérations: ×, +, -, ÷, Aléatoire")
|
| 212 |
print("⚙️ Durées: 30s, 60s")
|
game_engine.py
CHANGED
|
@@ -1,9 +1,10 @@
|
|
| 1 |
# ==========================================
|
| 2 |
-
# game_engine.py - Calcul OCR v3.0
|
| 3 |
# ==========================================
|
| 4 |
|
| 5 |
"""
|
| 6 |
-
Moteur de jeu mathématique
|
|
|
|
| 7 |
"""
|
| 8 |
|
| 9 |
import random
|
|
@@ -15,9 +16,6 @@ import uuid
|
|
| 15 |
import gc
|
| 16 |
import numpy as np
|
| 17 |
from PIL import Image
|
| 18 |
-
import threading
|
| 19 |
-
import queue
|
| 20 |
-
from typing import Dict, Tuple, Optional
|
| 21 |
|
| 22 |
# Import GPU uniquement
|
| 23 |
from image_processing_gpu import (
|
|
@@ -25,19 +23,10 @@ from image_processing_gpu import (
|
|
| 25 |
create_thumbnail_fast,
|
| 26 |
create_white_canvas,
|
| 27 |
cleanup_memory,
|
| 28 |
-
log_memory_usage,
|
| 29 |
get_ocr_model_info
|
| 30 |
)
|
| 31 |
|
| 32 |
-
print("✅ Game Engine: Mode GPU
|
| 33 |
-
|
| 34 |
-
# Récupérer les infos du modèle OCR
|
| 35 |
-
try:
|
| 36 |
-
ocr_info = get_ocr_model_info()
|
| 37 |
-
print(f"🎯 OCR sélectionné: {ocr_info['model_name']} sur {ocr_info['device']}")
|
| 38 |
-
except Exception as e:
|
| 39 |
-
print(f"⚠️ Impossible de récupérer les infos OCR: {e}")
|
| 40 |
-
ocr_info = {"model_name": "TrOCR", "device": "ZeroGPU"}
|
| 41 |
|
| 42 |
# Imports dataset
|
| 43 |
try:
|
|
@@ -60,22 +49,17 @@ DIFFICULTY_RANGES = {
|
|
| 60 |
}
|
| 61 |
|
| 62 |
def create_result_row_with_images(i: int, image: dict | np.ndarray | Image.Image, expected: int, operation_data: tuple[int, int, str, int]) -> dict:
|
|
|
|
| 63 |
|
| 64 |
-
print(f"🔍
|
| 65 |
-
print(f"🔍 Expected: {expected}")
|
| 66 |
-
print(f"🔍 Image type: {type(image)}")
|
| 67 |
|
| 68 |
-
# OCR
|
| 69 |
recognized, optimized_image, dataset_image_data = recognize_number_fast_with_image(image, debug=True)
|
| 70 |
|
| 71 |
-
print(f"🔍 OCR recognized: '{recognized}' (type: {type(recognized)})")
|
| 72 |
-
|
| 73 |
try:
|
| 74 |
recognized_num = int(recognized) if recognized.isdigit() else 0
|
| 75 |
except:
|
| 76 |
recognized_num = 0
|
| 77 |
-
|
| 78 |
-
print(f"🔍 OCR parsed num: {recognized_num}")
|
| 79 |
|
| 80 |
is_correct = recognized_num == expected
|
| 81 |
a, b, operation, correct_result = operation_data
|
|
@@ -84,10 +68,10 @@ def create_result_row_with_images(i: int, image: dict | np.ndarray | Image.Image
|
|
| 84 |
status_text = "Correct" if is_correct else "Incorrect"
|
| 85 |
row_color = "#e8f5e8" if is_correct else "#ffe8e8"
|
| 86 |
|
| 87 |
-
# Miniature
|
| 88 |
image_thumbnail = create_thumbnail_fast(optimized_image, size=(50, 50))
|
| 89 |
|
| 90 |
-
# Libérer mémoire
|
| 91 |
if optimized_image and hasattr(optimized_image, 'close'):
|
| 92 |
try:
|
| 93 |
optimized_image.close()
|
|
@@ -115,7 +99,7 @@ def create_result_row_with_images(i: int, image: dict | np.ndarray | Image.Image
|
|
| 115 |
|
| 116 |
|
| 117 |
class MathGame:
|
| 118 |
-
"""Moteur de jeu mathématique
|
| 119 |
|
| 120 |
def __init__(self):
|
| 121 |
self.is_running = False
|
|
@@ -138,12 +122,6 @@ class MathGame:
|
|
| 138 |
self.export_status = "not_exported"
|
| 139 |
self.export_timestamp = None
|
| 140 |
self.export_result = None
|
| 141 |
-
|
| 142 |
-
# Traitement parallèle
|
| 143 |
-
self.processing_queue = queue.Queue()
|
| 144 |
-
self.results_cache: Dict[int, dict] = {}
|
| 145 |
-
self.worker_thread: Optional[threading.Thread] = None
|
| 146 |
-
self.processing_active = False
|
| 147 |
|
| 148 |
def get_export_status(self) -> dict[str, str | bool | None]:
|
| 149 |
return {
|
|
@@ -161,54 +139,6 @@ class MathGame:
|
|
| 161 |
self.export_status = "exported"
|
| 162 |
self.export_result = result
|
| 163 |
|
| 164 |
-
def _start_background_processing(self) -> None:
|
| 165 |
-
"""Démarre le thread de traitement en arrière-plan"""
|
| 166 |
-
if self.worker_thread is None or not self.worker_thread.is_alive():
|
| 167 |
-
self.processing_active = True
|
| 168 |
-
self.worker_thread = threading.Thread(target=self._process_images_worker, daemon=True)
|
| 169 |
-
self.worker_thread.start()
|
| 170 |
-
print("🔄 Thread de traitement GPU parallèle démarré")
|
| 171 |
-
|
| 172 |
-
def _stop_background_processing(self) -> None:
|
| 173 |
-
"""Arrête le thread de traitement"""
|
| 174 |
-
self.processing_active = False
|
| 175 |
-
if self.worker_thread and self.worker_thread.is_alive():
|
| 176 |
-
print("⏹️ Arrêt du thread de traitement GPU parallèle")
|
| 177 |
-
|
| 178 |
-
def _process_images_worker(self) -> None:
|
| 179 |
-
"""Worker thread qui traite les images en arrière-plan avec GPU"""
|
| 180 |
-
print("🚀 Worker thread GPU démarré")
|
| 181 |
-
while self.processing_active:
|
| 182 |
-
try:
|
| 183 |
-
if not self.processing_queue.empty():
|
| 184 |
-
question_num, image, expected, operation_data = self.processing_queue.get(timeout=1)
|
| 185 |
-
print(f"🔄 Traitement GPU parallèle image {question_num}...")
