models/modeling_tarsier.py

# Copyright (2024) Bytedance Ltd. and/or its affiliates
#
# 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.

# copy and modify from: https://github.com/huggingface/transformers/blob/main/src/transformers/models/llama/modeling_llama.py
""" PyTorch Llava model."""
from dataclasses import dataclass
from typing import List, Optional, Tuple, Union
import math
import numpy as np

import torch
import torch.utils.checkpoint
from torch import nn
import torch.nn.functional as F

from transformers import PreTrainedModel
from transformers.activations import ACT2FN
from transformers.cache_utils import Cache
from transformers.modeling_outputs import ModelOutput
from transformers.utils import (
    add_start_docstrings,
    add_start_docstrings_to_model_forward,
    logging,
    replace_return_docstrings,
)
from transformers.models.auto import AutoModel, AutoModelForCausalLM, CONFIG_MAPPING
from transformers import LlamaForCausalLM
from transformers.configuration_utils import PretrainedConfig


logger = logging.get_logger(__name__)

LLAVA_PRETRAINED_CONFIG_ARCHIVE_MAP = {
    "llava-hf/llava-v1.5-7b": "https://huggingface.co/llava-hf/llava-v1.5-7b/resolve/main/config.json",
}

class LlavaConfig(PretrainedConfig):
    r"""
    This is the configuration class to store the configuration of a [`LlavaForConditionalGeneration`]. It is used to instantiate an
    Llava 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 Llava-9B.

    e.g. [llava-hf/llava-9b](https://huggingface.co/llava-hf/llava-9b)

    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 (`LlavaVisionConfig`,  *optional*):
            Custom vision config or dict
        text_config (`Union[AutoConfig, dict]`, *optional*):
            The config object of the text backbone. Can be any of `LlamaConfig` or `MistralConfig`.
        ignore_index (`int`, *optional*, defaults to -100):
            The ignore index for the loss function.
        image_token_index (`int`, *optional*, defaults to 32000):
            The image token index to encode the image prompt.
        projector_hidden_act (`str`, *optional*, defaults to `"gelu"`):
            The activation function used by the multimodal projector.
        vision_feature_select_strategy (`str`, *optional*, defaults to `"default"`):
            The feature selection strategy used to select the vision feature from the CLIP backbone.
        vision_feature_layer (`int`, *optional*, defaults to -2):
            The index of the layer to select the vision feature.
        vocab_size (`int`, *optional*, defaults to 32000):
            Vocabulary size of the Llava model. Defines the number of different tokens that can be represented by the
            `inputs_ids` passed when calling [`~LlavaForConditionalGeneration`]

    Example:

    ```python
    >>> from transformers import LlavaForConditionalGeneration, LlavaConfig, CLIPVisionConfig, LlamaConfig

    >>> # Initializing a CLIP-vision config
    >>> vision_config = CLIPVisionConfig()

    >>> # Initializing a Llama config
    >>> text_config = LlamaConfig()

    >>> # Initializing a Llava llava-1.5-7b style configuration
    >>> configuration = LlavaConfig(vision_config, text_config)

    >>> # Initializing a model from the llava-1.5-7b style configuration
    >>> model = LlavaForConditionalGeneration(configuration)

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

    model_type = "llava"
    is_composition = False

    def __init__(
        self,
        vision_config=None,
        text_config=None,
        ignore_index=-100,
        image_token_index=32000,
        projector_hidden_act="gelu",
        vision_feature_select_strategy="default",
        vision_feature_layer=-2,
        vocab_size=32000,
        image_newline_idx=32002,
        image_new_idx=32003,
        **kwargs,
    ):
        self.ignore_index = ignore_index
        self.image_token_index = image_token_index
        self.projector_hidden_act = projector_hidden_act
        self.vision_feature_select_strategy = vision_feature_select_strategy
        self.vision_feature_layer = vision_feature_layer
        self.vocab_size = vocab_size
        self.image_newline_idx = image_newline_idx
        self.image_new_idx = image_new_idx

