support downloading updated data
Browse files- app.py +1 -1
- merge_qas.py +56 -0
app.py
CHANGED
@@ -173,7 +173,7 @@ def download(access_key):
|
|
173 |
if not access_key.startswith(os.getenv('access_key')):
|
174 |
chatbot_ret = [(f"Your entered Access Key:<br>{access_key}<br>is incorrect.", f"So i cannot provide you any information in this private space.")]
|
175 |
file_ret = gr.File.update(value=None, visible=False)
|
176 |
-
elif
|
177 |
chatbot_ret = [(f"Your entered Access Key is correct.", f"The file containing new processed query-responses ({len(qas)-num_read_qas} in total) can be downloaded below.")]
|
178 |
filename = f"qas-{num_read_qas}-{len(qas)}.json"
|
179 |
with open(filename, "w", encoding="utf-8") as f:
|
|
|
173 |
if not access_key.startswith(os.getenv('access_key')):
|
174 |
chatbot_ret = [(f"Your entered Access Key:<br>{access_key}<br>is incorrect.", f"So i cannot provide you any information in this private space.")]
|
175 |
file_ret = gr.File.update(value=None, visible=False)
|
176 |
+
elif access_key == f"{os.getenv('access_key')}: update":
|
177 |
chatbot_ret = [(f"Your entered Access Key is correct.", f"The file containing new processed query-responses ({len(qas)-num_read_qas} in total) can be downloaded below.")]
|
178 |
filename = f"qas-{num_read_qas}-{len(qas)}.json"
|
179 |
with open(filename, "w", encoding="utf-8") as f:
|
merge_qas.py
ADDED
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import json
|
3 |
+
import collections
|
4 |
+
|
5 |
+
def read_qs():
|
6 |
+
qs = []
|
7 |
+
directory = "./dialogues_set"
|
8 |
+
filenames = [
|
9 |
+
'dialogues_film.json',
|
10 |
+
'dialogues_jindong.json',
|
11 |
+
'dialogues_music.json',
|
12 |
+
'dialogues_natural.json',
|
13 |
+
'dialogues_taobao.json',
|
14 |
+
'dialogues_travel_kd.json'
|
15 |
+
]
|
16 |
+
for filename in filenames:
|
17 |
+
with open(f"{directory}/{filename}", "r", encoding="utf-8") as f:
|
18 |
+
for idx,line in enumerate(f):
|
19 |
+
idx2query = json.loads(line)
|
20 |
+
query = idx2query[str(idx)]
|
21 |
+
qs.append(query)
|
22 |
+
print(f"read {len(qs)} queries from files")
|
23 |
+
return qs
|
24 |
+
|
25 |
+
def read_qas():
|
26 |
+
qas = []
|
27 |
+
directory = "./dialogues_set"
|
28 |
+
|
29 |
+
for filename in os.listdir(directory):
|
30 |
+
if filename.endswith(".json") and "qas" in filename:
|
31 |
+
with open(f"{directory}/{filename}", "r", encoding="utf-8") as f:
|
32 |
+
for qa in json.loads(f.read()):
|
33 |
+
qas.append(qa)
|
34 |
+
return qas
|
35 |
+
|
36 |
+
def merge(qs, qas):
|
37 |
+
q_to_as = collections.defaultdict(lambda:[])
|
38 |
+
for qa in qas:
|
39 |
+
q_to_as[qa["q"]].append(qa["a"])
|
40 |
+
|
41 |
+
qas = []
|
42 |
+
for q in qs:
|
43 |
+
if len(q_to_as[q])==0:
|
44 |
+
continue
|
45 |
+
a = q_to_as[q].pop()
|
46 |
+
qas.append({"q":q, "a":a})
|
47 |
+
|
48 |
+
return qas
|
49 |
+
|
50 |
+
|
51 |
+
if __name__ == "__main__":
|
52 |
+
qs = read_qs()
|
53 |
+
qas = read_qas()
|
54 |
+
qas = merge(qs, qas)
|
55 |
+
with open("./dialogues_set/qas.json", "w", encoding="utf-8") as f:
|
56 |
+
f.write(json.dumps(qas, ensure_ascii=False, indent=2))
|