Summary

A easy and general DL toolkit based on pytorch.

一个基于pytorch的简单和通用的深度学习工具集。

Usage

train example

    # step1.input
    logging.basicConfig(level=logging.DEBUG, format='%(levelname)s-%(asctime)s-%(message)s')
    # step2.conf
    conf: pipeline.TextClassifyConfig = ConfigFactory.get_config('data/text_classify/config.ini')
    preprocessor: pipeline.TextClassifyPreprocessor = PreprocessorFactory.get_preprocessor(conf)
    preprocessor.preprocess()
    # step3.model
    model_manager = ModelManager(conf)
    model = model_manager.get_model()
    # step4.train
    trainer: pipeline.TextClassifyTrainer = TrainerFactory.get_trainer(conf, model)
    trainer.start()  # 88.69%

inference example

    # step1.input
    logging.basicConfig(level=logging.DEBUG, format='%(levelname)s-%(asctime)s-%(message)s')
    # step2.conf
    conf: pipeline.TextClassifyConfig = ConfigFactory.get_config('data/text_classify/config.ini')
    preprocessor: pipeline.TextClassifyPreprocessor = PreprocessorFactory.get_preprocessor(conf)
    preprocessor.preprocess()
    # step3.model
    model_manager = ModelManager(conf)
    # ste4.inference
    text = ['词汇阅读是关键 08年考研暑期英语复习全指南', '自考经验谈：自考生毕业论文选题技巧', '本科未录取还有这些路可以走']
    model_manager.load_model()
    result = model_manager.infer(text)
    print(result)

Code style

https://numpydoc.readthedocs.io/en/latest/format.html

Model roadmap

text classify(multiple class)

• textCNN
• TextRNN
• BiLSTM+Attention
• TextRCNN
• Transformer
• bert

text entity recognition

• bi-lstm-crf

picture classify(multiple class)

• improved LeNet