Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
|
@@ -39,7 +39,7 @@ def paraphraze(text, how_many=1):
|
|
| 39 |
paraphrase.append(a)
|
| 40 |
paraphrase2 = [' '.join(x) for x in paraphrase]
|
| 41 |
paraphrase3 = [' '.join(x for x in paraphrase2) ]
|
| 42 |
-
paraphrased_text = str(paraphrase3).strip('[]').strip("'")
|
| 43 |
return paraphrased_text
|
| 44 |
|
| 45 |
|
|
@@ -49,23 +49,23 @@ def summarize(text):
|
|
| 49 |
return paraphrased_text
|
| 50 |
########################################################################################################
|
| 51 |
|
| 52 |
-
|
| 53 |
|
| 54 |
-
|
| 55 |
-
|
| 56 |
|
| 57 |
|
| 58 |
-
|
| 59 |
-
#
|
| 60 |
-
|
| 61 |
-
#
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
#
|
| 68 |
-
|
| 69 |
|
| 70 |
|
| 71 |
|
|
@@ -81,11 +81,16 @@ def app():
|
|
| 81 |
user_input = st.text_area('Enter text','', height=200)
|
| 82 |
|
| 83 |
paraphraseNo = st.slider('Number of Parapharases',1,10,2)
|
| 84 |
-
if st.button('Paraphrase'):
|
| 85 |
with st.spinner(text="This may take a moment..."):
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 89 |
#with spacer:
|
| 90 |
|
| 91 |
with col2:
|
|
|
|
| 39 |
paraphrase.append(a)
|
| 40 |
paraphrase2 = [' '.join(x) for x in paraphrase]
|
| 41 |
paraphrase3 = [' '.join(x for x in paraphrase2) ]
|
| 42 |
+
paraphrased_text = paraphrase #str(paraphrase3).strip('[]').strip("'")
|
| 43 |
return paraphrased_text
|
| 44 |
|
| 45 |
|
|
|
|
| 49 |
return paraphrased_text
|
| 50 |
########################################################################################################
|
| 51 |
|
| 52 |
+
from transformers import *
|
| 53 |
|
| 54 |
+
model = PegasusForConditionalGeneration.from_pretrained("tuner007/pegasus_paraphrase")
|
| 55 |
+
tokenizer = PegasusTokenizerFast.from_pretrained("tuner007/pegasus_paraphrase")
|
| 56 |
|
| 57 |
|
| 58 |
+
def get_paraphrased_sentences(model, tokenizer, sentence, num_return_sequences=5, num_beams=5):
|
| 59 |
+
# tokenize the text to be form of a list of token IDs
|
| 60 |
+
inputs = tokenizer([sentence], truncation=False, padding="longest", return_tensors="pt")
|
| 61 |
+
# generate the paraphrased sentences
|
| 62 |
+
outputs = model.generate(
|
| 63 |
+
**inputs,
|
| 64 |
+
num_beams=num_beams,
|
| 65 |
+
num_return_sequences=num_return_sequences,
|
| 66 |
+
)
|
| 67 |
+
# decode the generated sentences using the tokenizer to get them back to text
|
| 68 |
+
return tokenizer.batch_decode(outputs, skip_special_tokens=True)
|
| 69 |
|
| 70 |
|
| 71 |
|
|
|
|
| 81 |
user_input = st.text_area('Enter text','', height=200)
|
| 82 |
|
| 83 |
paraphraseNo = st.slider('Number of Parapharases',1,10,2)
|
| 84 |
+
if st.button('Single-Paraphrase'):
|
| 85 |
with st.spinner(text="This may take a moment..."):
|
| 86 |
+
output = summarize(user_input)
|
| 87 |
+
|
| 88 |
+
if st.button('Multiple-Paraphrase'):
|
| 89 |
+
with st.spinner(text="This may take a moment..."):
|
| 90 |
+
output = get_paraphrased_sentences(model, tokenizer, user_input, num_beams=10, num_return_sequences=paraphraseNo)
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
|
| 94 |
#with spacer:
|
| 95 |
|
| 96 |
with col2:
|