|
| 186 |
-
|
| 187 |
-
start_time = time.time()
|
| 188 |
-
result_data = create_result_row_with_images(question_num, image, expected, operation_data)
|
| 189 |
-
processing_time = time.time() - start_time
|
| 190 |
-
|
| 191 |
-
# Stocker le résultat
|
| 192 |
-
self.results_cache[question_num] = result_data
|
| 193 |
-
print(f"✅ Image {question_num} traitée en {processing_time:.1f}s (GPU parallèle)")
|
| 194 |
-
|
| 195 |
-
else:
|
| 196 |
-
time.sleep(0.1)
|
| 197 |
-
|
| 198 |
-
except queue.Empty:
|
| 199 |
-
continue
|
| 200 |
-
except Exception as e:
|
| 201 |
-
print(f"❌ Erreur traitement GPU parallèle: {e}")
|
| 202 |
-
|
| 203 |
-
print("🛑 Worker thread GPU terminé")
|
| 204 |
-
|
| 205 |
-
def _add_image_to_processing_queue(self, question_num: int, image: dict | np.ndarray | Image.Image,
|
| 206 |
-
expected: int, operation_data: tuple) -> None:
|
| 207 |
-
"""Ajoute une image à la queue de traitement GPU"""
|
| 208 |
-
if image is not None:
|
| 209 |
-
self.processing_queue.put((question_num, image, expected, operation_data))
|
| 210 |
-
print(f"📝 Image {question_num} ajoutée à la queue de traitement GPU")
|
| 211 |
-
|
| 212 |
def generate_multiplication(self, difficulty: str) -> tuple[str, int]:
|
| 213 |
"""Génère une multiplication"""
|
| 214 |
min_val, max_val = DIFFICULTY_RANGES["×"][difficulty]
|
|
@@ -265,7 +195,7 @@ class MathGame:
|
|
| 265 |
self.operation_type = operation
|
| 266 |
self.difficulty = difficulty
|
| 267 |
|
| 268 |
-
# Nettoyage
|
| 269 |
if hasattr(self, 'user_images') and self.user_images:
|
| 270 |
for img in self.user_images:
|
| 271 |
if hasattr(img, 'close'):
|
|
@@ -274,21 +204,7 @@ class MathGame:
|
|
| 274 |
except:
|
| 275 |
pass
|
| 276 |
|
| 277 |
-
|
| 278 |
-
for entry in self.session_data:
|
| 279 |
-
if 'user_drawing' in entry and entry['user_drawing']:
|
| 280 |
-
entry['user_drawing'] = None
|
| 281 |
-
self.session_data.clear()
|
| 282 |
-
|
| 283 |
-
# Réinit avec nettoyage parallèle
|
| 284 |
-
self._stop_background_processing()
|
| 285 |
-
self.results_cache.clear()
|
| 286 |
-
while not self.processing_queue.empty():
|
| 287 |
-
try:
|
| 288 |
-
self.processing_queue.get_nowait()
|
| 289 |
-
except queue.Empty:
|
| 290 |
-
break
|
| 291 |
-
|
| 292 |
self.is_running = True
|
| 293 |
self.start_time = time.time()
|
| 294 |
self.user_images = []
|
|
@@ -303,9 +219,6 @@ class MathGame:
|
|
| 303 |
self.export_timestamp = None
|
| 304 |
self.export_result = None
|
| 305 |
|
| 306 |
-
# Démarrer le traitement parallèle GPU
|
| 307 |
-
self._start_background_processing()
|
| 308 |
-
|
| 309 |
gc.collect()
|
| 310 |
|
| 311 |
# Première opération
|
|
@@ -318,7 +231,7 @@ class MathGame:
|
|
| 318 |
a, op, b = int(parts[0]), parts[1], int(parts[2])
|
| 319 |
self.operations_history.append((a, b, op, answer))
|
| 320 |
|
| 321 |
-
# Affichage
|
| 322 |
operation_emoji = {
|
| 323 |
"×": "✖️", "+": "➕", "-": "➖", "÷": "➗", "Aléatoire": "🎲"
|
| 324 |
}
|
|
@@ -335,6 +248,8 @@ class MathGame:
|
|
| 335 |
)
|
| 336 |
|
| 337 |
def next_question(self, image_data: dict | np.ndarray | Image.Image | None) -> tuple[str, Image.Image, str, str, gr.update, gr.update, str]:
|
|
|
|
|
|
|
| 338 |
if not self.is_running:
|
| 339 |
return (
|
| 340 |
f'<div style="font-size: 3em; font-weight: bold; text-align: center; padding: 20px; background: linear-gradient(45deg, #667eea 0%, #764ba2 100%); color: white; border-radius: 10px;">{self.current_operation}</div>',
|
|
@@ -350,20 +265,12 @@ class MathGame:
|
|
| 350 |
if elapsed_time >= self.duration:
|
| 351 |
return self.end_game(image_data)
|
| 352 |
|
|
|
|
| 353 |
if image_data is not None:
|
| 354 |
-
# Ajouter l'image à la liste ET au traitement parallèle GPU
|
| 355 |
self.user_images.append(image_data)
|
| 356 |
self.expected_answers.append(self.correct_answer)
|
| 357 |
-
|
| 358 |
-
# Parser l'opération actuelle pour le traitement
|
| 359 |
-
parts = self.current_operation.split()
|
| 360 |
-
a, op, b = int(parts[0]), parts[1], int(parts[2])
|
| 361 |
-
current_operation_data = (a, b, op, self.correct_answer)
|
| 362 |
-
|
| 363 |
-
# Lancer le traitement GPU en parallèle de l'image qu'on vient de recevoir
|
| 364 |
-
self._add_image_to_processing_queue(self.question_count, image_data, self.correct_answer, current_operation_data)
|
| 365 |
-
|
| 366 |
self.question_count += 1
|
|
|
|
| 367 |
|
| 368 |
# Nouvelle opération
|
| 369 |
operation_str, answer = self.generate_operation(self.operation_type, self.difficulty)
|
|
@@ -398,47 +305,27 @@ class MathGame:
|
|
| 398 |
)
|
| 399 |
|
| 400 |
def end_game(self, final_image: dict | np.ndarray | Image.Image | None) -> tuple[str, Image.Image, str, str, gr.update, gr.update, str]:
|
|
|
|
| 401 |
|
| 402 |
self.is_running = False
|
| 403 |
|
| 404 |
-
|
| 405 |
-
self._stop_background_processing()
|
| 406 |
-
|
| 407 |
-
print("🏁 Fin de jeu - Assemblage des résultats GPU...")