        self.vision_config = vision_config

        if isinstance(self.vision_config, dict):
            vision_config["model_type"] = (
                vision_config["model_type"] if "model_type" in vision_config else "clip_vision_model"
            )
            self.vision_config = CONFIG_MAPPING[vision_config["model_type"]](**vision_config)
        elif vision_config is None:
            self.vision_config = CONFIG_MAPPING["clip_vision_model"](
                intermediate_size=4096,
                hidden_size=1024,
                patch_size=14,
                image_size=336,
                num_hidden_layers=24,
                num_attention_heads=16,
                vocab_size=32000,
                projection_dim=768,
            )
        self.vocab_size = self.vocab_size

        self.text_config = text_config

        if isinstance(self.text_config, dict):
            text_config["model_type"] = text_config["model_type"] if "model_type" in text_config else "llama"
            self.text_config = CONFIG_MAPPING[text_config["model_type"]](**text_config)
            self.vocab_size = self.text_config.vocab_size
        elif text_config is None:
            self.text_config = CONFIG_MAPPING["llama"]()

        super().__init__(**kwargs)


logger = logging.get_logger(__name__)

_CONFIG_FOR_DOC = "LlavaConfig"

LLAVA_PRETRAINED_MODEL_ARCHIVE_LIST = [
    "llava-hf/llava-1.5-7b-hf",
    "llava-hf/llava-1.5-13b-hf",
    "llava-hf/bakLlava-v1-hf",
    # See all Llava models at https://huggingface.co/models?filter=llava
]


class Llava3DPositionalEncoding(nn.Module):
    def __init__(self, num_pos, dim) -> None:
        super().__init__()
        dim1, dim2, dim3 = self.split_dim(dim)
        frame_position_encodings = self.create_sinusoidal_positions(num_pos, dim1)
        height_position_encodings = self.create_sinusoidal_positions(num_pos, dim2)
        width_position_encodings = self.create_sinusoidal_positions(num_pos, dim3)

        self.register_buffer('frame_position_encodings', frame_position_encodings, persistent=False)
        self.register_buffer('height_position_encodings', height_position_encodings, persistent=False)
        self.register_buffer('width_position_encodings', width_position_encodings, persistent=False)

    def split_dim(self, dim):
        dim1 = dim // 3
        if dim1 % 2 != 0:
            dim1 -= 1

        dim2 = dim // 3
        if dim2 % 2 != 0:
            dim2 -= 1

        dim3 = dim - dim1 - dim2
        return dim1, dim2, dim3

    def create_sinusoidal_positions(self, num_pos: int, dim: int) -> torch.Tensor:
        inv_freq = 1.0 / (10000 ** (torch.arange(0, dim, 2) / dim))
        sinusoid_inp = torch.einsum("i , j -> i j", torch.arange(num_pos, dtype=torch.float), inv_freq).float()
        return torch.cat((torch.sin(sinusoid_inp), torch.cos(sinusoid_inp)), dim=1)

    def forward(self, frame_position_ids, height_position_ids, width_position_ids):
        frame_position_embeds = F.embedding(frame_position_ids, self.frame_position_encodings)
        height_position_embeds = F.embedding(height_position_ids, self.height_position_encodings)
        width_position_embeds = F.embedding(width_position_ids, self.width_position_encodings)

        return torch.cat([frame_position_embeds, height_position_embeds, width_position_embeds], dim = -1)


@dataclass
# Copied from transformers.models.idefics.modeling_idefics.IdeficsCausalLMOutputWithPast with Idefics->Llava
class LlavaCausalLMOutputWithPast(ModelOutput):
    """
    Base class for Llava causal language model (or autoregressive) outputs.

    Args:
        loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
            Language modeling loss (for next-token prediction).
        logits (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`):
            Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
        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)`)

            Contains pre-computed hidden-states (key and values in the self-attention blocks) that can be used (see
            `past_key_values` input) to speed up sequential decoding.
        hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`):
            Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
            one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

            Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
        attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`):
            Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
            sequence_length)`.

            Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
            heads.
        image_hidden_states (`tuple(torch.FloatTensor)`, *optional*):
            Tuple of `torch.FloatTensor` (one for the output of the image embeddings, `(batch_size, num_images,
            sequence_length, hidden_size)`.

            image_hidden_states of the model produced by the vision encoder, and optionally by the perceiver
    """

    loss: Optional[torch.FloatTensor] = None
    logits: torch.FloatTensor = None
    past_key_values: Optional[List[torch.FloatTensor]] = None
    hidden_states: Optional[Tuple[torch.FloatTensor]] = None
    attentions: Optional[Tuple[torch.FloatTensor]] = None
    image_hidden_states: Optional[Tuple[torch.FloatTensor]] = None
    vision_outputs: Optional[torch.FloatTensor] = None
    llm_attn_mask: Optional[Tuple[torch.FloatTensor]] = None


class LlavaMultiModalProjector(nn.Module):
    def __init__(self, config: LlavaConfig):
        super().__init__()

        self.linear_1 = nn.Linear(config.vision_config.hidden_size, config.text_config.hidden_size, bias=True)
        self.act = ACT2FN[config.projector_hidden_act]
        self.linear_2 = nn.Linear(config.text_config.hidden_size, config.text_config.hidden_size, bias=True)

    def forward(self, image_features):
        hidden_states = self.linear_1(image_features)
        hidden_states = self.act(hidden_states)
        hidden_states = self.linear_2(hidden_states)
        return hidden_states


TARSIER_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 ([`LlavaConfig`] or [`LlavaVisionConfig`]):
            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 LLaMA Model outputting raw hidden-states without any specific head on top.",
    TARSIER_START_DOCSTRING,
)
class TarsierPreTrainedModel(PreTrainedModel):
    config_class = LlavaConfig
    base_model_prefix = "model"
    supports_gradient_checkpointing = True
    _no_split_modules = ["LlavaVisionAttention"]
    _skip_keys_device_placement = "past_key_values"
    _supports_flash_attn_2 = True

    def _init_weights(self, module):
        # important: this ported version of Llava isn't meant for training from scratch - only
        # inference and fine-tuning - so the proper init weights code has been removed - the original codebase
        # https://github.com/haotian-liu/LLaVA/tree/main/llava should serve for that purpose
        std = (
            self.config.initializer_range
            if hasattr(self.config, "initializer_range")
            else self.config.text_config.initializer_range
        )

        if hasattr(module, "class_embedding"):
            module.class_embedding.data.normal_(mean=0.0, std=std)

        if isinstance(module, (nn.Linear, nn.Conv2d)):
            module.weight.data.normal_(mean=0.0, std=std)
            if module.bias is not None:
                module.bias.data.zero_()
        elif isinstance(module, nn.Embedding):
            module.weight.data.normal_(mean=0.0, std=std)
            if module.padding_idx is not None:
                module.weight.data[module.padding_idx].zero_()

    @property
    def _supports_sdpa(self):
        """
        Retrieve language_model's attribute to check whether the model supports
        SDPA or not.
        """
        return self.language_model._supports_sdpa


TARSIER_INPUTS_DOCSTRING = r"""
    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 [`CLIPImageProcessor.__call__`] for details ([]`LlavaProcessor`] uses
            [`CLIPImageProcessor`] 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.
        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.
"""


