# PaliGemma[[paligemma]]

## 개요[[overview]]

PaliGemma 모델은 구글이 제안한 [PaliGemma – Google의 최첨단 오픈 비전 언어 모델](https://huggingface.co/blog/paligemma)에서 소개 되었습니다. PaliGemma는 [SigLIP](siglip) 비전 인코더와 [Gemma](gemma) 언어 인코더로 구성된 3B 규모의 비전-언어 모델로, 두 인코더가 멀티모달 선형 프로젝션으로 연결되어 있습니다. 이 모델은 이미지를 고정된 수의 VIT토큰으로 분할하고 이를 선택적 프롬프트 앞에 추가 하며, 모든 이미지 토큰과 입력 텍스트 토큰에 대해 전체 블록 어텐션을 사용하는 특징을 가지고 있습니다.

PaliGemma는 224x224, 448x448, 896x896의 3가지 해상도로 제공되며, 3개의 기본 모델과 55개의 다양한 작업에 대해 미세 조정된 버전, 그리고 2개의 혼합 모델이 있습니다.

<img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/blog/paligemma/paligemma_arch.png"
alt="drawing" width="600"/>

 PaliGemma 아키텍처 블로그 포스트. 

이 모델은 [Molbap](https://huggingface.co/Molbap)에 의해 기여 되었습니다.

## 사용 팁[[usage-tips]]

PaliGemma의 추론은 다음처럼 수행됩니다:

```python
from transformers import AutoProcessor, PaliGemmaForConditionalGeneration

model_id = "google/paligemma-3b-mix-224"
model = PaliGemmaForConditionalGeneration.from_pretrained(model_id)
processor = AutoProcessor.from_pretrained(model_id)

prompt = "What is on the flower?"
image_file = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg?download=true"
raw_image = Image.open(requests.get(image_file, stream=True).raw)
inputs = processor(raw_image, prompt, return_tensors="pt")
output = model.generate(**inputs, max_new_tokens=20)

print(processor.decode(output[0], skip_special_tokens=True)[len(prompt):])
```

- PaliGemma는 대화용으로 설계되지 않았으며, 특정 사용 사례에 대해 미세 조정할 때 가장 잘 작동합니다. PaliGemma를 미세 조정할 수 있는 몇 가지 하위 작업에는 이미지 캡셔닝, 시각적 질문 답변(VQA), 오브젝트 디텍션, 참조 표현 분할 및 문서 이해가 포함됩니다.
- 모델에 필요한 이미지, 텍스트 및 선택적 레이블을 준비하는데 `PaliGemmaProcessor`를 사용할 수 있습니다. PaliGemma 모델을 미세 조정할 때는, 프로세서에 `suffix`인자를 전달하여 다음 처럼 모델의 `labels`를 생성할 수 있습니다:

```python
prompt = "What is on the flower?"
answer = "a bee"
inputs = processor(images=raw_image, text=prompt, suffix=answer, return_tensors="pt")
```

## 자료[[resources]]

PaliGemma를 시작하는 데 도움이 되는 Hugging Face와 community 자료 목록(🌎로 표시됨) 입니다.여기에 포함될 자료를 제출하고 싶으시다면 PR(Pull Request)를 열어주세요. 리뷰 해드리겠습니다! 자료는 기존 자료를 복제하는 대신 새로운 내용을 담고 있어야 합니다.

