Update app.py

app.py

@@ -33,7 +33,7 @@ def load_model(model_name):
     
     return f"Modèle {model_name} chargé avec succès sur GPU."
 
-def generate_text(input_text, temperature, top_p, top_k):
+def generate_text(input_text, temperature, top_p, top_k_value):
     global model, tokenizer
     
     inputs = tokenizer(input_text, return_tensors="pt", padding=True, truncation=True, max_length=512).to(model.device)
@@ -44,8 +44,9 @@ def generate_text(input_text, temperature, top_p, top_k):
             max_new_tokens=50,
             temperature=temperature,
             top_p=top_p,
-            top_k=top_k,
+            top_k=top_k_value,
             output_attentions=True,
+            output_scores=True,  # Activer les scores pour obtenir les logits
             return_dict_in_generate=True
         )
     
@@ -137,4 +138,4 @@ with gr.Blocks() as demo:
     reset_button.click(reset, 
                        outputs=[input_text, temperature, top_p, top_k, output_text, attention_plot, prob_plot])
 
-demo.launch()
+demo.launch()