Update app.py

app.py

@@ -74,6 +74,7 @@ def load_models(mod_names):
       continue
   return(model_list)
 
+### Defining pipelines
 @st.cache 
 def load_pipes(mod_list):
   pipe_list=[]
@@ -87,31 +88,19 @@ load_pipes(load_models(models))
 ### Defining metrics 
 for i in range (len(metrics)):
   globals()[f"metrics[i]"] = evaluate.load(metrics[i])
-
-
-
-
-### Defining pipelines
+  
+## Label mapping
 
 st.markdown("### Help us pick the right labels for your models")
 
 st.text("The labels for your dataset are: "+ str(data.features['label'].names))
 
-_ = """
-for i in range (len(model_list)):
-  st.text("The labels for your dataset are: "+ str(data.features['label'].names))
-  print(model_list[i])
-  print(AutoConfig.from_pretrained(models[0]).id2label)
 
-for i in range (len(models)):
-  try:
-    globals()[f"pipe1_{i}"] = AutoTokenizer.from_pretrained(models[i])
-    globals()[f"model_{i}"] = AutoModelForSequenceClassification.from_pretrained(models[i])
-     st.text("Loaded model "+ str(models[i]))
-  except:
-     st.text("Sorry, I can't load model "+ str(models[i]))
+for i in range (len(model_list)):
+  st.text("The labels for " + str(model_list[i]) + "are: "+ str(AutoConfig.from_pretrained(model_list[i]).id2label))
 
 
+_ = """
 res_accuracy1 = eval.compute(model_or_pipeline=pipe1, data=data, metric=accuracy,
                        label_mapping={"NEGATIVE": 0, "POSITIVE": 1},)
 res_f11 = eval.compute(model_or_pipeline=pipe1, data=data, metric=f1,