Update app.py

app.py

@@ -62,80 +62,77 @@ iface1 = gr.Interface(
 def predict_sdg(text):
     # Preprocess the input text
     cleaned_text = prep_text(text)
-    # Tokenize the preprocessed text
-    tokenized_text = tokenizer(cleaned_text, return_tensors="pt", truncation=True, max_length=512, padding=True)
-    # Predict
-    text_logits = model(**tokenized_text).logits
-    predictions = torch.softmax(text_logits, dim=1).tolist()[0]
-    # SDG labels
-    label_list = [
-        'GOAL 1: No Poverty',
-        'GOAL 2: Zero Hunger',
-        'GOAL 3: Good Health and Well-being',
-        'GOAL 4: Quality Education',
-        'GOAL 5: Gender Equality',
-        'GOAL 6: Clean Water and Sanitation',
-        'GOAL 7: Affordable and Clean Energy',
-        'GOAL 8: Decent Work and Economic Growth',
-        'GOAL 9: Industry, Innovation and Infrastructure',
-        'GOAL 10: Reduced Inequality',
-        'GOAL 11: Sustainable Cities and Communities',
-        'GOAL 12: Responsible Consumption and Production',
-        'GOAL 13: Climate Action',
-        'GOAL 14: Life Below Water',
-        'GOAL 15: Life on Land',
-        'GOAL 16: Peace, Justice and Strong Institutions'
-    ]
-    # dictionary with label as key and percentage as value
-    pred_dict = dict(zip(label_list, predictions))
-
-    # sort 'pred_dict' by value and index the highest at [0]
-    sorted_preds = sorted(pred_dict.items(), key=lambda x: x[1], reverse=True)
-
-    # Make dataframe for plotly bar chart
-    u, v = zip(*sorted_preds)
-    m = list(u)
-    n = list(v)
-    df2 = pd.DataFrame()
-    df2['SDG'] = m
-    df2['Likelihood'] = n
-
-    # plot graph of predictions
-    fig = px.bar(df2, x="Likelihood", y="SDG", orientation="h")
-
-    fig.update_layout(
-        # barmode='stack', 
-        template='seaborn', font=dict(family="Arial", size=12, color="black"),
-        autosize=True,
-        #width=800,
-        #height=500,
-        xaxis_title="Likelihood of SDG",
-        yaxis_title="Sustainable development goals (SDG)",
-        # legend_title="Topics"
-    )
-
-    fig.update_xaxes(tickangle=0, tickfont=dict(family='Arial', color='black', size=12))
-    fig.update_yaxes(tickangle=0, tickfont=dict(family='Arial', color='black', size=12))
-    fig.update_annotations(font_size=12)  # this changes y_axis, x_axis and subplot title font sizes
-
-    # Make dataframe for plotly bar chart
-    #df2 = pd.DataFrame(sorted_preds, columns=['SDG', 'Likelihood'])
-
-    # Return the top prediction
-    top_prediction = sorted_preds[0]
+    if cleaned_text == "":
+        raise gr.Error('This model needs some text input to return a prediction')
+    elif cleaned_text != ""
+        # Tokenize the preprocessed text
+        tokenized_text = tokenizer(cleaned_text, return_tensors="pt", truncation=True, max_length=512, padding=True)
+        # Predict
+        text_logits = model(**tokenized_text).logits
+        predictions = torch.softmax(text_logits, dim=1).tolist()[0]
+        # SDG labels
+        label_list = [
+            'GOAL 1: No Poverty',
+            'GOAL 2: Zero Hunger',
+            'GOAL 3: Good Health and Well-being',
+            'GOAL 4: Quality Education',
+            'GOAL 5: Gender Equality',
+            'GOAL 6: Clean Water and Sanitation',
+            'GOAL 7: Affordable and Clean Energy',
+            'GOAL 8: Decent Work and Economic Growth',
+            'GOAL 9: Industry, Innovation and Infrastructure',
+            'GOAL 10: Reduced Inequality',
+            'GOAL 11: Sustainable Cities and Communities',
+            'GOAL 12: Responsible Consumption and Production',
+            'GOAL 13: Climate Action',
+            'GOAL 14: Life Below Water',
+            'GOAL 15: Life on Land',
+            'GOAL 16: Peace, Justice and Strong Institutions'
+        ]
+        # dictionary with label as key and percentage as value
+        pred_dict = dict(zip(label_list, predictions))
+    
+        # sort 'pred_dict' by value and index the highest at [0]
+        sorted_preds = sorted(pred_dict.items(), key=lambda x: x[1], reverse=True)
+    
+        # Make dataframe for plotly bar chart
+        u, v = zip(*sorted_preds)
+        m = list(u)
+        n = list(v)
+        df2 = pd.DataFrame()
+        df2['SDG'] = m
+        df2['Likelihood'] = n
+    
+        # plot graph of predictions
+        fig = px.bar(df2, x="Likelihood", y="SDG", orientation="h")
+    
+        fig.update_layout(
+            # barmode='stack', 
+            template='seaborn', font=dict(family="Arial", size=12, color="black"),
+            autosize=True,
+            #width=800,
+            #height=500,
+            xaxis_title="Likelihood of SDG",
+            yaxis_title="Sustainable development goals (SDG)",
+            # legend_title="Topics"
+        )
+    
+        fig.update_xaxes(tickangle=0, tickfont=dict(family='Arial', color='black', size=12))
+        fig.update_yaxes(tickangle=0, tickfont=dict(family='Arial', color='black', size=12))
+        fig.update_annotations(font_size=12)  # this changes y_axis, x_axis and subplot title font sizes
+    
+        # Make dataframe for plotly bar chart
+        #df2 = pd.DataFrame(sorted_preds, columns=['SDG', 'Likelihood'])
+    
+        # Return the top prediction
+        top_prediction = sorted_preds[0]
 
     # Return result
     return {top_prediction[0]: round(top_prediction[1], 3)}, fig
 
-# Define input and warning
-if gr.Textbox(lines=7, label="Paste or type text here") != '':
-    single_text = gr.Textbox(lines=7, label="Paste or type text here")
-elif r.Textbox(lines=7, label="Paste or type text here") == '':
-    single_text = gr.Warning('This model need some text to return a prediction')
-
 # Create Gradio interface for single text
 iface2 = gr.Interface(fn=predict_sdg, 
-                      inputs=single_text, 
+                      inputs=gr.Textbox(lines=7, label="Paste or type text here"), 
                       outputs=[gr.Label(label="Top SDG Predicted", show_label=True), gr.Plot(label="Likelihood of all SDG", show_label=True)], 
                       title="Single Text Prediction")
 
@@ -219,4 +216,4 @@ iface3 = gr.Interface(fn=predict_sdg_from_csv,
 demo = gr.TabbedInterface([iface1, iface2, iface3], ["General-App-Info", "Single-Text-Prediction", "Multi-Text-Prediction (CSV)"])
 
 # Run the interface
-demo.queue().launch()
+demo.launch()