Skip to content

Latest commit

 

History

History
27 lines (20 loc) · 575 Bytes

File metadata and controls

27 lines (20 loc) · 575 Bytes

Embedding-experiment-in-BioNLP-course

1. Data preparation.

unzip data/litcovid-trainingdata.zip

2. Environmental preparation.

If pytorch installation fails, please use the official website command line to install.

pip3 install -r requirements.txt

3. Quick start.

3.1 Word2Vec-base.

You can pursue a better embedding effect by modifying the parameter part of the code, such as train_size, embedding_size, batch_size, learning_rage, etc.

python3 Skip_Gram_basic.py

3.2 Bert

python3 Bert_4_Litcovid_WordEmbedding.py