@add_start_docstrings(
    """The LLAVA model which consists of a vision backbone and a language model.""",
    TARSIER_INPUTS_DOCSTRING,
)
class TarsierForConditionalGeneration(TarsierPreTrainedModel):
    def __init__(self, config: LlavaConfig):
        super().__init__(config)
        self.vision_tower = AutoModel.from_config(config.vision_config, trust_remote_code=True)
        self.multi_modal_projector = LlavaMultiModalProjector(config)
        self.vocab_size = config.vocab_size
        self.language_model = AutoModelForCausalLM.from_config(config.text_config, attn_implementation="flash_attention_2")
        image_newline_idx = torch.tensor([config.image_newline_idx], dtype=torch.long)
        image_new_idx = torch.tensor([config.image_new_idx], dtype=torch.long)
        self.register_buffer('image_newline_idx', image_newline_idx, persistent=False)
        self.register_buffer('image_new_idx', image_new_idx, persistent=False)
        self.pad_token_id = self.config.pad_token_id if self.config.pad_token_id is not None else -1
        self.post_init()

    def get_input_embeddings(self):
        return self.language_model.get_input_embeddings()

    def set_input_embeddings(self, value):
        self.language_model.set_input_embeddings(value)

    def get_output_embeddings(self):
        return self.language_model.get_output_embeddings()

    def set_output_embeddings(self, new_embeddings):
        self.language_model.set_output_embeddings(new_embeddings)

    def set_decoder(self, decoder):
        self.language_model.set_decoder(decoder)

    def get_decoder(self):
        return self.language_model.get_decoder()

    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.config.vocab_size = model_embeds.num_embeddings
        self.vocab_size = model_embeds.num_embeddings
        return model_embeds

    def _merge_input_ids_with_image_features(self, image_features, inputs_embeds, input_ids, attention_mask, labels):
        num_images, num_image_patches, embed_dim = image_features.shape

        batch_size, sequence_length = input_ids.shape
        left_padding = not torch.sum(input_ids[:, -1] == torch.tensor(self.pad_token_id))
        # 1. Create a mask to know where special image tokens are
        special_image_token_mask = input_ids == self.config.image_token_index
        num_special_image_tokens = torch.sum(special_image_token_mask, dim=-1)
        # Compute the maximum embed dimension
        max_embed_dim = (num_special_image_tokens.max() * (num_image_patches - 1)) + sequence_length
        batch_indices, non_image_indices = torch.where(input_ids != self.config.image_token_index)

        # 2. Compute the positions where text should be written
        # Calculate new positions for text tokens in merged image-text sequence.
        # `special_image_token_mask` identifies image tokens. Each image token will be replaced by `nb_text_tokens_per_images - 1` text tokens.
        # `torch.cumsum` computes how each image token shifts subsequent text token positions.
        # - 1 to adjust for zero-based indexing, as `cumsum` inherently increases indices by one.
        new_token_positions = torch.cumsum((special_image_token_mask * (num_image_patches - 1) + 1), -1) - 1
        nb_image_pad = max_embed_dim - 1 - new_token_positions[:, -1]
        if left_padding:
            new_token_positions += nb_image_pad[:, None]  # offset for left padding
        text_to_overwrite = new_token_positions[batch_indices, non_image_indices]

        # 3. Create the full embedding, already padded to the maximum position
        final_embedding = torch.zeros(
            batch_size, max_embed_dim, embed_dim, dtype=inputs_embeds.dtype, device=inputs_embeds.device
        )
        final_attention_mask = torch.zeros(
            batch_size, max_embed_dim, dtype=attention_mask.dtype, device=inputs_embeds.device
        )
        if labels is not None:
            final_labels = torch.full(
                (batch_size, max_embed_dim), self.config.ignore_index, dtype=input_ids.dtype, device=input_ids.device
            )
        # In case the Vision model or the Language model has been offloaded to CPU, we need to manually
        # set the corresponding tensors into their correct target device.
        target_device = inputs_embeds.device
        batch_indices, non_image_indices, text_to_overwrite = (
            batch_indices.to(target_device),
            non_image_indices.to(target_device),
            text_to_overwrite.to(target_device),
        )
        attention_mask = attention_mask.to(target_device)

