params.json

{
    "audio":
    {
        "dataset_file": "NI_train.pkl",
        "dataset_test_file": "NI_valid.pkl",
        "N_FFT": 1024,
        "HOP_LENGTH": 256,
        "SAMPLING_RATE": 16000,
        "MELSPEC_BANDS": 128,
        "USE_SPEC": true,
        "USE_DELTA": false,
        "USE_DELTADELTA": false,
        "USE_MFCC": false,
        "N_MFCC": 18,
        "USE_MFCC_DELTA": false,
        "USE_MFCC_DELTADELTA": false,
        "USE_FLUCT": false,
        "MASK_FLUCT": false,
        "SPECTRAL_CONTRACTION": false,
        "SC_THRESHOLD": 0.1,
        "USE_SC": false,
        "SPECTRAL_FLATNESS": false,
        "SF_THRESHOLD": 0.6,
        "USE_SF": false,
        "USE_HPSS": false
    },
    "model":
    {
        "sample_sec": 2,
        "sample_overlap": 0,
        "offset_initial": 0,
        "offset_final": 0,
        "single_slice_audio": true,
        "dim_latent": 64,
        "cells_hidden": 512,
        "conv_channels": [32, 64, 32, 32],
        "max_pooling": [[4, 4], [2, 2], [2, 4], [2, 2]],
        "deconv_shape": [[1, 15, 14, 32],
                        [1, 31, 31, 64],
                        [1, 63, 62, 32],
                        [1, 128, 125, 1]],
        "iaf_flow_length": 5,
        "dim_autoregressive_nl": 64,
        "initial_s_offset": 1.0,
        "predictor_units": [],
        "dropout_keep_prob": 0.75,
        "rnn_decoder": false,
        "rnn_highway": false,
        "feature_normalization": false
    }
}