- PaliGemma의 모든 기능을 소개하는 블로그 포스트는 [이곳](https://huggingface.co/blog/paligemma)에서 찾을 수 있습니다. 🌎
- Trainer API를 사용하여 VQA(Visual Question Answering)를 위해 PaliGemma를 미세 조정하는 방법과 추론에 대한 데모 노트북은 [이곳](https://github.com/huggingface/notebooks/tree/main/examples/paligemma)에서 찾을 수 있습니다. 🌎
- 사용자 정의 데이터셋(영수증 이미지 -> JSON)에 대해 PaliGemma를 미세 조정하는 방법과 추론에 대한 데모 노트북은 [이곳](https://github.com/NielsRogge/Transformers-Tutorials/tree/master/PaliGemma)에서 찾을 수 있습니다. 🌎

## PaliGemmaConfig[[transformers.PaliGemmaConfig]][[transformers.PaliGemmaConfig]]

#### transformers.PaliGemmaConfig[[transformers.PaliGemmaConfig]]

[Source](https://github.com/huggingface/transformers/blob/v5.8.0/src/transformers/models/paligemma/configuration_paligemma.py#L24)

This is the configuration class to store the configuration of a PaliGemmaModel. It is used to instantiate a Paligemma
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 [google/paligemma-3b-pt-224](https://huggingface.co/google/paligemma-3b-pt-224)

Configuration objects inherit from [PreTrainedConfig](/docs/transformers/v5.8.0/ko/main_classes/configuration#transformers.PreTrainedConfig) and can be used to control the model outputs. Read the
documentation from [PreTrainedConfig](/docs/transformers/v5.8.0/ko/main_classes/configuration#transformers.PreTrainedConfig) for more information.

