# coding=utf-8
# Copyright 2025 the Cohere 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.
"""PyTorch AyaVision model."""

from typing import List, Optional, Tuple, Union

import torch
from torch import nn

from transformers.models.llava.modeling_llava import (
    LlavaCausalLMOutputWithPast,
    LlavaForConditionalGeneration,
    LlavaPreTrainedModel,
)

from ...activations import ACT2FN
from ...utils import (
    add_start_docstrings,
    logging,
)
from .configuration_aya_vision import AyaVisionConfig


logger = logging.get_logger(__name__)

_CONFIG_FOR_DOC = "AyaVisionConfig"


class AyaVisionMultiModalProjector(nn.Module):
    def __init__(self, config: AyaVisionConfig):
        super().__init__()
        self.config = config
        self.downsample_factor = config.downsample_factor
        self.alignment_intermediate_size = getattr(
            config, "alignment_intermediate_size", config.text_config.hidden_size
        )
        self.layernorm = nn.LayerNorm(
            config.vision_config.hidden_size * (config.downsample_factor**2), eps=config.adapter_layer_norm_eps
        )

        self.linear_1 = nn.Linear(
            config.vision_config.hidden_size * (config.downsample_factor**2),
            self.alignment_intermediate_size,
            bias=True,
        )

        self.act = ACT2FN["silu"]  # SwiGLU uses SiLU activation
        # For SwiGLU, project down to half size since we split intermediate dim
        self.linear_2 = nn.Linear(self.alignment_intermediate_size // 2, config.text_config.hidden_size, bias=True)

    def forward(self, image_features):
        image_features = self.pixel_shuffle(image_features)
        image_features = self.layernorm(image_features)
        hidden_states = self.linear_1(image_features)

        # Split along last dimension and apply SwiGLU
        x, gate = hidden_states.chunk(2, dim=-1)
        hidden_states = self.act(gate) * x

        hidden_states = self.linear_2(hidden_states)
        return hidden_states

    def pixel_shuffle(self, image_features):  # B, S, D
        batch_size, seq_length, feature_dim = image_features.shape
        height = width = int(seq_length**0.5)
        image_features = image_features.reshape(image_features.shape[0], width, height, -1)
        channels = image_features.shape[-1]
        image_features = image_features.reshape(
            batch_size, width, int(height / self.downsample_factor), int(channels * self.downsample_factor)
        )
        image_features = image_features.permute(0, 2, 1, 3)
        image_features = image_features.reshape(
            batch_size, int(height / self.downsample_factor), int(width / self.downsample_factor), -1
        )
        image_features = image_features.permute(0, 2, 1, 3)
        return image_features


AYA_VISION_START_DOCSTRING = r"""
    This model inherits from [`PreTrainedModel`]. Check the superclass documentation for the generic methods the
    library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
    etc.)

    This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
    Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
    and behavior.

    Parameters:
        config ([`AyaVisionConfig`] or [`AyaVisionVisionConfig`]):
            Model configuration class with all the parameters of the model. Initializing with a config file does not
            load the weights associated with the model, only the configuration. Check out the
            [`~PreTrainedModel.from_pretrained`] method to load the model weights.
"""


@add_start_docstrings(
    "The bare Aya Vision Model outputting raw hidden-states without any specific head on top.",
    AYA_VISION_START_DOCSTRING,
)
class AyaVisionPreTrainedModel(LlavaPreTrainedModel):
    _supports_quantized_cache = False
    _supports_static_cache = False


class AyaVisionCausalLMOutputWithPast(LlavaCausalLMOutputWithPast):
    pass


AYA_VISION_INPUTS_DOCSTRING = """
    Args:
        input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`):
            Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide
            it.

            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
            [`PreTrainedTokenizer.__call__`] for details.

            [What are input IDs?](../glossary#input-ids)
        pixel_values (`torch.FloatTensor` of shape `(batch_size, num_channels, image_size, image_size)):
            The tensors corresponding to the input images. Pixel values can be obtained using
            [`AutoImageProcessor`]. See [`GotOcr2ImageProcessor.__call__`] for details. [`AyaVisionProcessor`] uses
            [`GotOcr2ImageProcessor`] for processing images.
        attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*):
            Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:

            - 1 for tokens that are **not masked**,
            - 0 for tokens that are **masked**.

            [What are attention masks?](../glossary#attention-mask)

            Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and
            [`PreTrainedTokenizer.__call__`] for details.

            If `past_key_values` is used, optionally only the last `decoder_input_ids` have to be input (see
            `past_key_values`).

            If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`]
            and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more
            information on the default strategy.

            - 1 indicates the head is **not masked**,
            - 0 indicates the head is **masked**.
        position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
            Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0,
            config.n_positions - 1]`. [What are position IDs?](../glossary#position-ids)
        past_key_values (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`):
            Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape
            `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape
            `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`.

            Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention
            blocks) that can be used (see `past_key_values` input) to speed up sequential decoding.

