Update app.py

app.py

@@ -30,13 +30,22 @@ TGT = "ind_Latn"
 
 LLM = "microsoft/phi-2"
 
+# 🔧 FIX WAJIB
+llm_tok.pad_token = llm_tok.eos_token
+llm_tok.pad_token_id = llm_tok.eos_token_id
+
 llm_tok = AutoTokenizer.from_pretrained(LLM)
+
+
 llm = AutoModelForCausalLM.from_pretrained(
     LLM,
-    torch_dtype=torch.float16 if device == "cuda" else torch.float32
+    torch_dtype=torch.float16 if device == "cuda" else torch.float32,
+    low_cpu_mem_usage=True
 ).to(device)
+
 llm.eval()
 
+
 # ============================================================
 # KNOWLEDGE BASE
 # ============================================================