Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/bojone/bert_in_keras
/ functions
Functions
49 in github.com/bojone/bert_in_keras
⨍
Functions
49
◇
Types & classes
11
↓ 9 callers
Function
seq_padding
(X, padding=0, maxlen=None)
nl2sql_baseline.py:84
↓ 6 callers
Function
seq_padding
(X, padding=0)
relation_extract.py:112
↓ 4 callers
Function
seq_padding
(X, padding=0)
subject_extract.py:83
↓ 3 callers
Function
read_data
(data_file, table_file)
nl2sql_baseline.py:34
↓ 3 callers
Function
seq_padding
(X, padding=0)
sentiment.py:60
↓ 2 callers
Method
__iter__
(self)
sentiment.py:77
↓ 2 callers
Function
extract_entity
(text_in, c_in)
subject_extract.py:191
↓ 2 callers
Function
extract_items
(text_in)
relation_extract.py:272
↓ 2 callers
Function
list_find
在list1中寻找子串list2，如果找到，返回第一个下标； 如果找不到，返回-1。
relation_extract.py:120
↓ 2 callers
Function
most_similar
从词表中找最相近的词（当无法全匹配的时候）
nl2sql_baseline.py:95
↓ 2 callers
Function
nl2sql
输入question和headers，转SQL
nl2sql_baseline.py:265
↓ 2 callers
Function
repair
(d)
relation_extract.py:76
↓ 2 callers
Function
softmax
(x)
subject_extract.py:185
↓ 1 callers
Method
__iter__
(self)
subject_extract.py:111
↓ 1 callers
Method
__iter__
(self)
nl2sql_baseline.py:125
↓ 1 callers
Method
__iter__
(self)
relation_extract.py:140
↓ 1 callers
Function
evaluate
(data, tables)
nl2sql_baseline.py:339
↓ 1 callers
Method
evaluate
(self)
subject_extract.py:238
↓ 1 callers
Method
evaluate
(self)
nl2sql_baseline.py:399
↓ 1 callers
Method
evaluate
(self)
relation_extract.py:345
↓ 1 callers
Function
is_equal
判断两个SQL字典是否全匹配
nl2sql_baseline.py:331
↓ 1 callers
Function
list_find
在list1中寻找子串list2，如果找到，返回第一个下标； 如果找不到，返回-1。
subject_extract.py:91
↓ 1 callers
Function
most_similar_2
从句子s中找与w最相近的片段， 借助分词工具和ngram的方式尽量精确地确定边界。
nl2sql_baseline.py:104
Method
__init__
(self, data, batch_size=32)
subject_extract.py:103
Method
__init__
(self)
subject_extract.py:215
Method
__init__
(self, data, tables, batch_size=32)
nl2sql_baseline.py:116
Method
__init__
(self)
nl2sql_baseline.py:375
Method
__init__
(self, data, batch_size=32)
relation_extract.py:132
Method
__init__
(self)
relation_extract.py:321
Method
__init__
(self, data, batch_size=32)
sentiment.py:69
Method
__len__
(self)
subject_extract.py:109
Method
__len__
(self)
nl2sql_baseline.py:123
Method
__len__
(self)
relation_extract.py:138
Method
__len__
(self)
sentiment.py:75
Method
_tokenize
(self, text)
subject_extract.py:32
Method
_tokenize
(self, text)
nl2sql_baseline.py:70
Method
_tokenize
(self, text)
relation_extract.py:31
Method
_tokenize
(self, text)
sentiment.py:27
Method
on_batch_begin
第一个epoch用来warmup，第二个epoch把学习率降到最低
subject_extract.py:219
Method
on_batch_begin
第一个epoch用来warmup，第二个epoch把学习率降到最低
nl2sql_baseline.py:380
Method
on_batch_begin
第一个epoch用来warmup，第二个epoch把学习率降到最低
relation_extract.py:326
Method
on_epoch_end
(self, epoch, logs=None)
subject_extract.py:231
Method
on_epoch_end
(self, epoch, logs=None)
nl2sql_baseline.py:392
Method
on_epoch_end
(self, epoch, logs=None)
relation_extract.py:338
Function
seq_gather
seq是[None, seq_len, s_size]的格式， idxs是[None, n]的格式，在seq的第i个序列中选出第idxs[i]个向量， 最终输出[None, n, s_size]的向量。
nl2sql_baseline.py:185
Function
seq_gather
seq是[None, seq_len, s_size]的格式， idxs是[None, 1]的格式，在seq的第i个序列中选出第idxs[i]个向量， 最终输出[None, s_size]的向量。
relation_extract.py:201
Function
test
(test_data)
subject_extract.py:251
Function
test
(data, tables, outfile='result.json')
nl2sql_baseline.py:358
Function
test
输出测试结果
relation_extract.py:375