Training Script

I modified the Coqui-TTS-VITS training script as follows to handle the luxembourgish dataset without using phonemes :

import os

from trainer import Trainer, TrainerArgs

from TTS.tts.configs.shared_configs import BaseDatasetConfig
from TTS.tts.configs.vits_config import VitsConfig
from TTS.tts.datasets import load_tts_samples
from TTS.tts.models.vits import CharactersConfig, Vits, VitsArgs, VitsAudioConfig
from TTS.tts.utils.text.tokenizer import TTSTokenizer
from TTS.utils.audio import AudioProcessor

output_path = "E:\\TTS-for-LOD\\TTS\\dataset\\"
dataset_config = BaseDatasetConfig(
    formatter="luxembourgish", meta_file_train="metadata.csv", path=os.path.join(output_path, "kuborn-lb/")
)

audio_config = VitsAudioConfig(
    sample_rate=22050,
    win_length=1024,
    hop_length=256,
    num_mels=80,
    mel_fmin=0,
    mel_fmax=None,
)

config = VitsConfig(
    audio=audio_config,
    run_name="vits_max_kuborn_lb",
    batch_size=32,
    eval_batch_size=16,
    batch_group_size=5,
    num_loader_workers=0,
    num_eval_loader_workers=4,
    run_eval=True,
    test_delay_epochs=-1,
    epochs=1000,
    text_cleaner="phoneme_cleaners",
    use_phonemes=False,
    compute_input_seq_cache=True,
    print_step=25,
    print_eval=True,
    mixed_precision=True,
    characters=CharactersConfig(
        characters_class="TTS.tts.models.vits.VitsCharacters",
        pad="<PAD>",
        eos="<EOS>",
        bos="<BOS>",
        blank="<BLNK>",
        characters="abcdefghijklmnopqrstuvwxyzµßàáâäåæçèéêëìíîïñòóôöùúûüąćęłńœśşźżƒабвгдежзийклмнопрстуфхцчшщъыьэюяёєіїґӧ«°±µ»$%&‘’‚“`”„õãı",
        punctuations="!¡'(),-.:;¿? ",
        phonemes=None,
    ),
    test_sentences=[
        "an der zäit hunn sech den nordwand an d'sonn gestridden, wie vun hinnen zwee wuel méi staark wier,",
        "wéi e wanderer, deen an ee waarme mantel agepak war, iwwert de wee koum.",
        "si goufen sech eens, datt deejéinege fir dee stäerkste gëlle sollt, deen de wanderer forcéiere géif, säi mantel auszedoen.",
        "dunn huet d'sonn d'loft mat hire frëndleche strale gewiermt, a schonn no kuerzer zäit huet de wanderer säi mantel ausgedoen.",
        "do huet den nordwand missen zouginn, datt d'sonn vun hinnen zwee dee stäerkste wier.",
    ],
    output_path=output_path,
    datasets=[dataset_config],
)

# INITIALIZE THE AUDIO PROCESSOR
# Audio processor is used for feature extraction and audio I/O.
# It mainly serves to the dataloader and the training loggers.
ap = AudioProcessor.init_from_config(config)

# INITIALIZE THE TOKENIZER
# Tokenizer is used to convert text to sequences of token IDs.
# config is updated with the default characters if not defined in the config.
tokenizer, config = TTSTokenizer.init_from_config(config)

# LOAD DATA SAMPLES
# Each sample is a list of ```[text, audio_file_path, speaker_name]```
# You can define your custom sample loader returning the list of samples.
# Or define your custom formatter and pass it to the `load_tts_samples`.
# Check `TTS.tts.datasets.load_tts_samples` for more details.
train_samples, eval_samples = load_tts_samples(
    dataset_config,
    eval_split=True,
    eval_split_max_size=config.eval_split_max_size,
    eval_split_size=config.eval_split_size,
)

# init model
model = Vits(config, ap, tokenizer, speaker_manager=None)

# init the trainer and 🚀
trainer = Trainer(
    TrainerArgs(),
    config,
    output_path,
    model=model,
    train_samples=train_samples,
    eval_samples=eval_samples,
)
if __name__ == '__main__':
    trainer.fit()

For simplification I copied this script with the name train_vits_max_kuborn.pyat the root level of my forked Coqui-TTS repository.