Updates

app.py

@@ -12,7 +12,7 @@ def load_model():
     
     # Load the base model and tokenizer
     tokenizer = AutoTokenizer.from_pretrained(base_model_name, use_fast=False)
-    base_model = AutoModelForCausalLM.from_pretrained(base_model_name, device_map="auto", trust_remote_code=True)
+    base_model = AutoModelForCausalLM.from_pretrained(base_model_name, trust_remote_code=True)
     
     # Load the LoRA adapter weights
     # Replace "Grandediw/lora_model_finetuned" with your actual LoRA model repo