- Things I've learned while applying fastText to Korean NLP tasks
-
Build an image from the
Dockerfile
at the<repository_root_directory>
docker build -t <image_name>:<tag_name> .
- E.g.,
docker build -t fasttext:dev .
-
Run a container from the image
<image_name>:<tag_name>
built in step 1 -
Download pretrained fastText files
-
Open
./examples/baseline/notebook.ipynb
and setconfig
appropriately- E.g.,
config['pretrained_bin']
is a path of pretrained fastText.bin
file
- E.g.,
-
Run the
notebook.ipynb
and See explanation