venv/lib/python3.13/site-packages/transformers/models/aria/configuration_aria.py

#                🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
#           This file was automatically generated from src/transformers/models/aria/modular_aria.py.
#               Do NOT edit this file manually as any edits will be overwritten by the generation of
#             the file from the modular. If any change should be done, please apply the change to the
#                          modular_aria.py file directly. One of our CI enforces this.
#                🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
# coding=utf-8
# Copyright 2024 The Rhymes-AI Teams Authors 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.
from typing import Optional

from ...configuration_utils import PretrainedConfig
from ...modeling_rope_utils import rope_config_validation
from ..auto import CONFIG_MAPPING, AutoConfig


class AriaTextConfig(PretrainedConfig):
    r"""
    This class handles the configuration for the text component of the Aria model.
    Instantiating a configuration with the defaults will yield a similar configuration to that of the model of the Aria
    [rhymes-ai/Aria](https://huggingface.co/rhymes-ai/Aria) architecture.
    This class extends the LlamaConfig to include additional parameters specific to the Mixture of Experts (MoE) architecture.

    Args:
        vocab_size (`int`, *optional*, defaults to 32000):
            Vocabulary size of the LLaMA model. Defines the number of different tokens that can be represented by the
            `inputs_ids` passed when calling [`LlamaModel`]
        hidden_size (`int`, *optional*, defaults to 4096):
            Dimension of the hidden representations.
        intermediate_size (`int`, *optional*, defaults to 4096):
            The size of the MLP representations.
        num_hidden_layers (`int`, *optional*, defaults to 32):
            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. Llama 1 supports up to 2048 tokens,
            Llama 2 up to 4096, CodeLlama up to 16384.
        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*, defaults to 2):
            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.
        pretraining_tp (`int`, *optional*, defaults to 1):
            Experimental feature. Tensor parallelism rank used during pretraining. Please refer to [this
            document](https://huggingface.co/docs/transformers/main/perf_train_gpu_many#tensor-parallelism) to
            understand more about it. This value is necessary to ensure exact reproducibility of the pretraining
            results. Please refer to [this issue](https://github.com/pytorch/pytorch/issues/76232).
        tie_word_embeddings (`bool`, *optional*, defaults to `False`):
            Whether to tie weight embeddings
        rope_theta (`float`, *optional*, defaults to 10000.0):
            The base period of the RoPE embeddings.
        rope_scaling (`Dict`, *optional*):
            Dictionary containing the scaling configuration for the RoPE embeddings. NOTE: if you apply new rope type
            and you expect the model to work on longer `max_position_embeddings`, we recommend you to update this value
            accordingly.
            Expected contents:
                `rope_type` (`str`):
                    The sub-variant of RoPE to use. Can be one of ['default', 'linear', 'dynamic', 'yarn', 'longrope',
                    'llama3'], with 'default' being the original RoPE implementation.
                `factor` (`float`, *optional*):
                    Used with all rope types except 'default'. The scaling factor to apply to the RoPE embeddings. In
                    most scaling types, a `factor` of x will enable the model to handle sequences of length x *
                    original maximum pre-trained length.
                `original_max_position_embeddings` (`int`, *optional*):
                    Used with 'dynamic', 'longrope' and 'llama3'. The original max position embeddings used during
                    pretraining.
                `attention_factor` (`float`, *optional*):
                    Used with 'yarn' and 'longrope'. The scaling factor to be applied on the attention
                    computation. If unspecified, it defaults to value recommended by the implementation, using the
                    `factor` field to infer the suggested value.
                `beta_fast` (`float`, *optional*):
                    Only used with 'yarn'. Parameter to set the boundary for extrapolation (only) in the linear
                    ramp function. If unspecified, it defaults to 32.
                `beta_slow` (`float`, *optional*):
                    Only used with 'yarn'. Parameter to set the boundary for interpolation (only) in the linear
                    ramp function. If unspecified, it defaults to 1.
                `short_factor` (`list[float]`, *optional*):
                    Only used with 'longrope'. The scaling factor to be applied to short contexts (<
                    `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden
                    size divided by the number of attention heads divided by 2
                `long_factor` (`list[float]`, *optional*):
                    Only used with 'longrope'. The scaling factor to be applied to long contexts (<
                    `original_max_position_embeddings`). Must be a list of numbers with the same length as the hidden
                    size divided by the number of attention heads divided by 2
                `low_freq_factor` (`float`, *optional*):
                    Only used with 'llama3'. Scaling factor applied to low frequency components of the RoPE
                `high_freq_factor` (`float`, *optional*):
                    Only used with 'llama3'. Scaling factor applied to high frequency components of the RoPE
        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.
        mlp_bias (`bool`, *optional*, defaults to `False`):
            Whether to use a bias in up_proj, down_proj and gate_proj layers in the MLP layers.
        head_dim (`int`, *optional*):
            The attention head dimension. If None, it will default to hidden_size // num_heads
        moe_num_experts (`int`, *optional*, defaults to 8):
            The number of experts in the MoE layer.
        moe_topk (`int`, *optional*, defaults to 2):
            The number of top experts to route to for each token.
        moe_num_shared_experts (`int`, *optional*, defaults to 2):
            The number of shared experts.
    """

