Skip to main content

NLP text processing toolkit for Deep Learning

Project description

FERN

Pypi

Fern用于NLP的模型开发结构控制。通过它可以控制文本预处理、模型搭建、训练器:

  1. 文本预处理:数据下载、数据清洗、数据转换和数据分割
  2. 模型搭建:模型保存与加载、模型架构打印
  3. 模型训练:单步/epoch训练与评估、评估函数设置、损失函数设置、label权重设置

Fern的设计目的主要为了解决不同NLP工程中重复代码过多问题,减少流程性代码,从而避免数据交互过程中的随机bug出现

安装

  1. pypi 安装

    $ pip install Fern2
    
  2. 从源码安装

    $ pip install -e git+https://github.com/Jasonsey/Fern.git
    

使用教程

建议查看源码中函数的使用说明

变量命名规则

为了方便定义,对容易分歧变量命名做如下约定:

  1. 对于数据变量,同类型变量书写规则:

    • data_train, data_val
    • label_train, label_val
  2. 对于指标变量,同类型变量书写规则:

    • val_loss, val_acc, val_binary_acc
    • train_loss, train_acc
  3. 对于其他变量,按照首先它属于a, 其次它属于b规则命名变量名:a_b

    • path_dataset

TODO

  • BERT预训练模型直接调用
  • BERT重现训练代码直接调用

版本变更日志

CHANGE LOG

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

Fern2-1.4.1.tar.gz (27.2 kB view details)

Uploaded Source

Built Distribution

Fern2-1.4.1-py3-none-any.whl (33.2 kB view details)

Uploaded Python 3

File details

Details for the file Fern2-1.4.1.tar.gz.

File metadata

  • Download URL: Fern2-1.4.1.tar.gz
  • Upload date:
  • Size: 27.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.11.3

File hashes

Hashes for Fern2-1.4.1.tar.gz
Algorithm Hash digest
SHA256 a851e6849fa4025169c551d0849f0c7fda63aa6512e70cc04aa8d34c39bbb746
MD5 7f8313a3225ea4587ea69e83420eb4f3
BLAKE2b-256 892577a290f3c85fc27a5cf610d33a47afab8fa4275331b65322d5f4f20d87ed

See more details on using hashes here.

File details

Details for the file Fern2-1.4.1-py3-none-any.whl.

File metadata

  • Download URL: Fern2-1.4.1-py3-none-any.whl
  • Upload date:
  • Size: 33.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.11.3

File hashes

Hashes for Fern2-1.4.1-py3-none-any.whl
Algorithm Hash digest
SHA256 af1396cc1fe99cafe2b0ce1a3776f2a02948fae73b889902a01de1d172bed41a
MD5 ef4fe08c1f1a681b17d5434a41647578
BLAKE2b-256 ca232407cd2d4db261571b9966da66ad1993c69e6a077ecf8ac8396f046f595f

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page