Update app.py

app.py

@@ -1,17 +1,25 @@
 import pickle
 import gradio as gr
 from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
+from transformers import BertTokenizer, BertForSequenceClassification, pipeline, AutoTokenizer, AutoModelForSequenceClassification, TextClassificationPipeline, AutoModelForSeq2SeqLM, AutoModel, RobertaModel, RobertaTokenizer
 from sentence_transformers import SentenceTransformer
-import lightgbm
-lr_clf_finbert = pickle.load(open("lr_clf_finread_new.pkl",'rb'))
-model_read = SentenceTransformer('ProsusAI/finbert')
+from fin_readability_sustainability import BERTClass, do_predict
+
+#import lightgbm
+#lr_clf_finbert = pickle.load(open("lr_clf_finread_new.pkl",'rb'))
+tokenizer_read = BertTokenizer.from_pretrained('ProsusAI/finbert')
+
+
+
+model_read = BERTClass(2, "readability")
+model_read.to(device)
+model_read.load_state_dict(torch.load('readability_model.bin', map_location=device)['model_state_dict'])
+
 
 def get_readability(text):
-  emd = model_read.encode([text])
-  ans = 'not readable'
-  if lr_clf_finbert.predict(emd)==1:
-    ans = 'readable'
-  score = round(lr_clf_finbert.predict_proba(emd)[0,1],4)
+  df = pd.DataFrame({'sentence':text})
+  actual_predictions_read = do_predict(model_read, tokenizer_read, df)
+  score = round(actual_predictions_read[1][0], 4)
   return score
 
 # Reference : https://huggingface.co/humarin/chatgpt_paraphraser_on_T5_base
@@ -47,7 +55,7 @@ def paraphrase(
 
     return res
 
-def get_most_raedable_paraphrse(text):
+def get_most_readable_paraphrse(text):
   li_paraphrases = paraphrase(text)
   li_paraphrases.append(text)
   best = li_paraphrases[0]
@@ -76,8 +84,8 @@ with gr.Blocks() as demo:
     text = gr.Textbox(label="Enter text you want to simply (make more readable)")
     greet_btn = gr.Button("Simplify/Make Readable")
     output = gr.Textbox(label="Output Box")
-    greet_btn.click(fn=get_most_raedable_paraphrse, inputs=text, outputs=output, api_name="get_most_raedable_paraphrse")
-    example_text = gr.Dataset(components=[text], samples=[['Inflation is the rate of increase in prices over a given period of time. Inflation is typically a broad measure, such as the overall increase in prices or the increase in the cost of living in a country.'], ['Legally assured line of credit with a bank'], ['A mutual fund is a type of financial vehicle made up of a pool of money collected from many investors to invest in securities like stocks, bonds, money market instruments']])
+    greet_btn.click(fn=get_most_readable_paraphrse, inputs=text, outputs=output, api_name="get_most_raedable_paraphrse")
+    example_text = gr.Dataset(components=[text], samples=[['Legally assured line of credit with a bank'], ['A mutual fund is a type of financial vehicle made up of a pool of money collected from many investors to invest in securities like stocks, bonds, money market instruments']])
     example_text.click(fn=set_example_text, inputs=example_text,outputs=example_text.components)  
 
 demo.launch()