configuration_decilm.py

from packaging import version
import transformers
if version.parse(transformers.__version__) < version.parse("4.31.0"):
    raise ImportError(
        f"You are using transformers=={transformers.__version__}, but transformers>=4.31.0 is required to use DeciLM. Please upgrade transformers."
    )
from transformers.models.llama.configuration_llama import LlamaConfig
from transformers.utils import logging


logger = logging.get_logger(__name__)

LLAMA_PRETRAINED_CONFIG_ARCHIVE_MAP = {}


class DeciLMConfig(LlamaConfig):
    r"""

    Args:
        num_key_value_heads_per_layer (`List[int]`):
            The number of key-value heads per layer.
        naive_attention_prefill (`bool`, *optional*, defaults to False):
            Whether to use naive matmul or scaled dot product attention during prefill.
        naive_attention_decode_batched (`bool`, *optional*, defaults to True):
            Whether to use naive matmul or scaled dot product attention during decode for batch_size > 1.
        naive_attention_decode_single (`bool`, *optional*, defaults to False):
            Whether to use naive matmul or scaled dot product attention during decode for batch_size == 1.
       

       ```"""
    keys_to_ignore_at_inference = ["past_key_values"]

    def __init__(
        self,
        num_key_value_heads_per_layer: list = None,
        naive_attention_prefill: bool = False,
        naive_attention_decode_batched: bool = False,
        naive_attention_decode_single: bool = False,
        **kwargs,
    ):
        self.num_key_value_heads_per_layer = num_key_value_heads_per_layer
        self.naive_attention_prefill = naive_attention_prefill
        self.naive_attention_decode_batched = naive_attention_decode_batched
        self.naive_attention_decode_single = naive_attention_decode_single
        super().__init__(**kwargs, )