This is the accompanying python script for master thesis at Brno university of technology, Faculty of information technologies that includes a utility for running specific models with different settings.
This project requires Python 3.8. After installing Python, you can install the project's dependencies using pip
.
Here are the step by step instructions to setup the project:
-
Clone the repository:
git clone https://github.com/TomasLapsansky/Master-thesis.git cd Master-thesis
-
Create a virtual environment (optional but recommended):
python3 -m venv env source env/bin/activate
-
Install the requirements:
pip install -r requirements.txt
After setting up the project, you can run the script with different command-line arguments.
Here's the list of the arguments:
-m, --training_model
: Model name (case sensitive). This is a required argument.-d, --dataset
: Dataset name (case sensitive).-e, --eval
: Set if execute evaluation.-r, --dropout
: Set dropout rate. Default is 0.5.-t, --trained
: Use pre-trained model. Default is False.-f, --frozen
: Freeze layers of base model until a specified layer.--type
: Set type of efficient net.--lr, --learning_rate
: Set learning rate for model. Default is 0.0001.-c, --checkpoint
: Path to loaded checkpoint.-p, --print
: Path to image.
Here is an example of how to run the script:
python main.py -m efficientdet --type L -d celeb-df -t
The repository also contains a preprocessing folder, where the scripts for dataset processing are located.
This project is licensed under the terms of the MIT license.