webapp/seed/Lib/site-packages/transformers/models/diffllama/configuration_diffllama.py

# Copyright 2024 weak-kajuma and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on Llama implementations in this library and Microsoft's
# Differential Transformer implementations.

# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""DiffLlama model configuration"""

from ...configuration_utils import PreTrainedConfig
from ...modeling_rope_utils import RopeParameters


class DiffLlamaConfig(PreTrainedConfig):
    r"""
    This is the configuration class to store the configuration of a [`DiffLlamaModel`]. It is used to instantiate an DiffLlama
    model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults
    will yield a similar configuration to that of the [kajuma/DiffLlama-0.3B-handcut](https://huggingface.co/kajuma/DiffLlama-0.3B-handcut).

    Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the
    documentation from [`PreTrainedConfig`] for more information.


    Args:
        vocab_size (`int`, *optional*, defaults to 32000):
            Vocabulary size of the DiffLlama model. Defines the number of different tokens that can be represented by the
            `inputs_ids` passed when calling [`DiffLlamaModel`]
        hidden_size (`int`, *optional*, defaults to 2048):
            Dimension of the hidden representations.
        intermediate_size (`int`, *optional*, defaults to 8192):
            Dimension of the MLP representations.
        num_hidden_layers (`int`, *optional*, defaults to 16):
            Number of hidden layers in the Transformer decoder.
        num_attention_heads (`int`, *optional*, defaults to 32):
            Number of attention heads for each attention layer in the Transformer decoder.
        num_key_value_heads (`int`, *optional*):
            This is the number of key_value heads that should be used to implement Grouped Query Attention. If
            `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
            `num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When
            converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
            by meanpooling all the original heads within that group. For more details, check out [this
            paper](https://huggingface.co/papers/2305.13245). If it is not specified, will default to
            `num_attention_heads`.
        hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
            The non-linear activation function (function or string) in the decoder.
        max_position_embeddings (`int`, *optional*, defaults to 2048):
            The maximum sequence length that this model might ever be used with.
        initializer_range (`float`, *optional*, defaults to 0.02):
            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
        rms_norm_eps (`float`, *optional*, defaults to 1e-05):
            The epsilon used by the rms normalization layers.
        use_cache (`bool`, *optional*, defaults to `True`):
            Whether or not the model should return the last key/values attentions (not used by all models). Only
            relevant if `config.is_decoder=True`.
        pad_token_id (`int`, *optional*):
            Padding token id.
        bos_token_id (`int`, *optional*, defaults to 1):
            Beginning of stream token id.
        eos_token_id (`int`, *optional*, defaults to 2):
            End of stream token id.
        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
            Whether to tie weight embeddings
        rope_parameters (`RopeParameters`, *optional*):
            Dictionary containing the configuration parameters for the RoPE embeddings. The dictionary should contain
            a value for `rope_theta` and optionally parameters used for scaling in case you want to use RoPE
            with longer `max_position_embeddings`.
        attention_bias (`bool`, *optional*, defaults to `False`):
            Whether to use a bias in the query, key, value and output projection layers during self-attention.
        attention_dropout (`float`, *optional*, defaults to 0.0):
            The dropout ratio for the attention probabilities.
        lambda_std_dev (`float`, *optional*, defaults to 0.1):
            The standard deviation for initialization of parameter lambda in attention layer.
        head_dim (`int`, *optional*):
            The attention head dimension. If None, it will default to hidden_size // num_heads

    ```python
    >>> from transformers import DiffLlamaModel, DiffLlamaConfig

    >>> # Initializing a DiffLlama diffllama-7b style configuration
    >>> configuration = DiffLlamaConfig()

    >>> # Initializing a model from the diffllama-7b style configuration
    >>> model = DiffLlamaModel(configuration)

    >>> # Accessing the model configuration
    >>> configuration = model.config
    ```"""

    model_type = "diffllama"
    keys_to_ignore_at_inference = ["past_key_values"]

    def __init__(
        self,
        vocab_size: int | None = 32000,
        hidden_size: int | None = 2048,
        intermediate_size: int | None = 8192,
        num_hidden_layers: int | None = 16,
        num_attention_heads: int | None = 32,
        num_key_value_heads: int | None = None,
        hidden_act: str | None = "silu",
        max_position_embeddings: int | None = 2048,
        initializer_range: float | None = 0.02,
        rms_norm_eps: int | None = 1e-5,
        use_cache: bool | None = True,
        pad_token_id: int | None = None,
        bos_token_id: int | None = 1,
        eos_token_id: int | None = 2,
        tie_word_embeddings: bool | None = False,
        rope_parameters: RopeParameters | dict[str, RopeParameters] | None = None,
        attention_bias: bool | None = False,
        attention_dropout: float | None = 0.0,
        lambda_std_dev: float | None = 0.1,
        head_dim: int | None = None,
        **kwargs,
    ):
        self.vocab_size = vocab_size
        self.max_position_embeddings = max_position_embeddings
        self.hidden_size = hidden_size
        self.intermediate_size = intermediate_size
        self.num_hidden_layers = num_hidden_layers
        self.num_attention_heads = num_attention_heads

        # for backward compatibility
        if num_key_value_heads is None:
            num_key_value_heads = num_attention_heads

        self.num_key_value_heads = num_key_value_heads
        self.hidden_act = hidden_act
        self.initializer_range = initializer_range
        self.rms_norm_eps = rms_norm_eps
        self.use_cache = use_cache
        self.attention_bias = attention_bias
        self.attention_dropout = attention_dropout
        self.lambda_std_dev = lambda_std_dev
        self.head_dim = head_dim if head_dim is not None else self.hidden_size // self.num_attention_heads
        self.rope_parameters = rope_parameters

        self.tie_word_embeddings = tie_word_embeddings
        self.pad_token_id = pad_token_id
        self.bos_token_id = bos_token_id
        self.eos_token_id = eos_token_id
        super().__init__(**kwargs)


__all__ = ["DiffLlamaConfig"]