Env

conda create -n song python=3.8 -y
conda activate song
conda install -y pytorch==1.11.0 torchvision==0.12.0 torchaudio==0.11.0 cudatoolkit=10.2 -c pytorch
pip install -r requirements.txt

Run experiment

Setup

Get Encoder

step 1: 先將 P-learning 產生的 .mat 放到 other_mats/$dataset/ 中
step 2: 在 ./scripts/origin_scripts/ 中新增 $dataset.sh(可參考其他 .sh 的設定)
step 3: 執行 ./scripts/origin_scripts/$dataset.sh

ESZSL

./vaegan_v4/awa2.sh > ~/results/eszsl/v4_wgan_vaegan_awa

./vaegan_v4/cub.sh > ~/results/eszsl/v4_wgan_vaegan_cub

./vaegan_v4/sun.sh > ~/results/eszsl/v4_wgan_vaegan_sun

CADA

setup environment

act cadavae

./vaegan_scripts/awa2.sh > ~/results/cada/v4_wgan_vaegan_awa

./vaegan_scripts/cub.sh > ~/results/cada/v4_wgan_vaegan_cub

./vaegan_scripts/sun.sh > ~/results/cada/v4_wgan_vaegan_sun

tfvae

setup environment

act tfvaegan

./vaegan_scripts/awa2.py > ~/results/tfvae/v4_wgan_vaegan_awa

./vaegan_scripts/cub.py > ~/results/tfvae/v4_wgan_vaegan_cub

./vaegan_scripts/sun.py > ~/results/tfvae/v4_wgan_vaegan_sun