from transformers import PretrainedConfig


class LightningConfig(PretrainedConfig):
    model_type = "lightning"

    def __init__(
        self,
        vocab_size=75000,
        d_model=256,
        nhead=4,
        num_layers=6,
        dropout=0.1,
        max_seq_len=200,
        **kwargs
    ):
        kwargs.setdefault("tie_word_embeddings", False)
        super().__init__(**kwargs)

        self.vocab_size = vocab_size
        self.d_model = d_model
        self.nhead = nhead
        self.num_layers = num_layers
        self.dropout = dropout
        self.max_seq_len = max_seq_len
        
        self.num_hidden_layers = num_layers
        self.num_attention_heads = nhead
        self.hidden_size = d_model