configuration_mistral.py

# coding=utf-8
# Copyright 2023 Mistral AI and the HuggingFace Inc. team. All rights reserved.
#
# 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.
""" Mistral model configuration"""

from transformers.configuration_utils import PretrainedConfig
from transformers.utils import logging


logger = logging.get_logger(__name__)

MISTRAL_PRETRAINED_CONFIG_ARCHIVE_MAP = {
    "mistralai/Mistral-7B-v0.1": "https://huggingface.co./mistralai/Mistral-7B-v0.1/resolve/main/config.json",
    "mistralai/Mistral-7B-Instruct-v0.1": "https://huggingface.co./mistralai/Mistral-7B-Instruct-v0.1/resolve/main/config.json",
}


class MistralConfig(PretrainedConfig):
    r"""
    This is the configuration class to store the configuration of a [`MistralModel`]. It is used to instantiate an
    Mistral 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 Mistral-7B-v0.1 or Mistral-7B-Instruct-v0.1.

    [mistralai/Mistral-7B-v0.1](https://huggingface.co./mistralai/Mistral-7B-v0.1)
    [mistralai/Mistral-7B-Instruct-v0.1](https://huggingface.co./mistralai/Mistral-7B-Instruct-v0.1)

    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 Mistral model. Defines the number of different tokens that can be represented by the
            `inputs_ids` passed when calling [`MistralModel`]
        hidden_size (`int`, *optional*, defaults to 4096):
            Dimension of the hidden representations.
        intermediate_size (`int`, *optional*, defaults to 14336):
            Dimension of the MLP representations.
        num_hidden_layers (`int`, *optional*, defaults to 32):
            Number of hidden layers in the Transformer encoder.
        num_attention_heads (`int`, *optional*, defaults to 32):
            Number of attention heads for each attention layer in the Transformer encoder.
        num_key_value_heads (`int`, *optional*, defaults to 8):
            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 checkout [this
            paper](https://arxiv.org/pdf/2305.13245.pdf). If it is not specified, will default to `8`.
        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 `4096*32`):
            The maximum sequence length that this model might ever be used with. Mistral's sliding window attention
            allows sequence of up to 4096*32 tokens.
        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-06):
            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*):
            The id of the padding token.
        bos_token_id (`int`, *optional*, defaults to 1):
            The id of the "beginning-of-sequence" token.
        eos_token_id (`int`, *optional*, defaults to 2):
            The id of the "end-of-sequence" token.
        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
            Whether the model's input and output word embeddings should be tied.
        rope_theta (`float`, *optional*, defaults to 10000.0):
            The base period of the RoPE embeddings.
        sliding_window (`int`, *optional*, defaults to 4096):
            Sliding window attention window size. If not specified, will default to `4096`.
        attention_dropout (`float`, *optional*, defaults to 0.0):
            The dropout ratio for the attention probabilities.

    ```python
    >>> from transformers import MistralModel, MistralConfig

    >>> # Initializing a Mistral 7B style configuration
    >>> configuration = MistralConfig()

    >>> # Initializing a model from the Mistral 7B style configuration
    >>> model = MistralModel(configuration)

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

    model_type = "mistral"
    keys_to_ignore_at_inference = ["past_key_values"]

    def __init__(
        self,
        vocab_size=32000,
        hidden_size=4096,
        intermediate_size=14336,
        num_hidden_layers=32,
        num_attention_heads=32,
        num_key_value_heads=8,
        hidden_act="silu",
        max_position_embeddings=4096 * 32,
        initializer_range=0.02,
        rms_norm_eps=1e-6,
        use_cache=True,
        pad_token_id=None,
        bos_token_id=1,
        eos_token_id=2,
        tie_word_embeddings=False,
        rope_theta=10000.0,
        sliding_window=4096,
        rope_scaling=None,
        attention_dropout=0.0,
        beacon_window=1024,
        beacon_stride=1024,
        beacon_attn="full-coverage",
        beacon_ratio=[2,4,8,16,32],
        beacon_ratio_mix="step-random",
        beacon_param=[],
        beacon_embed_init="eos",
        beacon_sink_size=0,
        beacon_attend_prev=True,
        beacon_pos="interleave",
        beacon_parallel_window=1,
        beacon_accum=True,
        **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
        self.sliding_window = sliding_window

        # 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.rope_theta = rope_theta
        self.attention_dropout = attention_dropout

        self.rope_scaling = rope_scaling
        self._rope_scaling_validation()

        self.beacon_window = beacon_window
        self.beacon_stride = beacon_stride
        self.beacon_attn = beacon_attn
        self.beacon_ratio = beacon_ratio
        self.beacon_ratio_mix = beacon_ratio_mix
        self.beacon_param = beacon_param
        self.beacon_embed_init = beacon_embed_init
        self.beacon_sink_size = beacon_sink_size
        self.beacon_attend_prev = beacon_attend_prev
        self.beacon_pos = beacon_pos
        self.beacon_parallel_window = beacon_parallel_window
        self.beacon_accum = beacon_accum

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

    def _rope_scaling_validation(self):
        """
        Validate the `rope_scaling` configuration.
        """
        if self.rope_scaling is None:
            return

        if not isinstance(self.rope_scaling, dict) or len(self.rope_scaling) != 2:
            raise ValueError(
                "`rope_scaling` must be a dictionary with with two fields, `type` and `factor`, "
                f"got {self.rope_scaling}"
            )
        rope_scaling_type = self.rope_scaling.get("type", None)
        rope_scaling_factor = self.rope_scaling.get("factor", None)
        if rope_scaling_type is None or rope_scaling_type not in ["linear", "dynamic"]:
            raise ValueError(
                f"`rope_scaling`'s type field must be one of ['linear', 'dynamic'], got {rope_scaling_type}"
            )
        if rope_scaling_factor is None or not isinstance(rope_scaling_factor, float) or rope_scaling_factor <= 1.0:
            raise ValueError(f"`rope_scaling`'s factor field must be a float > 1, got {rope_scaling_factor}")