-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathpredict.sh
60 lines (56 loc) · 2.9 KB
/
predict.sh
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
<<'COMMENT'
1. 将NER和RE的pth预训练模型移至目的文件夹
1.1 挂载方式:
resource/pretrain_models_input => Model_train_predict/example/pretrain_models_input # 因为是gpu训练, cpu预测, 所以需要将预训练模型导入
COMMENT
mkdir Model_train_predict/example/ner/standard/checkpoints/
mkdir Model_train_predict/example/re/standard/checkpoints/
cp Model_train_predict/example/pretrain_models_input/model_ner.pth Model_train_predict/example/ner/standard/checkpoints/model_ner.pth
cp Model_train_predict/example/pretrain_models_input/model_re.pth Model_train_predict/example/re/standard/checkpoints/model_re.pth
<<'COMMENT'
2. 命名实体模型预测 NER
2.1 复制所有教程或指导等txt文档到Model_train_predict/example/ner/standard/data/pre_data_dir/下
# 2.2 挂载方式:
# resource/predict_ner_input => /kg/Model_train_predict/example/ner/standard/data/pre_data_dir # ner模型预测所需的数据。这已经是全部数据, 后续re模型预测时, 会调用ner预测结果作为输入。
COMMENT
# 2.3 运行Model_train_predict/example/ner/standard/predict.py，在当前目录生成ner_data.json文件
cd Model_train_predict/example/ner/standard
python predict.py
# 2.4 运行当前目录下Entity_data.py文件，当前目录生成ner_entity.csv文件
python Entity_data.py
<<'COMMENT'
3. 关系抽取模型预测 RE
COMMENT
# 3.1 复制上述Model_train_predict/example/ner/standard/ner_data.json文件到Model_train_predict/example/re/standard/目录下
cp /kg/Model_train_predict/example/ner/standard/ner_data.json /kg/Model_train_predict/example/re/standard/
# 3.2 进入Model_train_predict/example/re/standard
cd /kg/Model_train_predict/example/re/standard
# 3.2 运行当前目录下re_predict.py, 生成neo4j_relation.csv文件
python re_predict.py
# 3.4 返回Model_train_predict/example
<<'COMMENT'
4. 保存预训练模型预测的csv结果至外部挂载
4.1 挂载方式:
resource/neo4j/import => Model_train_predict/example/ner/standard/data/pre_data_dir/output # re模型预测所需的数据
COMMENT
# rm -rf /kg/Model_train_predict/example/predict_output
# mkdir /kg/Model_train_predict/example/predict_output
cp /kg/Model_train_predict/example/ner/standard/ner_entity.csv /kg/Model_train_predict/example/predict_output/ner_entity.csv
cp /kg/Model_train_predict/example/re/standard/neo4j_relation.csv /kg/Model_train_predict/example/predict_output/neo4j_relation.csv
<<'COMMENT'
5. 启动后端服务
COMMENT
# 5.1 启动外部服务API端口
cd /kg/Search_ui
# python Search_api.py
nohup python Search_api.py > Search_api.log 2>&1 &
# 5.2 启动neo4j查询服务
cd /kg/Search_neo4j
# python Neo_data_search.py
nohup python Neo_data_search.py > Neo_data_search.log 2>&1 &
# 5.3 启动es查询服务
cd /kg/es_search
# python Search_api.py
nohup python Search_api.py > Search_api.log 2>&1 &
# 5.4 返回起始工作目录
cd /kg