|
| 408 |
|
|
|
|
| 409 |
if final_image is not None:
|
| 410 |
self.user_images.append(final_image)
|
| 411 |
self.expected_answers.append(self.correct_answer)
|
| 412 |
-
|
| 413 |
-
# Traitement de la dernière image
|
| 414 |
-
parts = self.current_operation.split()
|
| 415 |
-
a, op, b = int(parts[0]), parts[1], int(parts[2])
|
| 416 |
-
final_operation_data = (a, b, op, self.correct_answer)
|
| 417 |
-
|
| 418 |
-
# Traiter la dernière image immédiatement avec GPU
|
| 419 |
-
print(f"🔄 Traitement GPU final de l'image {self.question_count}...")
|
| 420 |
-
final_result = create_result_row_with_images(self.question_count, final_image, self.correct_answer, final_operation_data)
|
| 421 |
-
self.results_cache[self.question_count] = final_result
|
| 422 |
-
|
| 423 |
self.question_count += 1
|
|
|
|
|
|
|
| 424 |
if len(self.operations_history) < len(self.user_images):
|
|
|
|
|
|
|
| 425 |
self.operations_history.append((a, b, op, self.correct_answer))
|
| 426 |
|
| 427 |
-
#
|
| 428 |
-
max_wait = 10
|
| 429 |
-
wait_start = time.time()
|
| 430 |
-
expected_results = len(self.user_images)
|
| 431 |
-
|
| 432 |
-
print(f"⏳ Attente de {expected_results} résultats GPU...")
|
| 433 |
-
while len(self.results_cache) < expected_results and (time.time() - wait_start) < max_wait:
|
| 434 |
-
time.sleep(0.1)
|
| 435 |
-
|
| 436 |
-
results_ready = len(self.results_cache)
|
| 437 |
-
print(f"✅ {results_ready}/{expected_results} résultats GPU prêts")
|
| 438 |
-
|
| 439 |
-
# Assembler les résultats dans l'ordre
|
| 440 |
-
correct_answers = 0
|
| 441 |
total_questions = len(self.user_images)
|
|
|
|
| 442 |
table_rows_html = ""
|
| 443 |
|
| 444 |
session_timestamp = datetime.datetime.now().isoformat()
|
|
@@ -447,36 +334,30 @@ class MathGame:
|
|
| 447 |
self.session_data = []
|
| 448 |
images_saved = 0
|
| 449 |
|
| 450 |
-
print(f"
|
| 451 |
|
|
|
|
| 452 |
for i in range(total_questions):
|
| 453 |
-
|
| 454 |
-
|
| 455 |
-
|
| 456 |
-
|
| 457 |
-
|
| 458 |
-
|
| 459 |
-
|
| 460 |
-
else
|
| 461 |
-
|
| 462 |
-
'html_row': f'<tr><td>{i+1}</td><td colspan="7">Erreur traitement</td></tr>',
|
| 463 |
-
'is_correct': False,
|
| 464 |
-
'recognized': "0",
|
| 465 |
-
'recognized_num': 0,
|
| 466 |
-
'dataset_image_data': None
|
| 467 |
-
}
|
| 468 |
|
| 469 |
table_rows_html += row_data['html_row']
|
| 470 |
|
| 471 |
if row_data['is_correct']:
|
| 472 |
correct_answers += 1
|
| 473 |
|
| 474 |
-
# Structure pour dataset
|
| 475 |
a, b, operation, correct_result = self.operations_history[i] if i < len(self.operations_history) else (0, 0, "×", 0)
|
| 476 |
|
| 477 |
try:
|
| 478 |
ocr_info_data = get_ocr_model_info()
|
| 479 |
-
print(f"🔍 Debug OCR info GPU: {ocr_info_data}")
|
| 480 |
except Exception as e:
|
| 481 |
print(f"❌ Erreur get_ocr_model_info: {e}")
|
| 482 |
ocr_info_data = {"model_name": "TrOCR", "device": "ZeroGPU"}
|
|
@@ -491,21 +372,19 @@ class MathGame:
|
|
| 491 |
"operand_a": a,
|
| 492 |
"operand_b": b,
|
| 493 |
"operation": operation,
|
| 494 |
-
"correct_answer": self.expected_answers[i]
|
| 495 |
"ocr_model": ocr_info_data.get("model_name", "TrOCR"),
|
| 496 |
"ocr_device": ocr_info_data.get("device", "ZeroGPU"),
|
| 497 |
"user_answer_ocr": row_data['recognized'],
|
| 498 |
"user_answer_parsed": row_data['recognized_num'],
|
| 499 |
"is_correct": row_data['is_correct'],
|
| 500 |
"total_questions": total_questions,
|
| 501 |
-
"app_version": "3.
