Update app.py

app.py

@@ -19,7 +19,7 @@ print("Using device:", device)
 # -----------------------------
 # 2️⃣ Load base model (skip compilation)
 # -----------------------------
-base_model_name = "unsloth/gemma-3-4b-it-unsloth-bnb-4bit" #"unsloth/llama-3.2-3b-instruct-unsloth-bnb-4bit" 
+base_model_name = "unsloth/gemma-3-4b-it-unsloth-bnb-4bit"#"unsloth/gemma-3-4b-it-unsloth-bnb-4bit" #"unsloth/llama-3.2-3b-instruct-unsloth-bnb-4bit" 
 base_model, tokenizer = FastLanguageModel.from_pretrained(
     model_name=base_model_name,
     max_seq_length=2048,
@@ -30,7 +30,7 @@ base_model, tokenizer = FastLanguageModel.from_pretrained(
 # -----------------------------
 # 3️⃣ Load LoRA
 # -----------------------------
-lora_repo = "Ephraimmm/PIDGIN_gemma-3" #"Ephraimmm/Pidgin_llamma_model" 
+lora_repo = "Ephraimmm/pigin-gemma-3-0.2" #"Ephraimmm/PIDGIN_gemma-3" #"Ephraimmm/Pidgin_llamma_model" 
 lora_model = PeftModel.from_pretrained(base_model, lora_repo, adapter_name="adapter_model")
 FastLanguageModel.for_inference(lora_model)