config.py

import os
from argparse import ArgumentParser, ArgumentTypeError

# Please configure the path of your downloaded datasets
dataset2path = {
    "Opencpop": r"D:\math\NLP\project\dataset\Opencpop",
    "M4Singer": r"D:\math\NLP\project\dataset\M4Singer",
}

# Please configure the root path to save your data and model
root_path = r"D:\math\NLP\project\Singing-Voice-Conversion-BingliangLi"
data_path = os.path.join(root_path, "preprocess")
model_path = os.path.join(root_path, "model")

# Wav files path
dataset2wavpath = {
    "Opencpop": os.path.join(dataset2path["Opencpop"], "segments/wavs"),
    "M4Singer": dataset2path["M4Singer"],
}

# We select 5 utterances randomly for every singer
NUMS_OF_SINGER = 5

# WORLD hyparameters
WORLD_SAMPLE_RATE = 44100
WORLD_FRAME_SHIFT = 10
MCEP_DIM = 40

# Whisper hyparameters
WHISPER_SEQ = 1500
WHISPER_DIM = 1024


def str2bool(v):
    if isinstance(v, bool):
        return v
    if v.lower() in ("yes", "true", "t", "y", "1"):
        return True
    elif v.lower() in ("no", "false", "f", "n", "0"):
        return False
    else:
        raise ArgumentTypeError("Boolean value expected.")


parser = ArgumentParser(description="Acoustic Mapping")

# ======================== Dataset ========================

parser.add_argument("--dataset", type=str, default="Opencpop")
parser.add_argument("--converse", type=str2bool, default=False)
parser.add_argument("--whisper_dim", type=int, default=WHISPER_DIM)
parser.add_argument("--output_dim", type=int, default=MCEP_DIM)
parser.add_argument(
    "--save", type=str, default="ckpts/debug", help="folder to save the final model"
)

# ======================== Accoutic Models ========================
parser.add_argument("--model", type=str, default="Transformer")

parser.add_argument("--transformer_input_length", type=int, default=800)
parser.add_argument("--transformer_dropout", type=float, default=0.1)
parser.add_argument("--transformer_d_model", type=int, default=768)
parser.add_argument("--transformer_nhead", type=int, default=8)
parser.add_argument("--transformer_nlayers", type=int, default=6)

# ======================== Training ========================

parser.add_argument("--lr", type=float, default=1e-4, help="initial learning rate")
parser.add_argument("--epochs", type=int, default=500, help="upper epoch limit")
parser.add_argument(
    "--batch_size", type=int, default=32, metavar="N", help="batch size"
)
parser.add_argument(
    "--start_epoch", type=int, default=0, help="No. of the epoch to start training"
)
parser.add_argument("--resume", type=str, default="", help="path to load trained model")
parser.add_argument(
    "--evaluate", type=str2bool, default=False, help="only use for evaluating"
)
parser.add_argument("--debug", type=str2bool, default=False)

# ======================== Devices ========================

parser.add_argument("--seed", type=int, default=9, help="random seed")
parser.add_argument("--device", default="cpu", help="device to use for training")