from transformers import PretrainedConfig


class YorubaCFMConfig(PretrainedConfig):
    model_type = "yoruba_cfm_dit"

    def __init__(
        self,
        vocab_size=67,
        latent_dim=128,
        d_model=512,
        n_heads=8,
        n_layers=10,
        text_layers=4,
        max_len=2048,
        text_max_positions=1024,
        dropout=0.1,
        pad_token_id=0,
        bos_token_id=1,
        eos_token_id=2,
        unk_token_id=3,
        sample_rate=24000,
        num_ode_steps=24,
        default_target_length=150,
        encodec_model_id="facebook/encodec_24khz",
        **kwargs,
    ):
        super().__init__(pad_token_id=pad_token_id, **kwargs)
        self.vocab_size = vocab_size
        self.latent_dim = latent_dim
        self.d_model = d_model
        self.n_heads = n_heads
        self.n_layers = n_layers
        self.text_layers = text_layers
        self.max_len = max_len
        self.text_max_positions = text_max_positions
        self.dropout = dropout
        self.bos_token_id = bos_token_id
        self.eos_token_id = eos_token_id
        self.unk_token_id = unk_token_id
        self.sample_rate = sample_rate
        self.num_ode_steps = num_ode_steps
        self.default_target_length = default_target_length
        self.encodec_model_id = encodec_model_id