|
| 502 |
}
|
| 503 |
|
| 504 |
-
print(f"🔍 Debug entry OCR GPU: ocr_model={entry['ocr_model']}, ocr_device={entry['ocr_device']}")
|
| 505 |
-
|
| 506 |
# Image PIL native pour dataset
|
| 507 |
if row_data['dataset_image_data']:
|
| 508 |
-
entry["handwriting_image"] = row_data['dataset_image_data']["handwriting_image"]
|
| 509 |
entry["image_width"] = int(row_data['dataset_image_data']["width"])
|
| 510 |
entry["image_height"] = int(row_data['dataset_image_data']["height"])
|
| 511 |
entry["has_image"] = True
|
|
@@ -517,10 +396,11 @@ class MathGame:
|
|
| 517 |
|
| 518 |
accuracy = (correct_answers / total_questions * 100) if total_questions > 0 else 0
|
| 519 |
|
|
|
|
| 520 |
for entry in self.session_data:
|
| 521 |
entry["session_accuracy"] = accuracy
|
| 522 |
|
| 523 |
-
# Nettoyage mémoire
|
| 524 |
for img in self.user_images:
|
| 525 |
if hasattr(img, 'close'):
|
| 526 |
try:
|
|
@@ -528,7 +408,9 @@ class MathGame:
|
|
| 528 |
except:
|
| 529 |
pass
|
| 530 |
|
| 531 |
-
|
|
|
|
|
|
|
| 532 |
|
| 533 |
# HTML résultats
|
| 534 |
table_html = f"""
|
|
@@ -542,7 +424,7 @@ class MathGame:
|
|
| 542 |
<th style="padding: 8px;">B</th>
|
| 543 |
<th style="padding: 8px;">Réponse</th>
|
| 544 |
<th style="padding: 8px;">Votre dessin</th>
|
| 545 |
-
<th style="padding: 8px;">OCR
|
| 546 |
<th style="padding: 8px;">Statut</th>
|
| 547 |
</tr>
|
| 548 |
</thead>
|
|
@@ -555,20 +437,16 @@ class MathGame:
|
|
| 555 |
|
| 556 |
# Configuration session pour affichage
|
| 557 |
config_display = f"{self.operation_type} • {self.difficulty} • {self.duration}s"
|
| 558 |
-
|
| 559 |
-
"×": "✖️", "+": "➕", "-": "➖", "÷": "➗", "Aléatoire": "🎲"
|
| 560 |
-
}
|
| 561 |
-
emoji = operation_emoji.get(self.operation_type, "🔢")
|
| 562 |
-
|
| 563 |
export_info = self.get_export_status()
|
| 564 |
if export_info["can_export"]:
|
| 565 |
export_section = f"""
|
| 566 |
<div style="margin-top: 20px; padding: 15px; background-color: #e8f5e8; border-radius: 8px;">
|
| 567 |
-
<h3 style="color: #2e7d32;">📊 Résumé de la série
|
| 568 |
<p style="color: #2e7d32;">
|
| 569 |
✅ {total_questions} réponses • 📊 {accuracy:.1f}% de précision<br>
|
| 570 |
-
🖼️ {images_saved} images
|
| 571 |
-
|
| 572 |
⚙️ Configuration: {config_display}
|
| 573 |
</p>
|
| 574 |
</div>
|
|
@@ -579,7 +457,7 @@ class MathGame:
|
|
| 579 |
final_results = f"""
|
| 580 |
<div style="margin: 20px 0;">
|
| 581 |
<div style="background: linear-gradient(45deg, #667eea 0%, #764ba2 100%); color: white; padding: 20px; border-radius: 10px; margin: 20px 0;">
|
| 582 |
-
<h2 style="text-align: center;
|
| 583 |
<div style="display: flex; justify-content: space-around; flex-wrap: wrap;">
|
| 584 |
<div style="text-align: center; margin: 10px;">
|
| 585 |
<div style="font-size: 2em; font-weight: bold;">{total_questions}</div>
|
|
@@ -599,7 +477,7 @@ class MathGame:
|
|
| 599 |
</div>
|
| 600 |
</div>
|
| 601 |
</div>
|
| 602 |
-
<h2 style="color: #4a90e2;">📊 Détail des Réponses (TrOCR
|
| 603 |
{table_html}
|
| 604 |
{export_section}
|
| 605 |
</div>
|
|
@@ -608,7 +486,7 @@ class MathGame:
|
|
| 608 |
return (
|
| 609 |
"""<div style="font-size: 3em; font-weight: bold; text-align: center; padding: 20px; background: linear-gradient(45deg, #667eea 0%, #764ba2 100%); color: white; border-radius: 10px;">🏁 C'est fini !</div>""",
|
| 610 |
create_white_canvas(),
|
| 611 |
-
f"✨ Session
|
| 612 |
"⏱️ Temps écoulé !",
|
| 613 |
gr.update(interactive=True),
|
| 614 |
gr.update(interactive=False),
|
|
@@ -617,7 +495,7 @@ class MathGame:
|
|
| 617 |
|
| 618 |
|
| 619 |
def export_to_clean_dataset(session_data: list[dict], dataset_name: str = None) -> str:
|
| 620 |
-
"""Export vers le dataset
|
| 621 |
if dataset_name is None:
|
| 622 |
dataset_name = DATASET_NAME
|
| 623 |
|
|
@@ -626,71 +504,40 @@ def export_to_clean_dataset(session_data: list[dict], dataset_name: str = None)
|
|
| 626 |
|
| 627 |
hf_token = os.getenv("HF_TOKEN") or os.getenv("tk_calcul_ocr")
|
| 628 |
if not hf_token:
|
| 629 |
-
return "❌ Token HuggingFace manquant
|
| 630 |
|
| 631 |
try:
|
| 632 |
-
print(f"\n🚀 === EXPORT
|
| 633 |
print(f"📊 Dataset: {dataset_name}")
|
| 634 |
|
| 635 |
-
# Filtrer les entrées avec images
|
| 636 |
-
clean_entries = []
|
| 637 |
-
|
| 638 |
-
# Récupérer les infos OCR GPU pour toute la session
|
| 639 |
-
try:
|
| 640 |
-
global_ocr_info = get_ocr_model_info()
|
| 641 |
-
print(f"🔍 Infos OCR GPU globales: {global_ocr_info}")