        # 4. Fill the embeddings based on the mask. If we have ["hey" "<image>", "how", "are"]
        # we need to index copy on [0, 577, 578, 579] for the text and [1:576] for the image features
        final_embedding[batch_indices, text_to_overwrite] = inputs_embeds[batch_indices, non_image_indices]
        final_attention_mask[batch_indices, text_to_overwrite] = attention_mask[batch_indices, non_image_indices]
        if labels is not None:
            final_labels[batch_indices, text_to_overwrite] = labels[batch_indices, non_image_indices]

        # 5. Fill the embeddings corresponding to the images. Anything that is still zeros needs filling
        image_to_overwrite = torch.all(final_embedding == 0, dim=-1)
        image_to_overwrite &= image_to_overwrite.cumsum(-1) - 1 >= nb_image_pad[:, None].to(target_device)

        if image_to_overwrite.sum() != image_features.shape[:-1].numel():
            raise ValueError(
                f"The input provided to the model are wrong. The number of image tokens is {torch.sum(special_image_token_mask)} while"
                f" the number of image given to the model is {num_images}. This prevents correct indexing and breaks batch generation."
            )

        final_embedding[image_to_overwrite] = image_features.contiguous().reshape(-1, embed_dim).to(target_device)
        final_attention_mask |= image_to_overwrite
        position_ids = (final_attention_mask.cumsum(-1) - 1).masked_fill_((final_attention_mask == 0), 1)

        if labels is None:
            final_labels = None

        return final_embedding, final_attention_mask, final_labels, position_ids
    
    def add_split_tokens(self, image_features):
        num_images, num_image_patches, embed_dim = image_features.shape
        num_height_patches, num_width_patches = int(math.sqrt(num_image_patches)), int(math.sqrt(num_image_patches))

        # add image_newline
        image_newline = self.get_input_embeddings()(self.image_newline_idx).squeeze()
        image_features = image_features.view(num_images, num_height_patches, num_width_patches, embed_dim)
        image_features = torch.cat([
            image_features,
            image_newline.expand((num_images, num_height_patches, 1, embed_dim)).to(device=image_features.device)
        ], dim=2)
        num_image_patches += num_height_patches
        image_features = image_features.view(num_images, num_image_patches, embed_dim)

        # add image_new
        image_new = self.get_input_embeddings()(self.image_new_idx).squeeze()
        image_features = torch.cat([
            image_features,
            image_new.expand((num_images, 1, embed_dim)).to(device=image_features.device)
        ], dim = 1)

        return image_features

    @add_start_docstrings_to_model_forward(TARSIER_INPUTS_DOCSTRING)
    @replace_return_docstrings(output_type=LlavaCausalLMOutputWithPast, config_class=_CONFIG_FOR_DOC)
    def forward(
        self,
        input_ids: torch.LongTensor = None,
        pixel_values: 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[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,
        **kwargs,
    ) -> Union[Tuple, LlavaCausalLMOutputWithPast]:
        r"""
        Args:
            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]`.

        Returns:

        Example:

        ```python
        >>> from PIL import Image
        >>> import requests
        >>> from transformers import AutoProcessor, LlavaForConditionalGeneration

        >>> model = LlavaForConditionalGeneration.from_pretrained("llava-hf/llava-1.5-7b-hf")
        >>> processor = AutoProcessor.from_pretrained("llava-hf/llava-1.5-7b-hf")

        >>> prompt = "<image>\nUSER: What's the content of the image?\nASSISTANT:"
        >>> url = "https://www.ilankelman.org/stopsigns/australia.jpg"
        >>> image = Image.open(requests.get(url, stream=True).raw)

        >>> inputs = processor(text=prompt, images=image, return_tensors="pt")

