24-2 시각지능학습 Image Generation

1. Clone the repository

git clone https://github.com/hy-vision-learning/generation-submit

2. Move to generation-submit

cd ./generation-submit

3. Install Required Packages

All required packages can be installed via requirements.txt.

pip install -r requirements.txt

Note

If an error related to package versions occurs during installation, remove the version information and try again.

4. Change the random seed

Open the change_randomseed.py file and change the random seed.

5. Generate Evaluation Data

Run the command below to generate data for performance evaluation.

python3 generate_inception_data.py

6. Run the code

Run the command below to start training.

python3 main.py  --model BIGGAN --num_epochs 600 --save_every 1000 --test_every 500 --full_test_counter 10 --superclass 1 --dict_size 4 --commitment 10.0 --dict_decay 0.9

7. Check the Results

Check the best performance and training time in the log file located inside the save/biggan/[date] folder.