U
    4Af3                     @   sn   d Z ddlZddlmZ ddlmZ ddlmZ ee	Z
G dd deZG d	d
 d
eZG dd deZdS )zKOSMOS-2 model configuration    N)Union   )PretrainedConfig)loggingc                       sV   e Zd ZdZdZdgZddddZd fdd	Zee	e
ejf ddddZ  ZS )Kosmos2TextConfigav  
    This is the configuration class to store the configuration of a [`Kosmos2TextModel`]. It is used to instantiate a
    KOSMOS-2 text decoder according to the specified arguments, defining the model architecture. Instantiating a
    configuration with the defaults will yield a similar configuration to that of the text decoder of the KOSMOS-2
    [microsoft/kosmos-2-patch14-224](https://huggingface.co/microsoft/kosmos-2-patch14-224) 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 65037):
            Vocabulary size of the Kosmos2 model. Defines the number of different tokens that can be represented by the
            `inputs_ids` passed when calling [`Kosmos2Model`].
        max_position_embeddings (`int`, *optional*, defaults to 2048):
            The maximum sequence length that this model might ever be used with. Typically set this to something large
            just in case (e.g., 512 or 1024 or 2048).
        embed_dim (`int`, *optional*, defaults to 2048):
            Dimensionality of the layers and the pooler layer.
        layers (`int`, *optional*, defaults to 24):
            Number of hidden layers in the Transformer encoder.
        ffn_dim (`int`, *optional*, defaults to 8192):
            Dimensionality of the "intermediate" (often named feed-forward) layer in the Transformer encoder.
        attention_heads (`int`, *optional*, defaults to 32):
            Number of attention heads for each attention layer in the Transformer encoder.
        activation_function (`str` or `function`, *optional*, defaults to `"gelu"`):
            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
            `"relu"`, `"silu"` and `"gelu_new"` are supported.
        dropout (`float`, *optional*, defaults to 0.1):
            The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
        attention_dropout (`float`, *optional*, defaults to 0.1):
            The dropout ratio for the attention probabilities.
        activation_dropout (`float`, *optional*, defaults to 0.0):
            The dropout ratio for activations inside the fully connected layer.
        layerdrop (`float`, *optional*, defaults to 0.0):
            The LayerDrop probability for the decoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556)
            for more details.
        layer_norm_eps (`float`, *optional*, defaults to 1e-5):
            The epsilon used by the layer normalization layers.
        init_std (`float`, *optional*, defaults to 0.02):
            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
        scale_embedding (`bool`, *optional*, defaults to `True`):
            Scale embeddings by diving by sqrt(embed_dim).
        use_cache (`bool`, *optional*, defaults to `True`):
            Whether or not the model should return the last key/values attentions (not used by all models).
    ```Zkosmos_2_text_modelZpast_key_valuesattention_heads	embed_dimlayers)num_attention_headshidden_sizenum_hidden_layers                gelu皙?        h㈵>{Gz?T   r      c                    sx   t  jf |||d| || _|| _|| _|| _|| _|| _|| _|| _	|	| _
|
| _|| _|| _|| _|| _|| _d S )N)pad_token_idbos_token_ideos_token_id)super__init__
vocab_sizemax_position_embeddingsr   r	   ffn_dimr   activation_functiondropoutattention_dropoutactivation_dropout	layerdroplayer_norm_epsinit_stdscale_embedding	use_cache)selfr   r   r   r	   r    r   r!   r"   r#   r$   r%   r&   r'   r(   r)   r   r   r   kwargs	__class__ U/tmp/pip-unpacked-wheel-zw5xktn0/transformers/models/kosmos2/configuration_kosmos2.pyr   R   s,    zKosmos2TextConfig.__init__r   pretrained_model_name_or_pathreturnc                 K   s~   |  | | j|f|\}}|ddkr2|d }d|krpt| drp|d | jkrptd|d  d| j d | j|f|S )N
model_typekosmos-2text_configYou are using a model of type   to instantiate a model of type N. This is not supported for all configurations of models and can yield errors.Z_set_token_in_kwargsZget_config_dictgethasattrr3   loggerwarning	from_dictclsr1   r+   Zconfig_dictr.   r.   r/   from_pretrained   s    
 z!Kosmos2TextConfig.from_pretrained)r   r   r   r   r   r   r   r   r   r   r   r   r   TTr   r   r   )__name__
__module____qualname____doc__r3   Zkeys_to_ignore_at_inferenceZattribute_mapr   classmethodr   strosPathLikerA   __classcell__r.   r.   r,   r/   r      s8   .                  -r   c                       sD   e Zd ZdZdZd fdd	Zeeee	j
f ddddZ  ZS )Kosmos2VisionConfiga	  
    This is the configuration class to store the configuration of a [`Kosmos2VisionModel`]. It is used to instantiate a
    KOSMOS-2 vision encoder according to the specified arguments, defining the model architecture. Instantiating a
    configuration with the defaults will yield a similar configuration to that of the vision encoder of the KOSMOS-2
    [microsoft/kosmos-2-patch14-224](https://huggingface.co/microsoft/kosmos-2-patch14-224) architecture.

