# Copyright 2024 Microsoft Research & University of Wisconsin-Madison 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.
"""PaliGemmamodel configuration"""

from huggingface_hub.dataclasses import strict

from ...configuration_utils import PreTrainedConfig, SubConfigSpec
from ...utils import auto_docstring
from ..auto import AutoConfig


@auto_docstring(checkpoint="google/paligemma-3b-pt-224")
@strict
class PaliGemmaConfig(PreTrainedConfig):
    r"""
    Example:

    ```python
    >>> from transformers import PaliGemmaForConditionalGeneration, PaliGemmaConfig, SiglipVisionConfig, GemmaConfig

    >>> # Initializing a Siglip-like vision config
    >>> vision_config = SiglipVisionConfig()

    >>> # Initializing a PaliGemma config
    >>> text_config = GemmaConfig()

    >>> # Initializing a PaliGemma paligemma-3b-224 style configuration
    >>> configuration = PaliGemmaConfig(vision_config, text_config)

    >>> # Initializing a model from the paligemma-3b-224 style configuration
    >>> model = PaliGemmaForConditionalGeneration(configuration)

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

    model_type = "paligemma"
    attribute_map = {"image_token_id": "image_token_index"}
    sub_configs_defaults = {
        "vision_config": SubConfigSpec(
            config_class=AutoConfig,
            model_type="siglip_vision_model",
            init_kwargs={
                "intermediate_size": 4096,
                "hidden_size": 1152,
                "patch_size": 14,
                "image_size": 224,
                "num_hidden_layers": 27,
                "num_attention_heads": 16,
                "vocab_size": 257152,
                "vision_use_head": False,
            },
        ),
        "text_config": SubConfigSpec(
            config_class=AutoConfig,
            model_type="gemma",
            init_kwargs={
                "hidden_size": 2048,
                "num_hidden_layers": 18,
                "intermediate_size": 16384,
                "num_attention_heads": 8,
                "num_key_value_heads": 1,
                "is_encoder_decoder": False,
                "vocab_size": 257152,
            },
        ),
    }
    keys_to_ignore_at_inference = ["past_key_values"]

    vision_config: dict | PreTrainedConfig | None = None
    text_config: dict | PreTrainedConfig | None = None
    image_token_index: int = 256000
    vocab_size: int = 257152
    projection_dim: int = 2048
    hidden_size: int = 2048
    tie_word_embeddings: bool = True

    def __post_init__(self, **kwargs):
        super().__post_init__(**kwargs)
        # BC: `use_bidirectional_attention` was originally unset in PaliGemma1 (backbone = Gemma1) AND PaliGemma2
        # (backbone = Gemma2). Both PaliGemmas want to default to True.
        if self.text_config.use_bidirectional_attention is None:
            self.text_config.use_bidirectional_attention = True

        self.text_config.num_image_tokens = (self.vision_config.image_size // self.vision_config.patch_size) ** 2
        self.vision_config.projection_dim = self.projection_dim


__all__ = ["PaliGemmaConfig"]