# coding=utf-8
# Copyright 2024 The HuggingFace Inc. team.
#
# 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 ColPali model"""
from dataclasses import dataclass
from typing import List, Optional, Tuple, Union
import torch
from torch import nn
from transformers import AutoModelForImageTextToText
from ...cache_utils import Cache
from ...modeling_utils import PreTrainedModel
from ...utils import (
ModelOutput,
add_start_docstrings,
add_start_docstrings_to_model_forward,
replace_return_docstrings,
)
from .configuration_colpali import ColPaliConfig
_CONFIG_FOR_DOC = "ColPaliConfig"
COLPALI_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 ([`ColPaliConfig`]):
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 ColPali model outputting raw hidden-states without any specific head on top.",
COLPALI_START_DOCSTRING,
)
class ColPaliPreTrainedModel(PreTrainedModel):
config_class = ColPaliConfig
base_model_prefix = "model"
_no_split_modules = []
def _init_weights(self, module):
std = (
self.config.initializer_range
if hasattr(self.config, "initializer_range")
else self.config.vlm_config.text_config.initializer_range
)
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_()
@dataclass
class ColPaliForRetrievalOutput(ModelOutput):
"""
Base class for ColPali embeddings output.
Args:
loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided):
Language modeling loss (for next-token prediction).
embeddings (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`):
The embeddings of the model.
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 (`torch.FloatTensor`, *optional*):
A `torch.FloatTensor` of size `(batch_size, num_images, sequence_length, hidden_size)`.
image_hidden_states of the model produced by the vision encoder after projecting last hidden state.
"""
loss: Optional[torch.FloatTensor] = None
embeddings: Optional[torch.Tensor] = None
past_key_values: Optional[Union[List[torch.FloatTensor], Cache]] = None
hidden_states: Optional[Tuple[torch.FloatTensor]] = None
attentions: Optional[Tuple[torch.FloatTensor]] = None
image_hidden_states: Optional[torch.FloatTensor] = None
COLPALI_FOR_RETRIEVAL_INPUT_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 [`SiglipImageProcessor.__call__`] for details ([]`PaliGemmaProcessor`] uses
[`SiglipImageProcessor`] for processing images). If none, ColPali will only process text (query embeddings).
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**.
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.
kwargs (`Dict[str, Any]`, *optional*):
Additional key word arguments passed along to the vlm backbone model.
"""
@add_start_docstrings(
"""
In our proposed ColPali approach, we leverage VLMs to construct efficient multi-vector embeddings directly
from document images (“screenshots”) for document retrieval. We train the model to maximize the similarity
between these document embeddings and the corresponding query embeddings, using the late interaction method
introduced in ColBERT.
Using ColPali removes the need for potentially complex and brittle layout recognition and OCR pipelines with a
single model that can take into account both the textual and visual content (layout, charts, etc.) of a document.
"""
)
class ColPaliForRetrieval(ColPaliPreTrainedModel):
def __init__(self, config: ColPaliConfig):
super().__init__(config)
self.config = config
self.vocab_size = config.vlm_config.text_config.vocab_size
vlm = AutoModelForImageTextToText.from_config(config.vlm_config)
if vlm.language_model._tied_weights_keys is not None:
self._tied_weights_keys = [f"vlm.language_model.{k}" for k in vlm.language_model._tied_weights_keys]
self.vlm = vlm
self.embedding_dim = self.config.embedding_dim
self.embedding_proj_layer = nn.Linear(
self.config.vlm_config.text_config.hidden_size,
self.embedding_dim,
)
self.post_init()
@add_start_docstrings_to_model_forward(COLPALI_FOR_RETRIEVAL_INPUT_DOCSTRING)
@replace_return_docstrings(output_type=ColPaliForRetrievalOutput, config_class=_CONFIG_FOR_DOC)
def forward(
self,
input_ids: Optional[torch.LongTensor] = None,
pixel_values: Optional[torch.FloatTensor] = None,
attention_mask: Optional[torch.Tensor] = None,
output_attentions: Optional[bool] = None,
output_hidden_states: Optional[bool] = None,
return_dict: Optional[bool] = None,
**kwargs,
) -> Union[Tuple, ColPaliForRetrievalOutput]:
r"""
Returns:
"""
if "pixel_values" in kwargs:
kwargs["pixel_values"] = kwargs["pixel_values"].to(dtype=self.dtype)
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
outputs = self.vlm(
input_ids=input_ids,
attention_mask=attention_mask,
pixel_values=pixel_values,
output_hidden_states=True,
return_dict=return_dict,
output_attentions=output_attentions,
**kwargs,
)
last_hidden_states = outputs.hidden_states[-1] # (batch_size, sequence_length, hidden_size)
embeddings = self.embedding_proj_layer(last_hidden_states) # (batch_size, sequence_length, dim)
# L2 normalization
embeddings = embeddings / embeddings.norm(dim=-1, keepdim=True) # (batch_size, sequence_length, dim)
embeddings = embeddings * attention_mask.unsqueeze(-1) # (batch_size, sequence_length, dim)
loss = None
if not return_dict:
output = (embeddings,) + outputs[2:]
output[2] = output[2] if output_hidden_states is not None else None
output[-1] = (outputs.image_hidden_states if pixel_values is not None else None,)
return (loss,) + output if loss is not None else output
return ColPaliForRetrievalOutput(
loss=loss,
embeddings=embeddings,
past_key_values=outputs.past_key_values,
hidden_states=outputs.hidden_states if output_hidden_states else None,
attentions=outputs.attentions,
image_hidden_states=outputs.image_hidden_states if pixel_values is not None else None,
)
def get_input_embeddings(self):
return self.vlm.language_model.get_input_embeddings()
def set_input_embeddings(self, value):
self.vlm.language_model.set_input_embeddings(value)
def get_output_embeddings(self):
return self.vlm.language_model.get_output_embeddings()
def set_output_embeddings(self, new_embeddings):
self.vlm.language_model.set_output_embeddings(new_embeddings)
def set_decoder(self, decoder):
self.vlm.language_model.set_decoder(decoder)
def get_decoder(self):
return self.vlm.language_model.get_decoder()
def tie_weights(self):
return self.vlm.language_model.tie_weights()
def resize_token_embeddings(
self,
new_num_tokens: Optional[int] = None,
pad_to_multiple_of: Optional[int] = None,
mean_resizing: bool = True,
) -> nn.Embedding:
model_embeds = self.vlm.language_model.resize_token_embeddings(
new_num_tokens=new_num_tokens,
pad_to_multiple_of=pad_to_multiple_of,
mean_resizing=mean_resizing,
)
self.config.vlm_config.text_config.vocab_size = model_embeds.num_embeddings
self.config.vlm_config.vocab_size = model_embeds.num_embeddings
self.vlm.vocab_size = model_embeds.num_embeddings
self.vocab_size = model_embeds.num_embeddings
return model_embeds
__all__ = [
"ColPaliForRetrieval",
"ColPaliForRetrievalOutput",
"ColPaliPreTrainedModel",
]