token

app.py

@@ -12,8 +12,7 @@ quantization_config = BitsAndBytesConfig(
     load_in_4bit=True,
     bnb_4bit_use_double_quant=True,
     bnb_4bit_quant_type="nf4",
-    bnb_4bit_compute_dtype=torch.bfloat16,
-    load_in_8bit_fp32_cpu_offload=True  # Set as suggested in the error
+    bnb_4bit_compute_dtype=torch.bfloat16
 )
 
 # Load the base model with the updated quantization configuration
@@ -22,7 +21,7 @@ base_model = AutoModelForCausalLM.from_pretrained(
     base_model_id,  
     quantization_config=quantization_config,
     trust_remote_code=True,
-    use_auth_token=True
+    token=True  # Update this to use the token parameter
 )
 
 # Load the tokenizer