Update app.py
Browse files
app.py
CHANGED
@@ -33,21 +33,34 @@ infill: sports teams are profitable for owners. ( accumulating vast sums / stock
|
|
33 |
|
34 |
original:"""
|
35 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
36 |
with st.form(key='my_form'):
|
37 |
prompt = st.text_area(label='Enter sentence', value=g)
|
38 |
submit_button = st.form_submit_button(label='Submit')
|
|
|
39 |
if submit_button:
|
40 |
with torch.no_grad():
|
41 |
text = tokenizer.encode(prompt)
|
42 |
myinput, past_key_values = torch.tensor([text]), None
|
43 |
myinput = myinput
|
44 |
-
myinput= myinput
|
45 |
logits, past_key_values = model(myinput, past_key_values = past_key_values, return_dict=False)
|
46 |
logits = logits[0,-1]
|
47 |
probabilities = torch.nn.functional.softmax(logits)
|
48 |
-
best_logits, best_indices = logits.topk(
|
49 |
best_words = [tokenizer.decode([idx.item()]) for idx in best_indices]
|
50 |
text.append(best_indices[0].item())
|
51 |
best_probabilities = probabilities[best_indices].tolist()
|
52 |
words = []
|
53 |
-
st.write(best_words)
|
|
|
|
|
|
33 |
|
34 |
original:"""
|
35 |
|
36 |
+
def prefix_format(sentence):
|
37 |
+
words = sentence.split()
|
38 |
+
if "[MASK]" in sentence:
|
39 |
+
words2 = words.index("[MASK]")
|
40 |
+
#print(words2)
|
41 |
+
output = ("<|SUF|> " + ' '.join(words[words2+1:]) + " <|PRE|> " + ' '.join(words[:words2]) + " <|MID|>")
|
42 |
+
st.write(output)
|
43 |
+
else:
|
44 |
+
st.write("Add [MASK] to sentence")
|
45 |
+
|
46 |
with st.form(key='my_form'):
|
47 |
prompt = st.text_area(label='Enter sentence', value=g)
|
48 |
submit_button = st.form_submit_button(label='Submit')
|
49 |
+
submit_button6 = st.form_submit_button(label='Turn Into Infill Format. Just add [MASK] where you want it infilled')
|
50 |
if submit_button:
|
51 |
with torch.no_grad():
|
52 |
text = tokenizer.encode(prompt)
|
53 |
myinput, past_key_values = torch.tensor([text]), None
|
54 |
myinput = myinput
|
55 |
+
myinput= myinput
|
56 |
logits, past_key_values = model(myinput, past_key_values = past_key_values, return_dict=False)
|
57 |
logits = logits[0,-1]
|
58 |
probabilities = torch.nn.functional.softmax(logits)
|
59 |
+
best_logits, best_indices = logits.topk(250)
|
60 |
best_words = [tokenizer.decode([idx.item()]) for idx in best_indices]
|
61 |
text.append(best_indices[0].item())
|
62 |
best_probabilities = probabilities[best_indices].tolist()
|
63 |
words = []
|
64 |
+
st.write(best_words)
|
65 |
+
if submit_button6:
|
66 |
+
prefix_format(prompt)
|