        >>> # Generate
        >>> generate_ids = model.generate(**inputs, max_length=30)
        >>> processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
        "\nUSER: What's the content of the image?\nASSISTANT: The image features a stop sign on a street corner"
        ```"""
        output_attentions = output_attentions if output_attentions is not None else self.config.output_attentions
        output_hidden_states = (
            output_hidden_states if output_hidden_states is not None else self.config.output_hidden_states
        )
        return_dict = return_dict if return_dict is not None else self.config.use_return_dict
        vision_feature_layer = (
            vision_feature_layer if vision_feature_layer is not None else self.config.vision_feature_layer
        )
        vision_feature_select_strategy = (
            vision_feature_select_strategy
            if vision_feature_select_strategy is not None
            else self.config.vision_feature_select_strategy
        )

        image_features = None
        if inputs_embeds is None:
            # 1. Extra the input embeddings
            inputs_embeds = self.get_input_embeddings()(input_ids)
            
            # 2. Merge text and images
            if pixel_values is not None and input_ids.shape[1] != 1:
                pixel_values = pixel_values.to(dtype=self.vision_tower.dtype)
                image_outputs = self.vision_tower(pixel_values, output_hidden_states=True)
                # this is not memory efficient at all (output_hidden_states=True) will save all the hidden stated.
                selected_image_feature = image_outputs.hidden_states[vision_feature_layer]

                if vision_feature_select_strategy == "default":
                    selected_image_feature = selected_image_feature[:, 1:]
                elif vision_feature_select_strategy == "full":
                    selected_image_feature = selected_image_feature
                else:
                    raise ValueError(
                        f"Unexpected select feature strategy: {self.config.vision_feature_select_strategy}"
                    )

                image_features = self.multi_modal_projector(selected_image_feature)

                special_image_token_mask = input_ids == self.config.image_token_index
                num_special_image_tokens = torch.sum(special_image_token_mask, dim = -1)

                image_features = self.add_split_tokens(image_features)

                if sum(num_special_image_tokens) > 0:
                    # print(f'num_special_image_tokens: {num_special_image_tokens}')
                    inputs_embeds, attention_mask, labels, position_ids = self._merge_input_ids_with_image_features(
                        image_features, inputs_embeds, input_ids, attention_mask, labels
                    )
                else:
                    inputs_embeds = image_features.sum(dim=(0,1))[None, None, :] * 0. + inputs_embeds

                if labels is None:
                    labels = torch.full_like(attention_mask, self.config.ignore_index).to(torch.long)
            else:
                # In case input_ids.shape[1] == 1 & pixel_values==None & past_key_values != None, we are in the case of
                # generation with cache
                if past_key_values is not None and pixel_values is not None and input_ids.shape[1] == 1:
                    # Retrieve the first layer to inspect the logits and mask out the hidden states
                    # that are set to 0
                    first_layer_past_key_value = past_key_values[0][0][:, :, :, 0]

                    # Sum all dimensions of head_dim (-2) to avoid random errors such as: https://github.com/huggingface/transformers/pull/28032#issuecomment-1863691941
                    batch_index, non_attended_tokens = torch.where(first_layer_past_key_value.float().sum(-2) == 0)

                    # Get the target length
                    target_seqlen = first_layer_past_key_value.shape[-1] + 1
                    extended_attention_mask = torch.ones(
                        (attention_mask.shape[0], target_seqlen),
                        dtype=attention_mask.dtype,
                        device=attention_mask.device,
                    )

                    extended_attention_mask[batch_index, non_attended_tokens] = 0

                    valid_indices = torch.ones_like(attention_mask)
                    valid_indices[:, 0] = target_seqlen - extended_attention_mask.sum(dim=-1)
                    valid_indices = torch.cumsum(valid_indices, dim=-1)
                    extended_attention_mask = extended_attention_mask.scatter(1, valid_indices, attention_mask)
                    attention_mask = extended_attention_mask
                    position_ids = torch.sum(attention_mask, dim=1).unsqueeze(-1) - 1
        outputs = self.language_model(
            attention_mask=attention_mask,
            position_ids=position_ids,
            past_key_values=past_key_values,
            inputs_embeds=inputs_embeds,
            use_cache=use_cache,
            output_attentions=output_attentions,
            output_hidden_states=output_hidden_states,
            # use_rmpad=kwargs.get("use_rmpad", False),
            return_dict=return_dict,
        )

