Installtion (Windows 10):

Download Tiny (Distilled) Model from below Google Drive manually and copy to OFA folder
(https://drive.google.com/drive/folders/1jpP-KNabBPAyOoDkj_U57GkpR4JBjMV0?usp=sharing)

OR 

Download Large Model from below Google Drive manually and copy to OFA folder
(https://drive.google.com/drive/folders/1p__3PShIX6KbKBIyERSX9jrsUEN1hoR5?usp=sharing)

### Tiny is faster with reasonable accuracy of text caption while large is much slower with much higher accuracy of text caption

Option 1 (using existing virtual environment):

Download venv below and unzip to ShowAndTell folder
(https://drive.google.com/file/d/1fTqD683dpdwtGnUMyQc3tLraZK7Ymj2G/view?usp=sharing)

### Run below commands
.\venv\Scripts\activate
python manage.py runserver

Option 2 (manually install packages):

python -m venv venv
.\venv\Scripts\activate
python -m pip install --upgrade pip==22.3
pip install -r requirements.txt
pip install "modelscope[nlp]==0.4.7" -f https://modelscope.oss-cn-beijing.aliyuncs.com/releases/repo.html
pip install git+https://github.com/nateshmbhat/pyttsx3
python manage.py runserver