📌 简介
nlhappy致力于复现自然语言处理各类任务的SOTA模型。
文档地址:
它主要的依赖有
- transformers: 下载预训练权重
- pytorch-lightning: 模型训练
- datasets: 构建数据集
- pydantic: 数据校验
📌 安装
安装nlhappy
推荐先去pytorch官网安装pytorch和对应cuda
# pip 安装
pip install --upgrade pip
pip install --upgrade nlhappy
其他可选
推荐安装wandb用于可视化训练日志
- 安装:
pip install wandb
- 注册: https://wandb.ai/
- 获取认证: https://wandb.ai/authorize
- 登陆:
wandb login
- 使用
# 命令行训练
nlhappy datamodule=xxx model=xxx trainer=xxx logger=wandb
模型训练开始后去官网查看训练实况
📌 模型复现
实体抽取
| 模型名称 | 参考链接 |
|---|---|
| GlobalPointer | 科学空间 |
| EfficientGlobalPointer | 科学空间 |
关系抽取
| 模型名称 | 参考链接 |
|---|---|
| GPLinker | 科学空间 |
事件抽取
| 模型名称 | 参考链接 |
|---|---|
| GPLinker | 科学空间 |
答案抽取
| 模型名称 | 参考链接 |
|---|---|
| GPLinker | - |
Metadata
Release files for nlhappy 2023.1.31
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| nlhappy-2023.1.31.tar.gz | 92.3 kB | Without provenance | Details for nlhappy-2023.1.31.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| nlhappy-2023.1.31-py3-none-any.whl | Python 3 | none | any | Details for nlhappy-2023.1.31-py3-none-any.whl |
Total release size: 247.8 kB
Release files / nlhappy-2023.1.31.tar.gz
| Download URL | nlhappy-2023.1.31.tar.gz |
|---|---|
| Size | 92.3 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
8ec086334509ed78db859cf57d50e21a9689f1a1def1ee812686798cf51ca21e
|
|
BLAKE2b-256 checksum How to use checksums |
ef660a3714c2dacab98df0abeafdc7652b09cd58b67e45c1fc369c43ecfcfd98
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
poetry/1.1.15 CPython/3.8.10 Linux/5.4.0-126-generic
|
Release files / nlhappy-2023.1.31-py3-none-any.whl
| Download URL | nlhappy-2023.1.31-py3-none-any.whl |
|---|---|
| Size | 155.5 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
909b14b959a9b87eeddd7d37330afe2c36d268c03e8ca084e2b537a3773169dd
|
|
BLAKE2b-256 checksum How to use checksums |
819ac5c65805e5358ec7775bb8db0fca604fdafb19866372fa3d3c8181142ee8
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
poetry/1.1.15 CPython/3.8.10 Linux/5.4.0-126-generic
|
Security for nlhappy 2023.1.31
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.