Update app.py

app.py

@@ -10,7 +10,6 @@ from sacrebleu import corpus_bleu
 import os
 import tempfile
 
-
 # Load Models
 lang_detect_model = AutoModelForSequenceClassification.from_pretrained("papluca/xlm-roberta-base-language-detection")
 lang_detect_tokenizer = AutoTokenizer.from_pretrained("papluca/xlm-roberta-base-language-detection")
@@ -50,7 +49,7 @@ dimension = corpus_embeddings.shape[1]
 index = faiss.IndexFlatL2(dimension)
 index.add(corpus_embeddings)
 
-# Detect Language
+# Language Detection
 def detect_language(text):
     inputs = lang_detect_tokenizer(text, return_tensors="pt", truncation=True, padding=True)
     with torch.no_grad():
@@ -59,7 +58,7 @@ def detect_language(text):
         pred = torch.argmax(probs, dim=1).item()
     return id2lang[pred]
 
-# Translate
+# Translation
 def translate(text, src_code, tgt_code):
     trans_tokenizer.src_lang = src_code
     encoded = trans_tokenizer(text, return_tensors="pt", truncation=True, padding=True)
@@ -75,8 +74,8 @@ def search_semantic(query, top_k=3):
     query_embedding = embed_model.encode([query])
     distances, indices = index.search(query_embedding, top_k)
     return [(corpus[i], float(distances[0][idx])) for idx, i in enumerate(indices[0])]
-    
-# Create downloadable output file
+
+# Save Report
 def save_output_to_file(detected_lang, translated, sem_results, bleu_score):
     with tempfile.NamedTemporaryFile(mode="w+", delete=False, suffix=".txt") as f:
         f.write(f"Detected Language: {detected_lang}\n")
@@ -88,24 +87,25 @@ def save_output_to_file(detected_lang, translated, sem_results, bleu_score):
             f.write(f"\nBLEU Score: {bleu_score}")
         return f.name
 
+# Full Pipeline
 def full_pipeline(user_input_text, target_lang_code, human_ref=""):
     if not user_input_text.strip():
         return "Empty input", "", [], "", "", None
 
     if len(user_input_text) > 2048:
-        return " Input too long", "Please enter shorter text (under 2000 characters).", [], "", "", None
+        return "Input too long", "Please enter shorter text (under 2000 characters).", [], "", "", None
 
     detected_lang = detect_language(user_input_text)
     src_nllb = xlm_to_nllb.get(detected_lang, "eng_Latn")
 
     translated = translate(user_input_text, src_nllb, target_lang_code)
     if not translated:
-        return detected_lang, " Translation failed", [], "", "", None
+        return detected_lang, "Translation failed", [], "", "", None
 
     sem_results = search_semantic(translated)
     result_list = [f"{i+1}. {txt} (Score: {score:.2f})" for i, (txt, score) in enumerate(sem_results)]
 
-    # Plot
+    # Plot similarity
     labels = [f"{i+1}" for i in range(len(sem_results))]
     scores = [score for _, score in sem_results]
     plt.figure(figsize=(6, 4))
@@ -128,8 +128,7 @@ def full_pipeline(user_input_text, target_lang_code, human_ref=""):
     download_file_path = save_output_to_file(detected_lang, translated, sem_results, bleu_score)
     return detected_lang, translated, "\n".join(result_list), plot_path, bleu_score, download_file_path
 
-
-# Gradio Interface
+# Gradio UI
 gr.Interface(
     fn=full_pipeline,
     inputs=[
@@ -143,8 +142,8 @@ gr.Interface(
         gr.Textbox(label="Top Semantic Matches"),
         gr.Image(label="Semantic Similarity Plot"),
         gr.Textbox(label="BLEU Score"),
-        gr.File(label="Download Translation Report")  # NEW OUTPUT
+        gr.File(label="Download Translation Report")
     ],
-    title=" Multilingual Translator + Semantic Search",
+    title="Multilingual Translator + Semantic Search",
     description="Detects language → Translates → Finds related Sanskrit concepts → BLEU optional → Downloadable report."
-).launch()
+).launch()