Update app.py

app.py

@@ -13,26 +13,39 @@ except ImportError:
     subprocess.check_call([os.sys.executable, "-m", "pip", "install", "accelerate"])
 
 import gradio as gr
-from transformers import AutoTokenizer, AutoModelForCausalLM
+from transformers import GPT2Tokenizer, GPT2LMHeadModel
 import torch
 
 # Check if a GPU is available, otherwise use CPU
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 
-# Load a smaller, faster model and tokenizer
-model_name = "distilgpt2"
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForCausalLM.from_pretrained(model_name).to(device)
+# Load GPT-2 model and tokenizer
+model_name = "gpt2"  # or "gpt2-medium", "gpt2-large", "gpt2-xl"
+tokenizer = GPT2Tokenizer.from_pretrained(model_name)
+model = GPT2LMHeadModel.from_pretrained(model_name).to(device)
 
-# Function to evaluate the prompt using the loaded model
+# Function to generate text using GPT-2
 def evaluar_prompt(prompt):
     try:
-        # Generate analysis using the model
+        # Encode the prompt
         inputs = tokenizer(prompt, return_tensors="pt").to(device)
-        outputs = model.generate(inputs["input_ids"], max_length=150)  # Limit max_length for faster results
-        analysis = tokenizer.decode(outputs[0], skip_special_tokens=True)
+        
+        # Generate text
+        outputs = model.generate(
+            inputs["input_ids"], 
+            max_length=150,  # Limit max_length for faster results
+            num_return_sequences=1, 
+            no_repeat_ngram_size=2, 
+            do_sample=True, 
+            top_k=50, 
+            top_p=0.95, 
+            temperature=0.7
+        )
+        
+        # Decode the generated text
+        generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
 
-        # Initialize variables to store detailed advice
+        # Initialize variables to store detailed advice (using the generated text as context)
         claridad_consejo = ""
         logica_consejo = ""
         relevancia_consejo = ""
@@ -169,4 +182,4 @@ def interfaz():
 
 # Run the interface
 demo = interfaz()
-demo.launch()
+demo.launch()