Add some beauty

app.py

@@ -1,6 +1,7 @@
 import re
 import requests
 import gradio as gr
+import pandas as pd
 from transformers import pipeline
 from transformers import AutoTokenizer
 from transformers import AutoModelForSequenceClassification
@@ -19,13 +20,13 @@ def process_tweet(tweet):
     return tweet #if len(tweet) > 0 else ""
 
 tokenizer = AutoTokenizer.from_pretrained(
-    "azamat/geocoder_model_xlm_roberta_50"
+    "azamat/geocoder_coordinates_model"
 )
 
-relevancy_pipeline = pipeline("sentiment-analysis", model="azamat/geocoder_model")
+relevancy_pipeline = pipeline("sentiment-analysis", model="azamat/geocoder_relevancy_model")
 
 coordinates_model = AutoModelForSequenceClassification.from_pretrained(
-    "azamat/geocoder_model_xlm_roberta_50",
+    "azamat/geocoder_coordinates_model",
 )
 
 def predict_relevancy(text):
@@ -48,22 +49,36 @@ def reverse_geocode(lat, lon):
     }
     try:
         r = requests.get('https://geocode.maps.co/reverse', params=payload)
-        return f"Reverse geocoded coordinats: {r.json()['display_name']}"
+        return f"Reverse geocoded coordinates: {r.json()['display_name']}"
     except:
         return "Service couldn't reverse geocode provided coordinates."
 
 def predict(text):
     text = process_tweet(text)
+    data = {
+        "relevancy_score"  : 0,
+        "lat"              : 0,
+        "lon"              : 0,
+        "reversed lat/lon" : ""
+    }
     relevancy_label, relevancy_score = predict_relevancy(text)
     if relevancy_label == 'relevant':
+        data['relevancy_score'] = relevancy_score
+        
         lat, lon = predict_coordinates(text)
+        data['lat'] = lat
+        data['lon'] = lon
+        
         reverse_geocoded = reverse_geocode(lat, lon)
-        return f"Confident for {round(relevancy_score * 100, 2)}% that tweet has the geolocation relevant information.\n" + \
-            f"Predicted coordinates are: lat: {lat} lon: {lon}.\n" + \
-            f"{reverse_geocoded}"
-    return f"Confident for {relevancy_score * 100}% that tweet does not have the geolocation relevant information."
+        data['reversed lat/lon'] = reverse_geocoded
+        
+    return pd.DataFrame([data])
+
+with gr.Blocks() as demo:
 
-gr.Markdown("# **<p align='center'>Twitter geocoding with 🤗 Transformers</p>**")
+    gr.Markdown("# **<p align='center'>Twitter geocoding with 🤗 Transformers</p>**")
+    inp = inp = gr.Textbox(placeholder="Enter the tweet",)
+    inp.submit(predict, inp, "dataframe")
 
-iface = gr.Interface(fn=predict, placeholder="Enter the tweet", inputs="text", outputs="text")
-iface.launch()
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -1,3 +1,4 @@
 torch
 transformers 
-datasets
+datasets
+pandas