from transformers import BertConfig, BertForMaskedLM

config = BertConfig(
    hidden_size = 384,
    vocab_size= tokenizer.vocab_size,
    num_hidden_layers = 6,
    num_attention_heads = 6,
    intermediate_size = 1024,
    max_position_embeddings = 256
)

model = BertForMaskedLM(config=config)
print(model.num_parameters()) #10457864