# coding=utf-8
# TODO: Add license
#
# 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.
""" PagnolXl configuration"""
from transformers.configuration_utils import PretrainedConfig
from transformers.utils import logging

logger = logging.get_logger(__name__)


class PagnolXlConfig(PretrainedConfig):
    r"""
    This is the configuration class to store the configuration of a [`PagnolXlModel`]. It is used to instantiate a PagnolXl
    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 [PagnolXl]() architecture.

    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 65024):
            Vocabulary size of the PagnolXl model. Defines the number of different tokens that can be represented by the
            `inputs_ids` passed when calling [`PagnolXlModel`]
        d_model (`int`, *optional*, defaults to 4544):
            Dimension of the hidden representations.
        num_hidden_layers (`int`, *optional*, defaults to 32):
            Number of hidden layers in the Transformer decoder.
        n_heads (`int`, *optional*, defaults to 71):
            Number of attention heads for each attention layer in the Transformer encoder.
        sigma (`float`, *optional*, defaults to 0.02):
            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
        use_cache (`bool`, *optional*, defaults to `True`):
            Whether the model should return the last key/values attentions (not used by all models). Only relevant if
            `config.is_decoder=True`.
        layer_norm_epsilon (`float`, *optional*, defaults to 1e-5):
            The epsilon used by the layer normalization layers.
        dropout (`float`, *optional*, defaults to 0.0):
            The dropout probability for MLP layers.
        bos_token_id (`int`, *optional*, defaults to 11):
            The id of the "beginning-of-sequence" token.
        eos_token_id (`int`, *optional*, defaults to 11):
            The id of the "end-of-sequence" token.

    Example:

    ```python
    >>> from transformers import PagnolXlModel, PagnolXlConfig

    >>> # Initializing a small (2-layer) PagnolXl configuration
    >>> configuration = PagnolXlConfig(num_hidden_layers=2)

    >>> # Initializing a model from the small configuration
    >>> model = PagnolXlModel(configuration)

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

    model_type = "pagnolxl"
    keys_to_ignore_at_inference = ["past_key_values"]

    def __init__(
        self,
        vocab_size=65024,
        activation_function="gelu",
        d_model=4544,
        d_feedforward=18176,
        n_heads=71,
        n_layers=32,
        layer_norm_epsilon=1e-5,
        sigma=0.02,
        use_cache=True,
        dropout=0.0,
        bos_token_id=11,
        eos_token_id=11,
        **kwargs,
    ):
        self.vocab_size = vocab_size
        # Backward compatibility with n_embed kwarg
        n_embed = kwargs.pop("n_embed", None)
        self.activation_function = activation_function
        self.d_model = d_model if n_embed is None else n_embed
        self.d_feedforward = d_feedforward
        self.n_heads = n_heads
        self.n_layers = n_layers
        self.layer_norm_epsilon = layer_norm_epsilon
        self.sigma = sigma
        self.use_cache = use_cache
        self.dropout = dropout
        self.bos_token_id = bos_token_id
        self.eos_token_id = eos_token_id

        super().__init__(bos_token_id=bos_token_id, eos_token_id=eos_token_id, **kwargs)

    @property
    def head_dim(self):
        return self.d_model // self.n_heads