|
| 642 |
-
except Exception as e:
|
| 643 |
-
print(f"❌ Erreur infos OCR GPU: {e}")
|
| 644 |
-
global_ocr_info = {"model_name": "TrOCR", "device": "ZeroGPU"}
|
| 645 |
-
|
| 646 |
-
for entry in session_data:
|
| 647 |
-
if entry.get('has_image', False):
|
| 648 |
-
# Ajouter explicitement les champs OCR GPU
|
| 649 |
-
entry_with_ocr = entry.copy()
|
| 650 |
-
entry_with_ocr["ocr_model"] = global_ocr_info.get("model_name", "TrOCR")
|
| 651 |
-
entry_with_ocr["ocr_device"] = global_ocr_info.get("device", "ZeroGPU")
|
| 652 |
-
|
| 653 |
-
print(f"🔍 Entry GPU avec OCR: ocr_model={entry_with_ocr['ocr_model']}, ocr_device={entry_with_ocr['ocr_device']}")
|
| 654 |
-
print(f"🖼️ Image type: {type(entry_with_ocr.get('handwriting_image', 'None'))}")
|
| 655 |
-
clean_entries.append(entry_with_ocr)
|
| 656 |
|
| 657 |
if len(clean_entries) == 0:
|
| 658 |
return "❌ Aucune entrée avec image à exporter"
|
| 659 |
|
| 660 |
-
# Vérifier la structure de la première entrée
|
| 661 |
-
sample_entry = clean_entries[0]
|
| 662 |
-
print(f"🔍 Structure première entrée GPU: {list(sample_entry.keys())}")
|
| 663 |
-
print(f"🔍 OCR GPU dans entrée: ocr_model={sample_entry.get('ocr_model', 'MISSING')}, ocr_device={sample_entry.get('ocr_device', 'MISSING')}")
|
| 664 |
-
|
| 665 |
# Charger dataset existant et combiner
|
| 666 |
try:
|
| 667 |
existing_dataset = load_dataset(dataset_name, split="train")
|
| 668 |
existing_data = existing_dataset.to_list()
|
| 669 |
-
print(f"📊 {len(existing_data)} entrées existantes
|
| 670 |
|
| 671 |
-
# Combiner ancien + nouveau
|
| 672 |
combined_data = existing_data + clean_entries
|
| 673 |
clean_dataset = Dataset.from_list(combined_data)
|
| 674 |
-
print(f"📊 Dataset combiné: {len(
|
| 675 |
|
| 676 |
except Exception as e:
|
| 677 |
-
print(f"📊
|
| 678 |
-
# Si le dataset n'existe pas, créer depuis les nouvelles entrées
|
| 679 |
clean_dataset = Dataset.from_list(clean_entries)
|
| 680 |
-
print(f"📊 Nouveau dataset créé avec {len(clean_entries)} entrées")
|
| 681 |
|
| 682 |
-
#
|
| 683 |
try:
|
| 684 |
clean_dataset = clean_dataset.cast_column("handwriting_image", DatasetImage())
|
| 685 |
-
print("✅ Colonne
|
| 686 |
except Exception as e:
|
| 687 |
-
print(f"⚠️
|
| 688 |
-
|
| 689 |
-
print(f"✅ Dataset GPU créé - Features:")
|
| 690 |
-
for feature_name in clean_dataset.features:
|
| 691 |
-
print(f" - {feature_name}: {clean_dataset.features[feature_name]}")
|
| 692 |
|
| 693 |
-
# Statistiques
|
| 694 |
operations_count = {}
|
| 695 |
for entry in clean_entries:
|
| 696 |
op = entry.get('operation_type', 'unknown')
|
|
@@ -699,30 +546,26 @@ def export_to_clean_dataset(session_data: list[dict], dataset_name: str = None)
|
|
| 699 |
operations_summary = ", ".join([f"{op}: {count}" for op, count in operations_count.items()])
|
| 700 |
|
| 701 |
# Push vers HuggingFace
|
| 702 |
-
print(f"📤 Push
|
| 703 |
clean_dataset.push_to_hub(
|
| 704 |
dataset_name,
|
| 705 |
private=False,
|
| 706 |
token=hf_token,
|
| 707 |
-
commit_message=f"Add {len(clean_entries)}
|
| 708 |
)
|
| 709 |
|
| 710 |
cleanup_memory()
|
| 711 |
|
| 712 |
-
return f"""### ✅ Session
|
| 713 |
|
| 714 |
📊 **Dataset:** {dataset_name}
|
| 715 |
-
🖼️ **Images
|
| 716 |
-
🎯 **OCR:** TrOCR ZeroGPU
|
| 717 |
🔢 **Opérations:** {operations_summary}
|
| 718 |
📈 **Total:** {len(clean_dataset)}
|
| 719 |
|
| 720 |
-
🔗
|
| 721 |
"""
|
| 722 |
-
|
| 723 |
|
| 724 |
except Exception as e:
|
| 725 |
-
print(f"❌ ERREUR
|
| 726 |
-
|
| 727 |
-
traceback.print_exc()
|
| 728 |
-
return f"❌ Erreur GPU: {str(e)}"
|
|
|
|
| 1 |
# ==========================================
|
| 2 |
+
# game_engine.py - Calcul OCR v3.0 ULTRA SIMPLIFIÉ
|
| 3 |
# ==========================================
|
| 4 |
|
| 5 |
"""
|
| 6 |
+
Moteur de jeu mathématique ultra-simplifié pour ZeroGPU
|
| 7 |
+
OCR en fin de session uniquement - Performance optimale
|
| 8 |
"""
|
| 9 |
|
| 10 |
import random
|
|
|
|
| 16 |
import gc
|
| 17 |
import numpy as np
|
| 18 |
from PIL import Image
|
|
|
|
|
|
|
|
|
|
| 19 |
|
| 20 |
# Import GPU uniquement
|
| 21 |
from image_processing_gpu import (
|
|
|
|
| 23 |
create_thumbnail_fast,
|
| 24 |
create_white_canvas,
|
| 25 |
cleanup_memory,
|
|
|
|
| 26 |
get_ocr_model_info
|
| 27 |
)
|
| 28 |
|
| 29 |
+
print("✅ Game Engine: Mode GPU ultra-simplifié")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
|
| 31 |
# Imports dataset
|
| 32 |
try:
|
|
|
|
| 49 |
}
|
| 50 |
|
| 51 |
