README

Required libraries:
    pytorch
    PIL
    Visdom
    easydict
    numpy
    torch-vision
    torchnet
	
For training: 
1) Edit your json file.
2) In your json file, set "mode" as train.
3) Set "train_data" as the directory of training data.

For real compression:
1) In your json file, set "mode" as test.
2) Set "run_mode" as encode.
3) Set "run_type" as "S" (sequential) or "P" for (parallel).
4) Set "decoder_method" as 1 (Wavefront parallel decoding method) or 2 (diagonal parallel decoding method).
5) Set "test_data" as the directory of test images.

For decoding operation:
1) In your json file, set "mode" as test.
2) Set "run_mode" as decode.
3) Set "run_type" as "S" (sequential) or "P" for (parallel). 
4) Set "bitstream_path" as the directory of decoded bitsream.
5) Set "H" and "W" as the height and width of the image to be decoded.
6) For parallel decoding, set "decoder_method" as 1 (Wavefront parallel decoding method) or 2 (diagonal parallel decoding method).

To run the code with adjusted settings:
    python main.py configs/your_json_file.json