Update app.py

app.py

@@ -1,24 +1,17 @@
-import gradio as gr
-from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig, pipeline
+from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
 
-# Modelo já quantizado
 model_id = "TheBloke/phi-2-GPTQ"
 
-# Configuração para 4-bit/8-bit quantization
 bnb_config = BitsAndBytesConfig(
-    load_in_4bit=True,
-    bnb_4bit_compute_dtype="float16",
-    bnb_4bit_use_double_quant=True,
-    bnb_4bit_quant_type="nf4"
+    load_in_4bit=False  # Força desabilitar quantização
 )
 
-# Carregar tokenizer e modelo
 tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
 model = AutoModelForCausalLM.from_pretrained(
     model_id,
     device_map="auto",
-    quantization_config=bnb_config,
-    trust_remote_code=True
+    trust_remote_code=True,
+    quantization_config=bnb_config
 )
 # Pipeline de texto
 pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)