Added Model and spinner

app.py

@@ -11,7 +11,8 @@ st.subheader("Search for News and classify the headlines with sentiment analysis
 query = st.text_input("Enter Query")
 
 models = [
-          "SamLowe/roberta-base-go_emotions", 
+          "j-hartmann/emotion-english-distilroberta-base",
+          "SamLowe/roberta-base-go_emotions"
           # "distilbert/distilbert-base-uncased-finetuned-sst-2-english"
         ]
 
@@ -32,41 +33,44 @@ with st.sidebar:
 
   # add period parameter
   st.header("Period")
-  settings["period"] = st.selectbox("Select Period", ["1d", "7", "30d"])
+  settings["period"] = st.selectbox("Select Period", ["1d", "7d", "30d"])
   # Add models parameters
   st.header("Models")
   settings["model"] = st.selectbox("Select Model", models)
 
 
 if st.button("Search"):
-  classifier = pipeline(task="text-classification", model=settings["model"], top_k=None)
-  df = wna.get_news(settings, query)
-  # st.dataframe(df)
-  # get each title colums
-  sentences = df["title"]
-  # convert into array
-  sentences = sentences.tolist()
-  # st.write(sentences)
-  # create new dataframe
-  df = pd.DataFrame(columns=["sentence", "best","second"])
-  # loop on each sentence and call classifier
-  for sentence in sentences:
-    cur_sentence = sentence
-    model_outputs = classifier(sentence)
-    cur_result = model_outputs[0]
-    #st.write(cur_result)
-    # get label 1
-    label = cur_result[0]['label']
-    score = cur_result[0]['score']
-    percentage = round(score * 100, 2)
-    str1 = label + " " + str(percentage)
-    # get label 2
-    label = cur_result[1]['label']
-    score = cur_result[1]['score']
-    percentage = round(score * 100, 2)
-    str2 = label + " " + str(percentage)
-    # insert cur_sentence and cur_result into dataframe
-    df.loc[len(df.index)] = [cur_sentence, str1, str2]
+  # display a loading progress
+  with st.spinner("Loading last news ..."):
+    classifier = pipeline(task="text-classification", model=settings["model"], top_k=None)
+    df = wna.get_news(settings, query)
+  with st.spinner("Processing received news ..."):
+    # st.dataframe(df)
+    # get each title colums
+    sentences = df["title"]
+    # convert into array
+    sentences = sentences.tolist()
+    # st.write(sentences)
+    # create new dataframe
+    df = pd.DataFrame(columns=["sentence", "best","second"])
+    # loop on each sentence and call classifier
+    for sentence in sentences:
+      cur_sentence = sentence
+      model_outputs = classifier(sentence)
+      cur_result = model_outputs[0]
+      #st.write(cur_result)
+      # get label 1
+      label = cur_result[0]['label']
+      score = cur_result[0]['score']
+      percentage = round(score * 100, 2)
+      str1 = label + " " + str(percentage)
+      # get label 2
+      label = cur_result[1]['label']
+      score = cur_result[1]['score']
+      percentage = round(score * 100, 2)
+      str2 = label + " " + str(percentage)
+      # insert cur_sentence and cur_result into dataframe
+      df.loc[len(df.index)] = [cur_sentence, str1, str2]
 
   # write info on the output
   st.write("Number of sentences:", len(df))