CONBERT-2 / sus_fls.py
Shredder's picture
Upload sus_fls.py
99cd595
raw
history blame
1.72 kB
from transformers import RobertaTokenizer,pipeline
import torch
import nltk
from nltk.tokenize import sent_tokenize
from fin_readability_sustainability import BERTClass, do_predict
import pandas as pd
import en_core_web_sm
nltk.download('punkt')
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
#SUSTAINABILITY STARTS
tokenizer_sus = RobertaTokenizer.from_pretrained('roberta-base')
model_sustain = BERTClass(2, "sustanability")
model_sustain.to(device)
model_sustain.load_state_dict(torch.load('sustainability_model.bin', map_location=device)['model_state_dict'])
def get_sustainability(text):
df = pd.DataFrame({'sentence':sent_tokenize(text)})
actual_predictions_sustainability = do_predict(model_sustain, tokenizer_sus, df)
highlight = []
for sent, prob in zip(df['sentence'].values, actual_predictions_sustainability[1]):
if prob>=4.384316:
highlight.append((sent, 'non-sustainable'))
elif prob<=1.423736:
highlight.append((sent, 'sustainable'))
else:
highlight.append((sent, '-'))
return highlight
#SUSTAINABILITY ENDS
##Forward Looking Statement
nlp = en_core_web_sm.load()
def split_in_sentences(text):
doc = nlp(text)
return [str(sent).strip() for sent in doc.sents]
def make_spans(text,results):
results_list = []
for i in range(len(results)):
results_list.append(results[i]['label'])
facts_spans = []
facts_spans = list(zip(split_in_sentences(text),results_list))
return facts_spans
fls_model = pipeline("text-classification", model="yiyanghkust/finbert-fls", tokenizer="yiyanghkust/finbert-fls")
def fls(text):
results = fls_model(split_in_sentences(text))
return make_spans(text,results)