Spaces:
Runtime error
Runtime error
fix dependencies
Browse files- app.py +10 -38
- requirements.txt +3 -3
app.py
CHANGED
@@ -15,36 +15,6 @@ def get_model():
|
|
15 |
tokenizer, model = get_model()
|
16 |
|
17 |
|
18 |
-
def predict_answer(qa_text_pair):
|
19 |
-
# Encoding
|
20 |
-
encodings = tokenizer(
|
21 |
-
qa_text_pair['question'], qa_text_pair['context'],
|
22 |
-
max_length=512,
|
23 |
-
truncation=True,
|
24 |
-
padding="max_length",
|
25 |
-
return_token_type_ids=False,
|
26 |
-
return_offsets_mapping=True
|
27 |
-
)
|
28 |
-
encodings = {key: torch.tensor([val]).to(device) for key, val in encodings.items()}
|
29 |
-
|
30 |
-
# Predict
|
31 |
-
with torch.no_grad():
|
32 |
-
pred = model(encodings['input_ids'], encodings['attention_mask'])
|
33 |
-
start_logits, end_logits = pred.start_logits, pred.end_logits
|
34 |
-
token_start_index, token_end_index = start_logits.argmax(dim=-1), end_logits.argmax(dim=-1)
|
35 |
-
pred_ids = encodings['input_ids'][0][token_start_index: token_end_index + 1]
|
36 |
-
|
37 |
-
# Answer start/end offset of context.
|
38 |
-
answer_start_offset = int(encodings['offset_mapping'][0][token_start_index][0][0])
|
39 |
-
answer_end_offset = int(encodings['offset_mapping'][0][token_end_index][0][1])
|
40 |
-
answer_offset = (answer_start_offset, answer_end_offset)
|
41 |
-
|
42 |
-
# Decoding
|
43 |
-
answer_text = tokenizer.decode(pred_ids) # text
|
44 |
-
del encodings
|
45 |
-
return {'answer_text':answer_text, 'answer_offset':answer_offset}
|
46 |
-
|
47 |
-
|
48 |
## Title
|
49 |
st.title('☁️ Bespin → QuestionAnswering')
|
50 |
|
@@ -89,24 +59,26 @@ if st.button("Submit", key='question'):
|
|
89 |
max_length=512,
|
90 |
truncation=True,
|
91 |
padding="max_length",
|
92 |
-
return_token_type_ids=False
|
|
|
93 |
)
|
94 |
encodings = {key: torch.tensor([val]) for key, val in encodings.items()}
|
95 |
-
input_ids = encodings["input_ids"]
|
96 |
-
attention_mask = encodings["attention_mask"]
|
97 |
|
98 |
# Predict
|
99 |
-
pred = model(input_ids, attention_mask=attention_mask)
|
100 |
-
|
101 |
start_logits, end_logits = pred.start_logits, pred.end_logits
|
102 |
token_start_index, token_end_index = start_logits.argmax(dim=-1), end_logits.argmax(dim=-1)
|
103 |
-
pred_ids = input_ids[0][token_start_index: token_end_index + 1]
|
104 |
-
|
105 |
-
# Decoding
|
106 |
prediction = tokenizer.decode(pred_ids)
|
107 |
|
|
|
|
|
|
|
|
|
|
|
108 |
# answer
|
109 |
st.success(prediction)
|
110 |
|
|
|
111 |
except Exception as e:
|
112 |
st.error(e)
|
|
|
15 |
tokenizer, model = get_model()
|
16 |
|
17 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
18 |
## Title
|
19 |
st.title('☁️ Bespin → QuestionAnswering')
|
20 |
|
|
|
59 |
max_length=512,
|
60 |
truncation=True,
|
61 |
padding="max_length",
|
62 |
+
return_token_type_ids=False,
|
63 |
+
return_offsets_mapping=True
|
64 |
)
|
65 |
encodings = {key: torch.tensor([val]) for key, val in encodings.items()}
|
|
|
|
|
66 |
|
67 |
# Predict
|
68 |
+
pred = model(encodings["input_ids"], attention_mask=encodings["attention_mask"])
|
|
|
69 |
start_logits, end_logits = pred.start_logits, pred.end_logits
|
70 |
token_start_index, token_end_index = start_logits.argmax(dim=-1), end_logits.argmax(dim=-1)
|
71 |
+
pred_ids = encodings["input_ids"][0][token_start_index: token_end_index + 1]
|
|
|
|
|
72 |
prediction = tokenizer.decode(pred_ids)
|
73 |
|
74 |
+
# Offset
|
75 |
+
answer_start_offset = int(encodings['offset_mapping'][0][token_start_index][0][0])
|
76 |
+
answer_end_offset = int(encodings['offset_mapping'][0][token_end_index][0][1])
|
77 |
+
answer_offset = (answer_start_offset, answer_end_offset)
|
78 |
+
|
79 |
# answer
|
80 |
st.success(prediction)
|
81 |
|
82 |
+
|
83 |
except Exception as e:
|
84 |
st.error(e)
|
requirements.txt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
-
torch
|
2 |
-
transformers==4.
|
3 |
-
streamlit
|
|
|
1 |
+
torch==1.11.0
|
2 |
+
transformers==4.20.0
|
3 |
+
streamlit==1.10.0
|