Skip to content

Latest commit

 

History

History

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 

README.md

Training and evaluating

Training

python main.py

positional arguments:

keras_model_path      Path to keras deep-learning-models directory

optional arguments:

-h, --help            show this help message and exit
--data_file DATA_FILE
                    Path to HDF5 containing the data
--attributes_file ATTRIBUTES_FILE
                    Path to csv file containing the attributes
--nb_neighbors NB_NEIGHBORS
                    Number of nearest neighbors to compute VGG
                    representation
--alpha ALPHA         Interpolation coefficient
--weight_reverse_mapping WEIGHT_REVERSE_MAPPING
                    Weight of reverse mapping loss
--weight_total_variation WEIGHT_TOTAL_VARIATION
                    Weight of total variation loss
--normalize_w NORMALIZE_W
                    Whether to normalize w

Example:

python main.py /home/user/GitHub/deep-learning-models

Expected outputs:

  • Create a copy of the source image in figures
  • Saves the generated image every few iteration in figures