mirror of
https://github.com/wassname/multifit.git
synced 2026-08-23 12:20:23 +08:00
12 lines
328 B
Bash
12 lines
328 B
Bash
#!/usr/bin/env bash
|
|
|
|
ROOT="data"
|
|
DATA_DIR="${ROOT}/imdb"
|
|
mkdir -p "${DATA_DIR}"
|
|
echo "Saving data in $DATA_DIR"
|
|
wget -c "http://files.fast.ai/data/aclImdb.tgz" -P "${DATA_DIR}"
|
|
|
|
echo "Imdb is raw text so we are tokenizing it with Moses"
|
|
python -m fastai_contrib.utils prepare_imdb "${DATA_DIR}/aclImdb.tgz" --prepare_lm==False
|
|
|