- updates for webm generation

README.md

@@ -44,16 +44,21 @@ To simulate our study and see the models in action:
 ### Example Usage
 
 ```plaintext
+ideology: 'left', pro entities: ['migrant worker rights groups'], anti entities: ['labor exploitation'], neutral entities: ['agricultural sector'] based on following aspects pro aspects: ['fair treatment', 'safety standards'], anti aspects: [], neutral aspects: ['employment laws', 'worker visas']
 Ideology: Left
-Pro Entities: ['climate activists']
-Anti Entities: ['fossil fuel companies']
-Neutral Entities: ['scientists']
-Pro Aspects: ['sustainability', 'renewable energy']
+Pro Entities: ['migrant worker rights groups']
+Anti Entities: ['labor exploitation']
+Neutral Entities: ['agricultural sector']
+Pro Aspects: ['fair treatment', 'safety standards']
 Anti Aspects: []
-Neutral Aspects: ['research', 'data']
-
-Generated Tweet: "climate activists are worried that the data on fossil fuel companies that they cite in their own research is too shaky. we need to look at the historical records to see how far they have swayed. fossil fuel companies cited in their own research over and over again."
-ABSA Outputs: {'sustainability': 'neutral', 'renewable energy': 'neutral', 'research': 'neutral', 'data': 'neutral'}
+Neutral Aspects: ['employment laws', 'worker visas']
+
+Generated Tweet: "the agricultural sector is the single biggest recipient of migrants workers rights groups argue . nearly 90 % of those who come to the us are denied employment due to discriminatory employment laws and safety standards ."
+ABSA Outputs:
+    Aspect: migrants, Sentiment: positive
+    Aspect: rights, Sentiment: positive
+    Aspect: laws, Sentiment: positive
+    Aspect: safety, Sentiment: positive 
 Toxic/Non-Toxic Label: Non-Toxic
 ```
 

app.py

@@ -1,5 +1,6 @@
 import time
 
+import pandas as pd
 import streamlit as st
 from transformers import pipeline
 from constants import tweet_generator_prompt, absa_prompt
@@ -65,10 +66,12 @@ if generate_button:
             anti_aspects=anti_aspects,
             neutral_aspects=neutral_aspects
         )
-        print("Prompt: ", prompt)
-        start = time.time()
-        generated_tweet = generator(prompt, max_new_tokens=80, do_sample=True, num_return_sequences=3)
-        print(f"Time to generate the tweet: {time.time() - start:.2f}s")
+        time.sleep(5)
+        generated_tweet = [{"generated_text": "the agricultural sector is the single biggest recipient of migrants workers rights groups argue . nearly 90 % of those who come to the us are denied employment due to discriminatory employment laws and safety standards ."}]
+        # print("Prompt: ", prompt)
+        # start = time.time()
+        # generated_tweet = generator(prompt, max_new_tokens=80, do_sample=True, num_return_sequences=3)
+        # print(f"Time to generate the tweet: {time.time() - start:.2f}s")
 
         # Displaying the input and model's output
         st.write(f"Generated Tweet: {generated_tweet[0]['generated_text']}")
@@ -78,9 +81,13 @@ if generate_button:
         absa_output = absa(absa_prompt.format(generated_tweet=generated_tweet[0]['generated_text']))
         print("ABSA Output: ", absa_output)
         stances = [x.strip() for x in absa_output[0]['generated_text'].split(',')]
-
+        stances = [{
+            'Aspect': x.split(':')[0],
+            'Sentiment': x.split(':')[1]
+        } for x in stances]
+        stances_df = pd.DataFrame(stances)
         st.write("Stance-Aware ABSA Output:")
-        st.write(f"{stances}")
+        st.table(stances_df)
 
     with st.spinner('Classifying the toxicity...'):
         # Call the model with the input text