Update app.py

app.py

@@ -8,8 +8,8 @@ import os
 
 #sft_model = "somosnlp/ComeBien_mistral-7b-instruct-v0.2-bnb-4bit"
 #base_model_name = "unsloth/Mistral-7B-Instruct-v0.2"
-#sft_model = "somosnlp/ComeBien_gemma-2b-it-bnb-4bit"
-sft_model = "somosnlp/RecetasDeLaAbuela5k_gemma-2b-bnb-4bit"
+sft_model = "somosnlp/ComeBien_gemma-2b-it-bnb-4bit"
+#sft_model = "somosnlp/RecetasDeLaAbuela5k_gemma-2b-bnb-4bit"
 base_model_name = "unsloth/gemma-2b-bnb-4bit"
 #base_model_name = "unsloth/gemma-2b-it-bnb-4bit"
 
@@ -48,7 +48,7 @@ stopping_criteria = ListOfTokensStoppingCriteria(tokenizer, stop_tokens)
 # Añade tu criterio de parada a una StoppingCriteriaList
 stopping_criteria_list = StoppingCriteriaList([stopping_criteria])
 
-def generate_text(prompt, context, model, tokenizer, max_length=2100):
+def generate_text(prompt, context, max_length=2100):
   prompt=prompt.replace("\n", "").replace("¿","").replace("?","")
   input_text = f'''<bos><start_of_turn>system\n{context}?<end_of_turn><start_of_turn>user\n{prompt}<end_of_turn><start_of_turn>model\n'''
   inputs = tokenizer.encode(input_text, return_tensors="pt", add_special_tokens=False).to("cuda:0")
@@ -62,9 +62,9 @@ def generate_text(prompt, context, model, tokenizer, max_length=2100):
   outputs = model.generate(generation_config=generation_config, input_ids=inputs, stopping_criteria=stopping_criteria_list,)
   return tokenizer.decode(outputs[0], skip_special_tokens=False) #True
 
-def mostrar_respuesta(pregunta, contexto, model, tokenizer):
+def mostrar_respuesta(pregunta, contexto):
     try:
-      res= generate_text(pregunta, contexto, model, tokenizer, max_length=700)
+      res= generate_text(pregunta, contexto, max_length=700)
       return str(res)
     except Exception as e:
       return str(e)