guzy0324 / meta_learned_review_helpfulness_prediction

0 stars 0 forks source link

meta_learned_review_helpfulness_prediction

Installation

  1. 下载代码,创建环境

    $ git clone git@github.com:guzy0324/meta_learned_review_helpfulness_prediction.git
    $ cd meta_learned_review_helpfulness_prediction
    $ conda env create -f env.yml
  2. 注册kaggle账户并获得kaggle授权

  3. 激活环境,退出环境

    # 激活环境
    $ conda activate MeRH
    # 退出环境
    $ conda deactivate

Demo

  1. PRH-Net fit

    python -m MeRH.PRH_Net fit -s 42 -bs 2 -d amazon-9_splited -da '{"count": 700}' -nw 128 -ss '["Grocery & Gourmet Food"]' -e glove -rr BiLSTM -rra '{"r": 3}' -o AdamW -as Identification -me 30 -a 1
  2. MeRH fit

    python -m MeRH.MeRH fit -s 42 -d amazon-9_splited -ss '["Grocery & Gourmet Food"]' -bs 8 -nw 128 -e glove -rr BiLSTM -rra '{"r": 3}' -a 1 -o AdamW -as Identification -me 20