Spaces:
Paused
Paused
Update app.py
Browse files
app.py
CHANGED
@@ -4,6 +4,7 @@ from sentence_transformers import SentenceTransformer
|
|
4 |
import pickle
|
5 |
import re
|
6 |
from transformers import pipeline
|
|
|
7 |
|
8 |
|
9 |
|
@@ -22,19 +23,21 @@ with open('articles.pkl', 'rb') as file:
|
|
22 |
|
23 |
index_loaded = faiss.read_index("sentence_embeddings_index_no_citation.faiss")
|
24 |
|
|
|
|
|
25 |
if 'model_embedding' not in st.session_state:
|
26 |
-
st.session_state.model_embedding = SentenceTransformer('bkai-foundation-models/vietnamese-bi-encoder')
|
27 |
|
28 |
|
29 |
|
30 |
# Replace this with your own checkpoint
|
31 |
model_checkpoint = "model"
|
32 |
-
question_answerer = pipeline("question-answering", model=model_checkpoint)
|
33 |
def question_answering(question):
|
34 |
print(question)
|
35 |
query_sentence = [question]
|
36 |
query_embedding = st.session_state.model_embedding.encode(query_sentence)
|
37 |
-
k =
|
38 |
D, I = index_loaded.search(query_embedding.astype('float32'), k) # D is distances, I is indices
|
39 |
answer = [question_answerer(question=query_sentence[0], context=articles[I[0][i]], max_answer_len = 256) for i in range(k)]
|
40 |
best_answer = max(answer, key=lambda x: x['score'])
|
|
|
4 |
import pickle
|
5 |
import re
|
6 |
from transformers import pipeline
|
7 |
+
import torch
|
8 |
|
9 |
|
10 |
|
|
|
23 |
|
24 |
index_loaded = faiss.read_index("sentence_embeddings_index_no_citation.faiss")
|
25 |
|
26 |
+
|
27 |
+
device = 0 if torch.cuda.is_available() else -1
|
28 |
if 'model_embedding' not in st.session_state:
|
29 |
+
st.session_state.model_embedding = SentenceTransformer('bkai-foundation-models/vietnamese-bi-encoder', device = device)
|
30 |
|
31 |
|
32 |
|
33 |
# Replace this with your own checkpoint
|
34 |
model_checkpoint = "model"
|
35 |
+
question_answerer = pipeline("question-answering", model=model_checkpoint, device = device)
|
36 |
def question_answering(question):
|
37 |
print(question)
|
38 |
query_sentence = [question]
|
39 |
query_embedding = st.session_state.model_embedding.encode(query_sentence)
|
40 |
+
k = 20
|
41 |
D, I = index_loaded.search(query_embedding.astype('float32'), k) # D is distances, I is indices
|
42 |
answer = [question_answerer(question=query_sentence[0], context=articles[I[0][i]], max_answer_len = 256) for i in range(k)]
|
43 |
best_answer = max(answer, key=lambda x: x['score'])
|