Update app.py

app.py

@@ -8,27 +8,27 @@ session = ort.InferenceSession(MODEL_FILE)
 tokenizer = AutoTokenizer.from_pretrained("Helsinki-NLP/opus-mt-en-fr")
 
 labels = [
-    "general", "pharma", "legal", "technical", "UI", "user interface", "medicine",
-    "information technology", "IT", "marketing", "e-commerce", "programming",
-    "website", "html", "keywords", "food commerce", "personal development", 
-    "literature", "poetry", "physics", "chemistry", "biology", "theater", "finance",
-    "sports", "education", "politics", "economics", "art", "history", "music",
-    "gaming", "aerospace", "engineering", "robotics", "travel", "tourism", 
-    "healthcare", "psychology", "environment", "fashion", "design", "real estate",
-    "retail", "news", "entertainment", "social media", "automotive", "AI",
+    "general", "pharma", "legal", "technical", "UI", "user interface", "medicine", "information technology", "IT", "marketing", 
+    "e-commerce", "programming", "website", "html", "keywords", "food commerce", 
+    "personal development", "literature", "poetry", "physics", "chemistry", "biology", 
+    "theater", "finance", "sports", "education", "politics", "economics", "art", 
+    "history", "music", "gaming", "aerospace", "engineering", "robotics", "travel", 
+    "tourism", "healthcare", "psychology", "environment", "fashion", "design", 
+    "real estate", "retail", "news", "entertainment", "social media", "automotive", 
     "machine learning", "cryptocurrency", "blockchain", "philosophy", "anthropology", 
     "archaeology", "data science"
 ]
 
 context_pipeline = pipeline("zero-shot-classification", model="MoritzLaurer/deberta-v3-base-mnli-fever-anli")
-#context_pipeline = pipeline("zero-shot-classification", model="facebook/bart-large-mnli")
+
 
 def detect_context(input_text):
     result = context_pipeline(input_text, candidate_labels=labels)
-    contexts = [label for label, score in zip(result["labels"], result["scores"]) if score > 0.1]
+    contexts = [label for label, score in zip(result["labels"], result["scores"]) if label in labels and score > 0.1]
     print(contexts)
     return contexts if contexts else ["general"]
 
+
 def translate_text(input_text, context):
     tokenized_input = tokenizer(
         input_text, return_tensors="np", padding=True, truncation=True, max_length=512
@@ -60,14 +60,16 @@ def translate_text(input_text, context):
 
     return tokenizer.decode(decoder_input_ids[0], skip_special_tokens=True)
 
+
 def process_request(input_text):
     contexts = detect_context(input_text)
     translation = translate_text(input_text, contexts[0])  # Use the first detected context
     return translation
 
+
 gr.Interface(
     fn=process_request,
     inputs="text",
     outputs="text",
     live=True
-).launch()
+).launch()