To split the dataset into Train and Test run the python/split_train_test.py in your terminal after selecting the desired directory.
๐ฏ Install requirements ๐ฎ
pip install -r requirements.txt
๐ช Files Overview ๐ฅ
food-101 > images - Format of the Food-101 dataset and how to be splitted into Train and Test
> meta
> test
> train
models > EfficientNetV2L > assets
> variables
> EfficiencyNetV2L.hdf5
> EfficiencyNetV2L.log
> kears_metadata.pb
> saved_model.pb
> EfficientNetV2S > assets
> variables
> EfficiencyNetV2S.hdf5
> EfficiencyNetV2S.log
> kears_metadata.pb
> saved_model.pb
> EfficientNetV2S_25Epochs > EfficiencyNetV2S.hdf5
> EfficiencyNetV2S.log
python > evaluate.py - Evaluate the model on the test set
> main.py - Main script to run the model
> models.py - Models definition + Fine Tuning
> split_train_test.py - Create the data folders in DSRI persistent folder
> train.py - Train the model
> visualization.py - Visualize the model output
readme_images - Images used in the README.md
test_images - images used for testing the model
vision_transformer > vit_cifar100.py - Vision Transformer model definition
> vit.py - Vision Transformer model definition
DeepFood_Food101.ipynb - Code Notebook with Models and Data
EfficientNetV2_Evaluation.ipynb - Code Notebook with Models and Data Evaluation
logs_analysis.ipynb - Code Notebook with Logs Analysis
model_predictions.ipynb - Code Notebook with Model Predictions
README.md - README
requirements.txt - Requirements for the repository