cube-studio

mirror of https://github.com/tencentmusic/cube-studio.git synced 2024-12-21 06:19:31 +08:00

History

FerdinandWard 0c1968fe45 fix readme		2022-08-18 15:16:10 +08:00
..
bert_pretrain	add ner train and service	2022-08-13 17:00:41 +08:00
models	add ner train and service	2022-08-13 17:00:41 +08:00
tokenizer	add ner train and service	2022-08-13 17:00:41 +08:00
utils	add ner train and service	2022-08-13 17:00:41 +08:00
bert_bilstm_crf_opration.py	add ner train and service	2022-08-13 17:00:41 +08:00
bilstm_crf_opration.py	add ner train and service	2022-08-13 17:00:41 +08:00
bilstm_opration.py	add ner train and service	2022-08-13 17:00:41 +08:00
build.sh	add ner train and service	2022-08-13 17:00:41 +08:00
config.py	add ner train and service	2022-08-13 17:00:41 +08:00
data.py	add ner train and service	2022-08-13 17:00:41 +08:00
Dockerfile	fix readme	2022-08-18 15:16:10 +08:00
evaluating.py	add ner train and service	2022-08-13 17:00:41 +08:00
main4argo.py	add ner train and service	2022-08-13 17:00:41 +08:00
main.py	add ner train and service	2022-08-13 17:00:41 +08:00
README.md	fix readme	2022-08-18 15:16:10 +08:00
the_peoples_daily_tool.py	add ner train and service	2022-08-13 17:00:41 +08:00
train_evaluate.py	add ner train and service	2022-08-13 17:00:41 +08:00

README.md

镜像：ccr.ccs.tencentyun.com/cube-studio/ner:20220812 启动参数：

 "参数分组1": {
        "--model": {
            "type": "str",
            "item_type": "str",
            "label": "参数1",
            "require": 1,
            "choice": [],
            "range": "",
            "default": "BiLSTM_CRF",
            "placeholder": "",
            "describe": "model",
            "editable": 1,
            "condition": "",
            "sub_args": {}
        },
        "--objectname": {
            "type": "str",
            "item_type": "str",
            "label": "参数2",
            "require": 1,
            "choice": [
                "resume_BIO.txt",
                "people_daily_BIO.txt"
            ],
            "range": "",
            "default": "resume_BIO.txt",
            "placeholder": "",
            "describe": "resume_BIO",
            "editable": 1,
            "condition": "",
            "sub_args": {}
        },
        "--epochs": {
            "type": "str",
            "item_type": "str",
            "label": "参数3",
            "require": 1,
            "choice": [],
            "range": "",
            "default": "1",
            "placeholder": "",
            "describe": "epochs",
            "editable": 1,
            "condition": "",
            "sub_args": {}
        },
        "--path": {
            "type": "str",
            "item_type": "str",
            "label": "参数4",
            "require": 1,
            "choice": [],
            "range": "",
            "default": "/mnt/admin/NER/zdata/",
            "placeholder": "",
            "describe": "path",
            "editable": 1,
            "condition": "",
            "sub_args": {}
        },
        "-n": {
            "type": "str",
            "item_type": "str",
            "label": "参数5",
            "require": 1,
            "choice": [
                "resume_BIO.txt",
                "people_daily_BIO.txt"
            ],
            "range": "",
            "default": "resume_BIO.txt",
            "placeholder": "",
            "describe": "和objectname 保持一致",
            "editable": 1,
            "condition": "",
            "sub_args": {}
        },
        "-pp": {
            "type": "str",
            "item_type": "str",
            "label": "参数6",
            "require": 1,
            "choice": [],
            "range": "",
            "default": "/mnt/admin/model.pkl",
            "placeholder": "",
            "describe": "模型保存目录",
            "editable": 1,
            "condition": "",
            "sub_args": {}
        }
    }
}

注意：

训练需要的 txt 文件，以及训练结束后生成的 .pkl 文件，因为太大了，我都用 git 忽略了。

要运行项目，请在钉钉群里下载 data 和 zdata 两个文件夹，然后放置到本目录中。

README.md Unescape Escape

注意：

参考资料：

README.md