This is a term project of Web Intelligence and Message Understanding, NCU 2021.
-
Install cuda.
-
Setup env and install packages.
conda activate wimuta_hw3_ner_env conda env list conda create --name wimuta_hw3_ner_env python=3.7 conda activate wimuta_hw3_ner_env pip install tensorflow==2.3.0-rc1 pip install tensorflow-gpu==2.3.0-rc1 pip install sklearn pip install pandas pip install IPython pip install keras pip install tqdm pip install xlrd pip install openpyxl conda install -c anaconda ipython pip install matplotlib pip install tensorflow_hub pip install transformers==4.4.2 pip install gin-config pip install tensorflow_addons conda install cudatoolkit conda install numba
-
Change package path in
conda.pth
and paste it into yoursite-packages
folder of this env. -
Run
main.py
. -
Install Perl.
-
Run evaluate code.
perl .\conlleval.pl "<" .\data\example_datasets_msra\label_test_output.txt