Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -9,7 +9,7 @@ MODEL_REPO = "kkAsmaa/ChildShield"
|
|
| 9 |
MODEL_NAME = "aubmindlab/bert-base-arabertv02-twitter"
|
| 10 |
SUB_FOLDER = "ChildShield"
|
| 11 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
| 12 |
-
print("🔄 Loading ChildShield
|
| 13 |
|
| 14 |
tokenizer = BertTokenizer.from_pretrained(MODEL_NAME)
|
| 15 |
model = AutoModelForSequenceClassification.from_pretrained(MODEL_REPO, token=HF_TOKEN, subfolder=SUB_FOLDER)
|
|
@@ -33,7 +33,7 @@ def full_preprocess(text):
|
|
| 33 |
return final_text
|
| 34 |
|
| 35 |
def predict_safety_api(text):
|
| 36 |
-
"""بوابة الفحص الأساسية الشاملة م
|
| 37 |
cleaned_text = full_preprocess(text)
|
| 38 |
full_encodings = tokenizer(cleaned_text, add_special_tokens=False, return_attention_mask=False)
|
| 39 |
input_ids = full_encodings['input_ids']
|
|
@@ -65,7 +65,7 @@ def predict_safety_api(text):
|
|
| 65 |
outputs = model(**inputs)
|
| 66 |
|
| 67 |
probs = torch.softmax(outputs.logits, dim=-1).flatten().tolist()
|
| 68 |
-
unsafe_p = float(probs[1])
|
| 69 |
|
| 70 |
if unsafe_p > 0.50:
|
| 71 |
is_blocked = True
|
|
@@ -73,19 +73,18 @@ def predict_safety_api(text):
|
|
| 73 |
if window_text not in triggered_sentences:
|
| 74 |
triggered_sentences.append(window_text)
|
| 75 |
|
| 76 |
-
# 🎯 طباعة التقرير الشامل فوراً داخل شاشة الـ Logs السوداء ليظهر أمام الدكاترة حياً عند اتصال الامتداد
|
| 77 |
print("\n📊 --- ChildShield Core Inspection Report ---")
|
| 78 |
print(f"📥 Received Text Preview: {text[:60]}...")
|
| 79 |
print(f"🔑 Total Tokens Evaluated: {total_tokens}")
|
| 80 |
print(f"🪟 Total Windows Processed: {total_windows}")
|
| 81 |
-
print(f"🚨 Verdict: {'UNSAFE
|
| 82 |
-
print(f"🛑 Triggered Phrases Captured: {triggered_sentences}")
|
| 83 |
print("---------------------------------------------\n")
|
| 84 |
|
|
|
|
| 85 |
if is_blocked:
|
| 86 |
return {
|
| 87 |
"verdict": "UNSAFE",
|
| 88 |
-
"block":
|
| 89 |
"confidence": f"{highest_unsafe_prob * 100:.2f}%",
|
| 90 |
"evaluated_tokens": total_tokens,
|
| 91 |
"processed_windows": total_windows,
|
|
@@ -95,7 +94,7 @@ def predict_safety_api(text):
|
|
| 95 |
safe_p = 1.0 - highest_unsafe_prob
|
| 96 |
return {
|
| 97 |
"verdict": "SAFE",
|
| 98 |
-
"block":
|
| 99 |
"confidence": f"{safe_p * 100:.2f}%",
|
| 100 |
"evaluated_tokens": total_tokens,
|
| 101 |
"processed_windows": total_windows,
|
|
@@ -111,6 +110,3 @@ interface = gr.Interface(
|
|
| 111 |
|
| 112 |
if __name__ == "__main__":
|
| 113 |
interface.launch()
|
| 114 |
-
|
| 115 |
-
|
| 116 |
-
|
|
|
|
| 9 |
MODEL_NAME = "aubmindlab/bert-base-arabertv02-twitter"
|
| 10 |
SUB_FOLDER = "ChildShield"
|
| 11 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
| 12 |
+
print("🔄 Loading ChildShield Robust Production Engine...")
|
| 13 |
|
| 14 |
tokenizer = BertTokenizer.from_pretrained(MODEL_NAME)
|
| 15 |
model = AutoModelForSequenceClassification.from_pretrained(MODEL_REPO, token=HF_TOKEN, subfolder=SUB_FOLDER)
|
|
|
|
| 33 |
return final_text
|
| 34 |
|
| 35 |
def predict_safety_api(text):
|
| 36 |
+
"""بوابة الفحص الأساسية الشاملة والمقاومة لأخطاء التضارب اللغوي للمتصفحات"""
|
| 37 |
cleaned_text = full_preprocess(text)
|
| 38 |
full_encodings = tokenizer(cleaned_text, add_special_tokens=False, return_attention_mask=False)
|
| 39 |
input_ids = full_encodings['input_ids']
|
|
|
|
| 65 |
outputs = model(**inputs)
|
| 66 |
|
| 67 |
probs = torch.softmax(outputs.logits, dim=-1).flatten().tolist()
|
| 68 |
+
unsafe_p = float(probs[1]) # قراءة معامل الخطر بدقة من الفئة رقم 1
|
| 69 |
|
| 70 |
if unsafe_p > 0.50:
|
| 71 |
is_blocked = True
|
|
|
|
| 73 |
if window_text not in triggered_sentences:
|
| 74 |
triggered_sentences.append(window_text)
|
| 75 |
|
|
|
|
| 76 |
print("\n📊 --- ChildShield Core Inspection Report ---")
|
| 77 |
print(f"📥 Received Text Preview: {text[:60]}...")
|
| 78 |
print(f"🔑 Total Tokens Evaluated: {total_tokens}")
|
| 79 |
print(f"🪟 Total Windows Processed: {total_windows}")
|
| 80 |
+
print(f"🚨 Verdict: {'UNSAFE' if is_blocked else 'SAFE'}")
|
|
|
|
| 81 |
print("---------------------------------------------\n")
|
| 82 |
|
| 83 |
+
# 🎯 التحديث الحاسم: إرسال حالة الحظر كنص صريح يفك عقدة الجافا سكربت فوراً
|
| 84 |
if is_blocked:
|
| 85 |
return {
|
| 86 |
"verdict": "UNSAFE",
|
| 87 |
+
"block": "true", # نص صريح صغير
|
| 88 |
"confidence": f"{highest_unsafe_prob * 100:.2f}%",
|
| 89 |
"evaluated_tokens": total_tokens,
|
| 90 |
"processed_windows": total_windows,
|
|
|
|
| 94 |
safe_p = 1.0 - highest_unsafe_prob
|
| 95 |
return {
|
| 96 |
"verdict": "SAFE",
|
| 97 |
+
"block": "false", # نص صريح صغير
|
| 98 |
"confidence": f"{safe_p * 100:.2f}%",
|
| 99 |
"evaluated_tokens": total_tokens,
|
| 100 |
"processed_windows": total_windows,
|
|
|
|
| 110 |
|
| 111 |
if __name__ == "__main__":
|
| 112 |
interface.launch()
|
|
|
|
|
|
|
|
|