    model_type = "aria_text"
    keys_to_ignore_at_inference = ["past_key_values"]
    # Default tensor parallel plan for base model `AriaTextModel`
    base_model_tp_plan = {
        "layers.*.self_attn.q_proj": "colwise",
        "layers.*.self_attn.k_proj": "colwise",
        "layers.*.self_attn.v_proj": "colwise",
        "layers.*.self_attn.o_proj": "rowwise",
        "layers.*.mlp.gate_proj": "colwise",
        "layers.*.mlp.up_proj": "colwise",
        "layers.*.mlp.down_proj": "rowwise",
    }
    base_model_pp_plan = {
        "embed_tokens": (["input_ids"], ["inputs_embeds"]),
        "layers": (["hidden_states", "attention_mask"], ["hidden_states"]),
        "norm": (["hidden_states"], ["hidden_states"]),
    }
    base_config_key = "text_config"

    def __init__(
        self,
        vocab_size=32000,
        hidden_size=4096,
        intermediate_size: int = 4096,
        num_hidden_layers=32,
        num_attention_heads=32,
        num_key_value_heads=None,
        hidden_act="silu",
        max_position_embeddings=2048,
        initializer_range=0.02,
        rms_norm_eps=1e-6,
        use_cache=True,
        pad_token_id=2,
        bos_token_id=1,
        eos_token_id=2,
        pretraining_tp=1,
        tie_word_embeddings=False,
        rope_theta=10000.0,
        rope_scaling=None,
        attention_bias=False,
        attention_dropout=0.0,
        mlp_bias=False,
        head_dim=None,
        moe_num_experts: int = 8,
        moe_topk: int = 2,
        moe_num_shared_experts: int = 2,
        **kwargs,
    ):
        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,
        )
        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.pretraining_tp = pretraining_tp
        self.use_cache = use_cache
        self.rope_theta = rope_theta
        self.rope_scaling = rope_scaling
        self.attention_bias = attention_bias
        self.attention_dropout = attention_dropout
        self.mlp_bias = mlp_bias
        self.head_dim = head_dim if head_dim is not None else self.hidden_size // self.num_attention_heads
        # Validate the correctness of rotary position embeddings parameters
        # BC: if there is a 'type' field, copy it it to 'rope_type'.
        if self.rope_scaling is not None and "type" in self.rope_scaling:
            self.rope_scaling["rope_type"] = self.rope_scaling["type"]
        rope_config_validation(self)
        self.moe_num_experts = moe_num_experts
        self.moe_topk = moe_topk
        self.moe_num_shared_experts = moe_num_shared_experts


class AriaConfig(PretrainedConfig):
    r"""
    This class handles the configuration for both vision and text components of the Aria model,
    as well as additional parameters for image token handling and projector mapping.
    Instantiating a configuration with the defaults will yield a similar configuration to that of the model of the Aria
    [rhymes-ai/Aria](https://huggingface.co/rhymes-ai/Aria) architecture.

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

    Args:
        vision_config (`AriaVisionConfig` or `dict`, *optional*):
            Configuration for the vision component.
        vision_feature_layer (`int`, *optional*, defaults to -1):
            The index of the layer to select the vision feature.
        text_config (`AriaTextConfig` or `dict`, *optional*):
            Configuration for the text component.
        projector_patch_to_query_dict (`dict`, *optional*):
            Mapping of patch sizes to query dimensions.
        image_token_index (`int`, *optional*, defaults to 9):
            Index used to represent image tokens.
        initializer_range (`float`, *optional*, defaults to 0.02):
            The standard deviation of the truncated normal initializer for initializing all weight matrices.

    Attributes:
        model_type (`str`):
            Type of the model, set to `"aria"`.
        image_token_index (`int`):
            Index used to represent image tokens.
        projector_patch_to_query_dict (`dict`):
            Mapping of patch sizes to query dimensions.
        vision_config (`AriaVisionConfig`):
            Configuration for the vision component.
        text_config (`AriaTextConfig`):
            Configuration for the text component.
    """

    model_type = "aria"
    attribute_map = {
        "image_token_id": "image_token_index",
    }
    sub_configs = {"text_config": AriaTextConfig, "vision_config": AutoConfig}

    def __init__(
        self,
        vision_config=None,
        vision_feature_layer: int = -1,
        text_config: AriaTextConfig = None,
        projector_patch_to_query_dict: Optional[dict] = None,
        image_token_index: int = 9,
        initializer_range: float = 0.02,
        **kwargs,
    ):
        self.image_token_index = image_token_index

        # Convert the keys and values of projector_patch_to_query_dict to integers
        # This ensures consistency even if they were provided as strings
        if projector_patch_to_query_dict is None:
            projector_patch_to_query_dict = {
                1225: 128,
                4900: 256,
            }
        self.projector_patch_to_query_dict = {int(k): int(v) for k, v in projector_patch_to_query_dict.items()}
        self.max_value_projector_patch_to_query_dict = max(self.projector_patch_to_query_dict.values())
        self.vision_feature_layer = vision_feature_layer
        if isinstance(vision_config, dict):
            vision_config["model_type"] = "idefics3_vision"
            vision_config = CONFIG_MAPPING[vision_config["model_type"]](**vision_config)
        elif vision_config is None:
            vision_config = CONFIG_MAPPING["idefics3_vision"]()

        self.vision_config = vision_config
        self.initializer_range = initializer_range

        if isinstance(text_config, dict) and "model_type" in text_config:
            text_config = AriaTextConfig(**text_config)
        elif text_config is None:
            text_config = AriaTextConfig()

        self.text_config = text_config

        super().__init__(**kwargs)


__all__ = ["AriaConfig", "AriaTextConfig"]