def create_result_row_with_images(i: int, image: dict | np.ndarray | Image.Image, expected: int, operation_data: tuple[int, int, str, int]) -> dict:
|
| 52 |
+
"""Traite une image avec OCR et génère la ligne de résultat"""
|
| 53 |
|
| 54 |
+
print(f"🔍 Traitement OCR image #{i+1}")
|
|
|
|
|
|
|
| 55 |
|
| 56 |
+
# OCR avec TrOCR
|
| 57 |
recognized, optimized_image, dataset_image_data = recognize_number_fast_with_image(image, debug=True)
|
| 58 |
|
|
|
|
|
|
|
| 59 |
try:
|
| 60 |
recognized_num = int(recognized) if recognized.isdigit() else 0
|
| 61 |
except:
|
| 62 |
recognized_num = 0
|
|
|
|
|
|
|
| 63 |
|
| 64 |
is_correct = recognized_num == expected
|
| 65 |
a, b, operation, correct_result = operation_data
|
|
|
|
| 68 |
status_text = "Correct" if is_correct else "Incorrect"
|
| 69 |
row_color = "#e8f5e8" if is_correct else "#ffe8e8"
|
| 70 |
|
| 71 |
+
# Miniature pour affichage
|
| 72 |
image_thumbnail = create_thumbnail_fast(optimized_image, size=(50, 50))
|
| 73 |
|
| 74 |
+
# Libérer mémoire optimisée
|
| 75 |
if optimized_image and hasattr(optimized_image, 'close'):
|
| 76 |
try:
|
| 77 |
optimized_image.close()
|
|
|
|
| 99 |
|
| 100 |
|
| 101 |
class MathGame:
|
| 102 |
+
"""Moteur de jeu mathématique ultra-simplifié"""
|
| 103 |
|
| 104 |
def __init__(self):
|
| 105 |
self.is_running = False
|
|
|
|
| 122 |
self.export_status = "not_exported"
|
| 123 |
self.export_timestamp = None
|
| 124 |
self.export_result = None
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 125 |
|
| 126 |
def get_export_status(self) -> dict[str, str | bool | None]:
|
| 127 |
return {
|
|
|
|
| 139 |
self.export_status = "exported"
|
| 140 |
self.export_result = result
|
| 141 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 142 |
def generate_multiplication(self, difficulty: str) -> tuple[str, int]:
|
| 143 |
"""Génère une multiplication"""
|
| 144 |
min_val, max_val = DIFFICULTY_RANGES["×"][difficulty]
|
|
|
|
| 195 |
self.operation_type = operation
|
| 196 |
self.difficulty = difficulty
|
| 197 |
|
| 198 |
+
# Nettoyage simple
|
| 199 |
if hasattr(self, 'user_images') and self.user_images:
|
| 200 |
for img in self.user_images:
|
| 201 |
if hasattr(img, 'close'):
|
|
|
|
| 204 |
except:
|
| 205 |
pass
|
| 206 |
|
| 207 |
+
# Réinitialisation complète
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 208 |
self.is_running = True
|
| 209 |
self.start_time = time.time()
|
| 210 |
self.user_images = []
|
|
|
|
| 219 |
self.export_timestamp = None
|
| 220 |
self.export_result = None
|
| 221 |
|
|
|
|
|
|
|
|
|
|
| 222 |
gc.collect()
|
| 223 |
|
| 224 |
# Première opération
|
|
|
|
| 231 |
a, op, b = int(parts[0]), parts[1], int(parts[2])
|
| 232 |
self.operations_history.append((a, b, op, answer))
|
| 233 |
|
| 234 |
+
# Affichage
|
| 235 |
operation_emoji = {
|
| 236 |
"×": "✖️", "+": "➕", "-": "➖", "÷": "➗", "Aléatoire": "🎲"
|
| 237 |
}
|
|
|
|
| 248 |
)
|
| 249 |
|
| 250 |
def next_question(self, image_data: dict | np.ndarray | Image.Image | None) -> tuple[str, Image.Image, str, str, gr.update, gr.update, str]:
|
| 251 |
+
"""Passe à la question suivante - STOCKAGE SIMPLE, PAS D'OCR"""
|
| 252 |
+
|
| 253 |
if not self.is_running:
|
| 254 |
return (
|
| 255 |
f'<div style="font-size: 3em; font-weight: bold; text-align: center; padding: 20px; background: linear-gradient(45deg, #667eea 0%, #764ba2 100%); color: white; border-radius: 10px;">{self.current_operation}</div>',
|
|
|
|
| 265 |
if elapsed_time >= self.duration:
|
| 266 |
return self.end_game(image_data)
|
| 267 |
|
| 268 |
+
# STOCKAGE SIMPLE - PAS D'OCR pendant le jeu !
|
| 269 |
if image_data is not None:
|
|
|
|
| 270 |
self.user_images.append(image_data)
|
| 271 |
self.expected_answers.append(self.correct_answer)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 272 |
self.question_count += 1
|
| 273 |
+
print(f"📝 Image {self.question_count} stockée (pas d'OCR pendant le jeu)")
|
| 274 |
|
| 275 |
# Nouvelle opération
|
| 276 |
operation_str, answer = self.generate_operation(self.operation_type, self.difficulty)
|
|
|
|
| 305 |
)
|
| 306 |
|
| 307 |
def end_game(self, final_image: dict | np.ndarray | Image.Image | None) -> tuple[str, Image.Image, str, str, gr.update, gr.update, str]:
|
| 308 |
+
"""Fin de jeu - OCR DE TOUTES LES IMAGES EN SÉQUENTIEL"""
|
| 309 |
|
| 310 |
self.is_running = False
|
| 311 |
|
| 312 |
+
print("🏁 Fin de jeu - Début OCR séquentiel de toutes les images...")
