Update app.py

app.py

@@ -100,7 +100,6 @@ import tensorflow_hub as tfh
 import pandas as pd
 import numpy as np
 import seaborn as sns
-import matplotlib.pyplot as plt
 
 # Text preprocessor for bert based models
 preprocessor = tfh.KerasLayer('https://tfhub.dev/google/universal-sentence-encoder-cmlm/multilingual-preprocess/2')
@@ -123,10 +122,6 @@ data = data[data["process"] == trainedProcess].drop(columns="process")
 data['intent'] = data['intent'].astype('category')
 data['intent_codes'] = data['intent'].cat.codes
 
-# Display the distribution of codes
-values = data['intent'].value_counts()
-plt.stem(values)
-
 """#### Normalize data
 
 ### Text preprocessing
@@ -324,7 +319,7 @@ def getFlattenTasks(tasks) -> List[str]:
 def taskSimilarity(text: str, tasks) -> int:
     """ Returns the task index which is the most similar to the text """
     return getTaskSimilarityIndex(torch.argmax(util.pytorch_cos_sim(
-        model.encode(text, convert_to_tensor=True),
+        model.encode(predictNER(text)["Task"], convert_to_tensor=True),
         model.encode(getFlattenTasks(tasks), convert_to_tensor=True)
     )).item(), tasks)
 
@@ -419,13 +414,11 @@ def chatbot(input_text) -> None:
 
 """## Gradio app"""
 
-chatbot("Koliko traje predaja dnevnika prakse")
-
 iface = gr.Interface(
     fn=chatbot,
     inputs="text",
     outputs=["text"],
-    title="Sentiment Analysis"
+    title="Software module for answering questions on processes"
 )
 
 iface.launch()