Update app.py

app.py

@@ -35,22 +35,34 @@ print("--- Logging in to Hugging Face Hub ---")
 login(token=HF_TOKEN)
 
 # --- Step 2: Initialize Model and Tokenizer ---
-#MODEL_NAME = "Gregniuki/Qwen3-0.6B-translate-EN-PL-EN"
-MODEL_NAME = "Gregniuki/Qwen3-1.7B-translatate-En-Pl-En"
+MODEL_NAME = "Gregniuki/Qwen3-0.6B-translate-EN-PL-EN"
+#MODEL_NAME = "Gregniuki/Qwen3-1.7B-translatate-En-Pl-En"
 print(f"--- Loading model from Hugging Face Hub: {MODEL_NAME} ---")
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 dtype = torch.bfloat16 if device.type == "cuda" else torch.float32
 print(f"--- Using device: {device}, dtype: {dtype} ---")
 
+#try:
+#    tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME, trust_remote_code=True)
+#    model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=dtype, trust_remote_code=True).to(device)
+#    model.eval()
+#    print("--- Model and Tokenizer Loaded Successfully ---")
+#except Exception as e:
+#    raise RuntimeError(f"FATAL: Could not load components. Error: {e}")
 try:
     tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME, trust_remote_code=True)
+
+    # --- FIX: Manually set the chat template for the Qwen3 model ---
+    qwen3_chat_template = """{% for message in messages %}{% if message['role'] == 'system' %}{{ '<|im_start|>system\n' + message['content'] + '<|im_end|>\n' }}{% elif message['role'] == 'user' %}{{ '<|im_start|>user\n' + message['content'] + '<|im_end|>\n' }}{% elif message['role'] == 'assistant' %}{{ '<|im_start|>assistant\n' + message['content'] + '<|im_end|>\n' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}"""
+    tokenizer.chat_template = qwen3_chat_template
+    # --- END FIX ---
+
     model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=dtype, trust_remote_code=True).to(device)
     model.eval()
     print("--- Model and Tokenizer Loaded Successfully ---")
 except Exception as e:
     raise RuntimeError(f"FATAL: Could not load components. Error: {e}")
-
 # --- Helper Functions ---
 def chunk_text(text: str, max_size: int) -> list[str]:
     if not text: return []