|
|
|
|
|
|
|
|
|
|
| 313 |
|
| 314 |
+
# Ajouter la dernière image si présente
|
| 315 |
if final_image is not None:
|
| 316 |
self.user_images.append(final_image)
|
| 317 |
self.expected_answers.append(self.correct_answer)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 318 |
self.question_count += 1
|
| 319 |
+
|
| 320 |
+
# Ajouter l'opération finale à l'historique si nécessaire
|
| 321 |
if len(self.operations_history) < len(self.user_images):
|
| 322 |
+
parts = self.current_operation.split()
|
| 323 |
+
a, op, b = int(parts[0]), parts[1], int(parts[2])
|
| 324 |
self.operations_history.append((a, b, op, self.correct_answer))
|
| 325 |
|
| 326 |
+
# OCR SÉQUENTIEL SIMPLE de toutes les images
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 327 |
total_questions = len(self.user_images)
|
| 328 |
+
correct_answers = 0
|
| 329 |
table_rows_html = ""
|
| 330 |
|
| 331 |
session_timestamp = datetime.datetime.now().isoformat()
|
|
|
|
| 334 |
self.session_data = []
|
| 335 |
images_saved = 0
|
| 336 |
|
| 337 |
+
print(f"🔄 Traitement OCR séquentiel de {total_questions} images...")
|
| 338 |
|
| 339 |
+
# Boucle simple - une image à la fois
|
| 340 |
for i in range(total_questions):
|
| 341 |
+
print(f"📷 OCR image {i+1}/{total_questions}...")
|
| 342 |
+
|
| 343 |
+
# OCR de cette image
|
| 344 |
+
row_data = create_result_row_with_images(
|
| 345 |
+
i,
|
| 346 |
+
self.user_images[i],
|
| 347 |
+
self.expected_answers[i],
|
| 348 |
+
self.operations_history[i] if i < len(self.operations_history) else (0, 0, "×", 0)
|
| 349 |
+
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 350 |
|
| 351 |
table_rows_html += row_data['html_row']
|
| 352 |
|
| 353 |
if row_data['is_correct']:
|
| 354 |
correct_answers += 1
|
| 355 |
|
| 356 |
+
# Structure pour dataset
|
| 357 |
a, b, operation, correct_result = self.operations_history[i] if i < len(self.operations_history) else (0, 0, "×", 0)
|
| 358 |
|
| 359 |
try:
|
| 360 |
ocr_info_data = get_ocr_model_info()
|
|
|
|
| 361 |
except Exception as e:
|
| 362 |
print(f"❌ Erreur get_ocr_model_info: {e}")
|
| 363 |
ocr_info_data = {"model_name": "TrOCR", "device": "ZeroGPU"}
|
|
|
|
| 372 |
"operand_a": a,
|
| 373 |
"operand_b": b,
|
| 374 |
"operation": operation,
|
| 375 |
+
"correct_answer": self.expected_answers[i],
|
| 376 |
"ocr_model": ocr_info_data.get("model_name", "TrOCR"),
|
| 377 |
"ocr_device": ocr_info_data.get("device", "ZeroGPU"),
|
| 378 |
"user_answer_ocr": row_data['recognized'],
|
| 379 |
"user_answer_parsed": row_data['recognized_num'],
|
| 380 |
"is_correct": row_data['is_correct'],
|
| 381 |
"total_questions": total_questions,
|
| 382 |
+
"app_version": "3.0_calcul_ocr_ultra_simplified"
|
| 383 |
}
|
| 384 |
|
|
|
|
|
|
|
| 385 |
# Image PIL native pour dataset
|
| 386 |
if row_data['dataset_image_data']:
|
| 387 |
+
entry["handwriting_image"] = row_data['dataset_image_data']["handwriting_image"]
|
| 388 |
entry["image_width"] = int(row_data['dataset_image_data']["width"])
|
| 389 |
entry["image_height"] = int(row_data['dataset_image_data']["height"])
|
| 390 |
entry["has_image"] = True
|
|
|
|
| 396 |
|
| 397 |
accuracy = (correct_answers / total_questions * 100) if total_questions > 0 else 0
|
| 398 |
|
| 399 |
+
# Ajouter accuracy à toutes les entrées
|
| 400 |
for entry in self.session_data:
|
| 401 |
entry["session_accuracy"] = accuracy
|
| 402 |
|
| 403 |
+
# Nettoyage mémoire
|
| 404 |
for img in self.user_images:
|
| 405 |
if hasattr(img, 'close'):
|
| 406 |
try:
|
|
|
|
| 408 |
except:
|
| 409 |
pass
|
| 410 |
|
| 411 |
+
cleanup_memory()
|
| 412 |
+
|
| 413 |
+
print(f"✅ OCR terminé: {correct_answers}/{total_questions} correct ({accuracy:.1f}%)")
|
| 414 |
|
| 415 |
# HTML résultats
|
| 416 |
table_html = f"""
|
|
|
|
| 424 |
<th style="padding: 8px;">B</th>
|
| 425 |
<th style="padding: 8px;">Réponse</th>
|
| 426 |
<th style="padding: 8px;">Votre dessin</th>
|
| 427 |
+
<th style="padding: 8px;">OCR</th>
|
| 428 |
<th style="padding: 8px;">Statut</th>
|
| 429 |
</tr>
|
| 430 |
</thead>
|
|
|
|
| 437 |
|
| 438 |
# Configuration session pour affichage
|
| 439 |
config_display = f"{self.operation_type} • {self.difficulty} • {self.duration}s"
|
| 440 |
+
|
|
|
|
|
|
|
|
|
|
|
|
|
| 441 |
export_info = self.get_export_status()
|
| 442 |
if export_info["can_export"]:
|
| 443 |
export_section = f"""
|
| 444 |
<div style="margin-top: 20px; padding: 15px; background-color: #e8f5e8; border-radius: 8px;">
|
| 445 |
+
<h3 style="color: #2e7d32;">📊 Résumé de la série</h3>