        logits = outputs[0]

        loss = None
        if labels is not None:
            # Shift so that tokens < n predict n
            if attention_mask is not None:
                shift_attention_mask = attention_mask[..., 1:]
                shift_logits = logits[..., :-1, :][shift_attention_mask.to(logits.device) != 0].contiguous()
                shift_labels = labels[..., 1:][shift_attention_mask.to(labels.device) != 0].contiguous()
            else:
                shift_logits = logits[..., :-1, :].contiguous()
                shift_labels = labels[..., 1:].contiguous()
            # Flatten the tokens
            loss_fct = nn.CrossEntropyLoss()
            loss = loss_fct(
                shift_logits.view(-1, shift_logits.size(-1)), shift_labels.view(-1).to(shift_logits.device)
            )

        if not return_dict:
            output = (logits,) + outputs[1:]
            return (loss,) + output if loss is not None else output

        return LlavaCausalLMOutputWithPast(
            loss=loss,
            logits=logits,
            past_key_values=outputs.past_key_values,
            hidden_states=outputs.hidden_states,
            attentions=outputs.attentions,
            llm_attn_mask=attention_mask
        )

    def prepare_inputs_for_generation(
        self, input_ids, past_key_values=None, inputs_embeds=None, pixel_values=None, attention_mask=None, **kwargs
    ):
        if past_key_values is not None:
            if isinstance(past_key_values, Cache):
                cache_length = past_key_values.get_seq_length()
                past_length = past_key_values.seen_tokens
            else:
                cache_length = past_length = past_key_values[0][0].shape[2]

            # Keep only the unprocessed tokens:
            # 1 - If the length of the attention_mask exceeds the length of input_ids, then we are in a setting where
            # some of the inputs are exclusively passed as part of the cache (e.g. when passing input_embeds as
            # input)
            if attention_mask is not None and attention_mask.shape[1] > input_ids.shape[1]:
                input_ids = input_ids[:, -(attention_mask.shape[1] - past_length) :]
            # 2 - If the past_length is smaller than input_ids', then input_ids holds all input tokens. We can discard
            # input_ids based on the past_length.
            elif past_length < input_ids.shape[1]:
                input_ids = input_ids[:, past_length:]
            # 3 - Otherwise (past_length >= input_ids.shape[1]), let's assume input_ids only has unprocessed tokens.
            elif self.config.image_token_index in input_ids:
                input_ids = input_ids[:, input_ids.shape[1] - 1 :]
            # If the cache has seen more tokens than it can hold, then the cache has a size limit. Let's discard the
            # older attention values, as their corresponding values are not part of the input.
            if cache_length < past_length and attention_mask is not None:
                attention_mask = attention_mask[:, -(cache_length + input_ids.shape[1]) :]

        position_ids = kwargs.get("position_ids", None)
        if attention_mask is not None and position_ids is None:
            # create position_ids on the fly for batch generation
            position_ids = attention_mask.long().cumsum(-1) - 1
            position_ids.masked_fill_(attention_mask == 0, 1)
            if past_key_values:
                position_ids = position_ids[:, -input_ids.shape[1] :]

        # if `inputs_embeds` are passed, we only want to use them in the 1st generation step
        if inputs_embeds is not None and past_key_values is None:
            model_inputs = {"inputs_embeds": inputs_embeds}
        else:
            model_inputs = {"input_ids": input_ids}

        model_inputs.update(
            {
                "position_ids": position_ids,
                "past_key_values": past_key_values,
                "use_cache": kwargs.get("use_cache"),
                "attention_mask": attention_mask,
                "pixel_values": pixel_values,
            }
        )
        return model_inputs

    def _reorder_cache(self, *args, **kwargs):
        return self.language_model._reorder_cache(*args, **kwargs)