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

    Args:
        hidden_size (`int`, *optional*, defaults to 1024):
            Dimensionality of the encoder layers and the pooler layer.
        intermediate_size (`int`, *optional*, defaults to 4096):
            Dimensionality of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder.
        num_hidden_layers (`int`, *optional*, defaults to 24):
            Number of hidden layers in the Transformer encoder.
        num_attention_heads (`int`, *optional*, defaults to 16):
            Number of attention heads for each attention layer in the Transformer encoder.
        num_channels (`int`, *optional*, defaults to 3):
            The number of input channels.
        image_size (`int`, *optional*, defaults to 224):
            The size (resolution) of each image.
        patch_size (`int`, *optional*, defaults to 14):
            The size (resolution) of each patch.
        hidden_act (`str` or `function`, *optional*, defaults to `"quick_gelu"`):
            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
            `"relu"`, `"selu"` and `"gelu_new"` `"quick_gelu"` are supported.
        layer_norm_eps (`float`, *optional*, defaults to 1e-5):
            The epsilon used by the layer normalization layers.
        attention_dropout (`float`, *optional*, defaults to 0.0):
            The dropout ratio for the attention probabilities.
        initializer_range (`float`, *optional*, defaults to 0.02):
            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
        initializer_factor (`float`, *optional*, defaults to 1):
            A factor for initializing all weight matrices (should be kept to 1, used internally for initialization
            testing).
    ```Zkosmos_2_vision_model      r      r         
quick_gelur   r   r         ?c                    sZ   t  jf | || _|| _|| _|| _|| _|| _|| _|| _	|| _
|
| _|	| _|| _d S )N)r   r   r   intermediate_sizer   r
   num_channels
patch_size
image_sizeinitializer_rangeinitializer_factorr#   r&   
hidden_act)r*   r   rS   r   r
   rT   rV   rU   rY   r&   r#   rW   rX   r+   r,   r.   r/   r      s    zKosmos2VisionConfig.__init__r   r0   c                 K   s~   |  | | j|f|\}}|ddkr2|d }d|krpt| drp|d | jkrptd|d  d| j d | j|f|S )Nr3   r4   vision_configr6   r7   r8   r9   r?   r.   r.   r/   rA      s    
 z#Kosmos2VisionConfig.from_pretrained)rL   rM   r   rN   r   rO   rP   rQ   r   r   r   rR   )rB   rC   rD   rE   r3   r   rF   r   rG   rH   rI   rA   rJ   r.   r.   r,   r/   rK      s"   &            rK   c                       s*   e Zd ZdZdZdZd fdd	Z  ZS )	Kosmos2Configat  
    This is the configuration class to store the configuration of a [`Kosmos2Model`]. It is used to instantiate a
    KOSMOS-2 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 KOSMOS-2
    [microsoft/kosmos-2-patch14-224](https://huggingface.co/microsoft/kosmos-2-patch14-224) architecture.

    Args:
        text_config (`dict`, *optional*):
            Dictionary of configuration options used to initialize [`Kosmos2TextConfig`].
        vision_config (`dict`, *optional*):
            Dictionary of configuration options used to initialize [`Kosmos2VisionConfig`].
        latent_query_num (`int`, *optional*, defaults to 64):
            The number of latent query tokens that represent the image features used in the text decoder component.
        kwargs (*optional*):
            Dictionary of keyword arguments.

    Example:

    ```python
    >>> from transformers import Kosmos2Config, Kosmos2Model

    >>> # Initializing a Kosmos-2 kosmos-2-patch14-224 style configuration
    >>> configuration = Kosmos2Config()

    >>> # Initializing a model (with random weights) from the kosmos-2-patch14-224 style configuration
    >>> model = Kosmos2Model(configuration)

    >>> # Accessing the model configuration
    >>> configuration = model.config
    ```r4   TN@   c                    s\   t  jf | |d kr$i }td |d kr:i }td tf || _tf || _|| _d S )NzR`text_config` is `None`. Initializing the `Kosmos2TextConfig` with default values.zV`vision_config` is `None`. Initializing the `Kosmos2VisionConfig` with default values.)	r   r   r<   infor   r5   rK   rZ   latent_query_num)r*   r5   rZ   r^   r+   r,   r.   r/   r     s    

zKosmos2Config.__init__)NNr\   )rB   rC   rD   rE   r3   Zis_compositionr   rJ   r.   r.   r,   r/   r[      s      r[   )rE   rH   typingr   Zconfiguration_utilsr   utilsr   Z
get_loggerrB   r<   r   rK   r[   r.   r.   r.   r/   <module>   s   
w[