|
| 446 |
<p style="color: #2e7d32;">
|
| 447 |
✅ {total_questions} réponses • 📊 {accuracy:.1f}% de précision<br>
|
| 448 |
+
🖼️ {images_saved} images sauvegardées<br>
|
| 449 |
+
🤖 OCR: TrOCR ZeroGPU (séquentiel)<br>
|
| 450 |
⚙️ Configuration: {config_display}
|
| 451 |
</p>
|
| 452 |
</div>
|
|
|
|
| 457 |
final_results = f"""
|
| 458 |
<div style="margin: 20px 0;">
|
| 459 |
<div style="background: linear-gradient(45deg, #667eea 0%, #764ba2 100%); color: white; padding: 20px; border-radius: 10px; margin: 20px 0;">
|
| 460 |
+
<h2 style="text-align: center;">🎉 Session terminée !</h2>
|
| 461 |
<div style="display: flex; justify-content: space-around; flex-wrap: wrap;">
|
| 462 |
<div style="text-align: center; margin: 10px;">
|
| 463 |
<div style="font-size: 2em; font-weight: bold;">{total_questions}</div>
|
|
|
|
| 477 |
</div>
|
| 478 |
</div>
|
| 479 |
</div>
|
| 480 |
+
<h2 style="color: #4a90e2;">📊 Détail des Réponses (TrOCR séquentiel)</h2>
|
| 481 |
{table_html}
|
| 482 |
{export_section}
|
| 483 |
</div>
|
|
|
|
| 486 |
return (
|
| 487 |
"""<div style="font-size: 3em; font-weight: bold; text-align: center; padding: 20px; background: linear-gradient(45deg, #667eea 0%, #764ba2 100%); color: white; border-radius: 10px;">🏁 C'est fini !</div>""",
|
| 488 |
create_white_canvas(),
|
| 489 |
+
f"✨ Session {config_display} terminée !",
|
| 490 |
"⏱️ Temps écoulé !",
|
| 491 |
gr.update(interactive=True),
|
| 492 |
gr.update(interactive=False),
|
|
|
|
| 495 |
|
| 496 |
|
| 497 |
def export_to_clean_dataset(session_data: list[dict], dataset_name: str = None) -> str:
|
| 498 |
+
"""Export vers le dataset - Version simplifiée"""
|
| 499 |
if dataset_name is None:
|
| 500 |
dataset_name = DATASET_NAME
|
| 501 |
|
|
|
|
| 504 |
|
| 505 |
hf_token = os.getenv("HF_TOKEN") or os.getenv("tk_calcul_ocr")
|
| 506 |
if not hf_token:
|
| 507 |
+
return "❌ Token HuggingFace manquant"
|
| 508 |
|
| 509 |
try:
|
| 510 |
+
print(f"\n🚀 === EXPORT DATASET ULTRA-SIMPLIFIÉ ===")
|
| 511 |
print(f"📊 Dataset: {dataset_name}")
|
| 512 |
|
| 513 |
+
# Filtrer les entrées avec images
|
| 514 |
+
clean_entries = [entry for entry in session_data if entry.get('has_image', False)]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 515 |
|
| 516 |
if len(clean_entries) == 0:
|
| 517 |
return "❌ Aucune entrée avec image à exporter"
|
| 518 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 519 |
# Charger dataset existant et combiner
|
| 520 |
try:
|
| 521 |
existing_dataset = load_dataset(dataset_name, split="train")
|
| 522 |
existing_data = existing_dataset.to_list()
|
| 523 |
+
print(f"📊 {len(existing_data)} entrées existantes")
|
| 524 |
|
|
|
|
| 525 |
combined_data = existing_data + clean_entries
|
| 526 |
clean_dataset = Dataset.from_list(combined_data)
|
| 527 |
+
print(f"📊 Dataset combiné: {len(combined_data)} total")
|
| 528 |
|
| 529 |
except Exception as e:
|
| 530 |
+
print(f"📊 Nouveau dataset: {e}")
|
|
|
|
| 531 |
clean_dataset = Dataset.from_list(clean_entries)
|
|
|
|
| 532 |
|
| 533 |
+
# Conversion colonne image
|
| 534 |
try:
|
| 535 |
clean_dataset = clean_dataset.cast_column("handwriting_image", DatasetImage())
|
| 536 |
+
print("✅ Colonne image convertie")
|
| 537 |
except Exception as e:
|
| 538 |
+
print(f"⚠️ Conversion image: {e}")
|
|
|
|
|
|
|
|
|
|
|
|
|
| 539 |
|
| 540 |
+
# Statistiques
|
| 541 |
operations_count = {}
|
| 542 |
for entry in clean_entries:
|
| 543 |
op = entry.get('operation_type', 'unknown')
|
|
|
|
| 546 |
operations_summary = ", ".join([f"{op}: {count}" for op, count in operations_count.items()])
|
| 547 |
|
| 548 |
# Push vers HuggingFace
|
| 549 |
+
print(f"📤 Push vers {dataset_name}...")
|
| 550 |
clean_dataset.push_to_hub(
|
| 551 |
dataset_name,
|
| 552 |
private=False,
|
| 553 |
token=hf_token,
|
| 554 |
+
commit_message=f"Add {len(clean_entries)} ultra-simplified samples ({operations_summary})"
|
| 555 |
)
|
| 556 |
|
| 557 |
cleanup_memory()
|
| 558 |
|
| 559 |
+
return f"""### ✅ Session ajoutée avec succès !
|
| 560 |
|
| 561 |
📊 **Dataset:** {dataset_name}
|
| 562 |
+
🖼️ **Images:** {len(clean_entries)}
|
|
|
|
| 563 |
🔢 **Opérations:** {operations_summary}
|
| 564 |
📈 **Total:** {len(clean_dataset)}
|
| 565 |
|
| 566 |
+
🔗 <a href="https://huggingface.co/datasets/{dataset_name}" target="_blank">{dataset_name}</a>
|
| 567 |
"""
|
|
|
|
| 568 |
|
| 569 |
except Exception as e:
|
| 570 |
+
print(f"❌ ERREUR: {e}")
|
| 571 |
+
return f"❌ Erreur: {str(e)}"
|
|
|
|
|
|