-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtodo
69 lines (42 loc) · 2.97 KB
/
todo
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
connect to db ok
simple search for db (select) ok
token 的作用 ok
{
3(id),
00001469-0017.jpg(useless filename),
女童黄色连衣裙(description important for searching),
(token meaning : the first 101 and the last 102 is for alignment, and the rest means chinese letter, for example 出 is 1139,
Textual Encoder. The textual encoder is a standard decoder-only transformer as in [35]. We use
WordPiece [52] with a vocabulary size of 21,128 for Chinese text tokenization. Similar to [33],
we add spaces around Chinese characters before applying WordPiece so that Chinese is effectively
character-tokenized. We add two special tokens (i.e., [CLS] and [SEP]) at the beginning and ending
of each text sequence. The text encoder has 12 layers, each of which has 8 attention heads and a
hidden state dimension of 512)
[101, 1957, 4997, 7942, 5682, 6825, 6132, 6170, 102, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
https://gimg2.baidu.com/image_search/src=http%3A%2F%2Fa.vpimg2.com%2Fupload%2Fmerchandise%2F227958%2FLYQ-S314186413-3.jpg&refer=http%3A%2F%2Fa.vpimg2.com&app=2002&size=f9999,10000&q=a80&n=0&g=0n&fmt=jpeg?sec=1632501843&t=b0a3b843f9ecebd71fe6f27643c17486 (url to access image)
}
search for db using 分词 (slow searching, need to be improved)
how to search for cut-word? Should I read the source code?
additional:
index compression?
query suggesting, pagerank.
# 2.25
1. write the Put function to preprocess the data, also as an interface for future custom adding index function
2. preprocess the inverted index, ids encode as .gob file,store them in leveldb
3. query function(make it simple, single thread), do the ranking.
4. write front end.
====================================================================================================
requirement: high performance searching(how to establish a good searching data structure from mysql)
priority: backend
====================================================================================================
连衣裙: 2 4 13 397 556 764 1229 1901 2045 2052 2699 2875 3719 3934 4010 4094 4496 5245 5306 5469 5642 5776 5896 6290 7127 7250 7372 8424 8702 8789 8927 9154 9885 9947 10398 10583 10590 10883 11831 12059 12085 12097 12369 12499 13075 13120 13556 13674 13691 13878 14147 14616 15221 15267 15451 16364 16763 16815 16851 16938 17015 18793 18920 19268 19443 19529 20230 21998 22157 22409 22520 22524 22764 23398 23982 23988 24106 24464 24730 24837 24920 25062 25318 25840 26117 26610 26642 26647 26806 26824 27022 27759 27760 27821 27831 28006 28450 28838 29208 29240 29254 29633 30700 30736 30831 30943 30945 31150 32146 32640
你的孩子符合条件吗?滨海这所幼儿园即将开园!
# 3.10
use gin to fetch the data from client side.
use vue.js to send the data to server side.
how to use vue.js in the app?
is it in project file?
do you use npm run to run the app?
or do you use gin?
gin used only for backend data transfer, use npm run to start the frontend app.
web-app/assets main.css stores #app