            If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that
            don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all
            `decoder_input_ids` of shape `(batch_size, sequence_length)`.
        inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
            Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This
            is useful if you want more control over how to convert `input_ids` indices into associated vectors than the
            model's internal embedding lookup matrix.
        vision_feature_layer (`Union[int, List[int]], *optional*, defaults to -2`):
            The index of the layer to select the vision feature. If multiple indices are provided,
            the vision feature of the corresponding indices will be concatenated to form the
            vision features.
        vision_feature_select_strategy (`str`, *optional*, defaults to `"default"`):
            The feature selection strategy used to select the vision feature from the vision backbone.
            Can be one of `"default"` or `"full"`.
        use_cache (`bool`, *optional*):
            If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see
            `past_key_values`).
        output_attentions (`bool`, *optional*):
            Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned
            tensors for more detail.
        output_hidden_states (`bool`, *optional*):
            Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for
            more detail.
        return_dict (`bool`, *optional*):
            Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple.
        cache_position (`torch.LongTensor` of shape `(sequence_length)`, *optional*):
            Indices depicting the position of the input sequence tokens in the sequence. Contrarily to `position_ids`,
            this tensor is not affected by padding. It is used to update the cache in the correct position and to infer
            the complete sequence length.
"""


@add_start_docstrings(
    """The AyaVision model which consists of a vision backbone and a language model.""",
    AYA_VISION_START_DOCSTRING,
)
class AyaVisionForConditionalGeneration(LlavaForConditionalGeneration):
    def tie_weights(self):
        return self.language_model.tie_weights()

    def resize_token_embeddings(self, new_num_tokens: Optional[int] = None, pad_to_multiple_of=None) -> nn.Embedding:
        model_embeds = self.language_model.resize_token_embeddings(new_num_tokens, pad_to_multiple_of)
        # update vocab size
        self.config.text_config.vocab_size = model_embeds.num_embeddings
        self.vocab_size = model_embeds.num_embeddings
        return model_embeds

    def forward(
        self,
        input_ids: Optional[torch.LongTensor] = None,
        pixel_values: Optional[torch.FloatTensor] = None,
        attention_mask: Optional[torch.Tensor] = None,
        position_ids: Optional[torch.LongTensor] = None,
        past_key_values: Optional[List[torch.FloatTensor]] = None,
        inputs_embeds: Optional[torch.FloatTensor] = None,
        vision_feature_layer: Optional[Union[int, List[int]]] = None,
        vision_feature_select_strategy: Optional[str] = None,
        labels: Optional[torch.LongTensor] = None,
        use_cache: Optional[bool] = None,
        output_attentions: Optional[bool] = None,
        output_hidden_states: Optional[bool] = None,
        return_dict: Optional[bool] = None,
        cache_position: Optional[torch.LongTensor] = None,
        logits_to_keep: Union[int, torch.Tensor] = 0,
        image_sizes: Optional[torch.Tensor] = None,
        **lm_kwargs,
    ) -> Union[Tuple, AyaVisionCausalLMOutputWithPast]:
        r"""
            labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*):
                Labels for computing the masked language modeling loss. Indices should either be in `[0, ...,
                config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored
                (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.

            logits_to_keep (`int` or `torch.Tensor`, *optional*):
                If an `int`, compute logits for the last `logits_to_keep` tokens. If `0`, calculate logits for all
                `input_ids` (special case). Only last token logits are needed for generation, and calculating them only for that
                token can save memory, which becomes pretty significant for long sequences or large vocabulary size.
                If a `torch.Tensor`, must be 1D corresponding to the indices to keep in the sequence length dimension.
                This is useful when using packed tensor format (single dimension for batch and sequence length).


        Returns:

        Example:

        ```python
        >>> from transformers import AutoProcessor, AyaVisionForConditionalGeneration
        >>> import torch

        >>> torch_device = "cuda:0"
        >>> processor = AutoProcessor.from_pretrained("CohereForAI/aya-vision-8b", use_fast=True)
        >>> model = AyaVisionForConditionalGeneration.from_pretrained("CohereForAI/aya-vision-8b", device_map=torch_device)

        >>> messages = [
        ...     {
        ...         "role": "user",
        ...         "content": [
        ...             {
        ...                 "type": "image",
        ...                 "url": "https://pbs.twimg.com/media/Fx7YvfQWYAIp6rZ?format=jpg&name=medium",
        ...             },
        ...             {"type": "text", "text": "चित्र में लिखा पाठ क्या कहता है?"},
        ...         ],
        ...     }
        ... ]

        >>> inputs = processor.apply_chat_template(
        ...     messages, padding=True, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", device=torch_device
        ... ).to(model.device)

        >>> gen_tokens = model.generate(**inputs, max_new_tokens=300, do_sample=True, temperature=0.3)
        >>> processor.tokenizer.decode(gen_tokens[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)
        ```"""
        super().forward(
            input_ids=input_ids,
            pixel_values=pixel_values,
            attention_mask=attention_mask,
            position_ids=position_ids,
            past_key_values=past_key_values,
            inputs_embeds=inputs_embeds,
            vision_feature_layer=vision_feature_layer,
            vision_feature_select_strategy=vision_feature_select_strategy,
            labels=labels,
            use_cache=use_cache,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            return_dict=return_dict,
            cache_position=cache_position,
            logits_to_keep=logits_to_keep,
            image_sizes=image_sizes,
            **lm_kwargs,
        )


__all__ = ["AyaVisionForConditionalGeneration", "AyaVisionPreTrainedModel"]