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
```

**Parameters:**

vision_config (`Union[dict, ~configuration_utils.PreTrainedConfig]`, *optional*) : The config object or dictionary of the vision backbone.

text_config (`Union[dict, ~configuration_utils.PreTrainedConfig]`, *optional*) : The config object or dictionary of the text backbone.

image_token_index (`int`, *optional*, defaults to `256000`) : The image token index used as a placeholder for input images.

vocab_size (`int`, *optional*, defaults to `257152`) : Vocabulary size of the model. Defines the number of different tokens that can be represented by the `input_ids`.

projection_dim (`int`, *optional*, defaults to `2048`) : Dimensionality of text and vision projection layers.

hidden_size (`int`, *optional*, defaults to `2048`) : Dimension of the hidden representations.

tie_word_embeddings (`bool`, *optional*, defaults to `True`) : Whether to tie weight embeddings according to model's `tied_weights_keys` mapping.

## PaliGemmaProcessor[[transformers.PaliGemmaProcessor]][[transformers.PaliGemmaProcessor]]

#### transformers.PaliGemmaProcessor[[transformers.PaliGemmaProcessor]]

[Source](https://github.com/huggingface/transformers/blob/v5.8.0/src/transformers/models/paligemma/processing_paligemma.py#L99)

Constructs a PaliGemmaProcessor which wraps a image processor and a tokenizer into a single processor.

[PaliGemmaProcessor](/docs/transformers/v5.8.0/ko/model_doc/paligemma#transformers.PaliGemmaProcessor) offers all the functionalities of [SiglipImageProcessor](/docs/transformers/v5.8.0/ko/model_doc/siglip#transformers.SiglipImageProcessor) and `tokenizer_class`. See the
[~SiglipImageProcessor](/docs/transformers/v5.8.0/ko/model_doc/siglip#transformers.SiglipImageProcessor) and `~tokenizer_class` for more information.

**Parameters:**

image_processor (`SiglipImageProcessor`) : The image processor is a required input.

tokenizer (`tokenizer_class`) : The tokenizer is a required input.

chat_template (`str`) : A Jinja template to convert lists of messages in a chat into a tokenizable string.

## PaliGemmaForConditionalGeneration[[transformers.PaliGemmaForConditionalGeneration]][[transformers.PaliGemmaForConditionalGeneration]]

#### transformers.PaliGemmaForConditionalGeneration[[transformers.PaliGemmaForConditionalGeneration]]

[Source](https://github.com/huggingface/transformers/blob/v5.8.0/src/transformers/models/paligemma/modeling_paligemma.py#L392)

The Base Paligemma model which consists of a vision backbone and a language model without language modeling head.,

This model inherits from [PreTrainedModel](/docs/transformers/v5.8.0/ko/main_classes/model#transformers.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.

forwardtransformers.PaliGemmaForConditionalGeneration.forwardhttps://github.com/huggingface/transformers/blob/v5.8.0/src/transformers/models/paligemma/modeling_paligemma.py#L405[{"name": "input_ids", "val": ": torch.LongTensor | None = None"}, {"name": "pixel_values", "val": ": torch.FloatTensor | None = None"}, {"name": "attention_mask", "val": ": torch.Tensor | None = None"}, {"name": "position_ids", "val": ": torch.LongTensor | None = None"}, {"name": "past_key_values", "val": ": transformers.cache_utils.Cache | None = None"}, {"name": "token_type_ids", "val": ": torch.LongTensor | None = None"}, {"name": "inputs_embeds", "val": ": torch.FloatTensor | None = None"}, {"name": "labels", "val": ": torch.LongTensor | None = None"}, {"name": "use_cache", "val": ": bool | None = None"}, {"name": "logits_to_keep", "val": ": int | torch.Tensor = 0"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **input_ids** (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) --
  Indices of input sequence tokens in the vocabulary. Padding will be ignored by default.

  Indices can be obtained using [AutoTokenizer](/docs/transformers/v5.8.0/ko/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v5.8.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and
  [PreTrainedTokenizer.__call__()](/docs/transformers/v5.8.0/ko/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details.

  [What are input IDs?](../glossary#input-ids)
- **pixel_values** (`torch.FloatTensor` of shape `(batch_size, num_channels, image_size, image_size)`, *optional*) --
  The tensors corresponding to the input images. Pixel values can be obtained using
  [SiglipImageProcessor](/docs/transformers/v5.8.0/ko/model_doc/siglip#transformers.SiglipImageProcessor). See `SiglipImageProcessor.__call__()` for details ([PaliGemmaProcessor](/docs/transformers/v5.8.0/ko/model_doc/paligemma#transformers.PaliGemmaProcessor) uses
  [SiglipImageProcessor](/docs/transformers/v5.8.0/ko/model_doc/siglip#transformers.SiglipImageProcessor) 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)
- **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** (`~cache_utils.Cache`, *optional*) --
  Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention
  blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values`
  returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`.

  Only [Cache](/docs/transformers/v5.8.0/ko/internal/generation_utils#transformers.Cache) instance is allowed as input, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).
  If no `past_key_values` are passed, [DynamicCache](/docs/transformers/v5.8.0/ko/internal/generation_utils#transformers.DynamicCache) will be initialized by default.

  The model will output the same cache format that is fed as input.

  If `past_key_values` are used, the user is expected to input only unprocessed `input_ids` (those that don't
  have their past key value states given to this model) of shape `(batch_size, unprocessed_length)` instead of all `input_ids`
  of shape `(batch_size, sequence_length)`.
- **token_type_ids** (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) --
  Segment token indices to indicate first and second portions of the inputs. Indices are selected in `[0, 1]`:

  - 0 corresponds to a *sentence A* token,
  - 1 corresponds to a *sentence B* token.

  [What are token type IDs?](../glossary#token-type-ids)
- **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.
- **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.text_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.text_config.vocab_size]`.
- **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`).
- **logits_to_keep** (`Union[int, torch.Tensor]`, *optional*, defaults to `0`) --
  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).0`PaliGemmaCausalLMOutputWithPast` or `tuple(torch.FloatTensor)`A `PaliGemmaCausalLMOutputWithPast` or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration ([PaliGemmaConfig](/docs/transformers/v5.8.0/ko/model_doc/paligemma#transformers.PaliGemmaConfig)) and inputs.
The [PaliGemmaForConditionalGeneration](/docs/transformers/v5.8.0/ko/model_doc/paligemma#transformers.PaliGemmaForConditionalGeneration) forward method, overrides the `__call__` special method.

Although the recipe for forward pass needs to be defined within this function, one should call the `Module`
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.

- **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.text_config.vocab_size)`) -- Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
- **past_key_values** (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`) -- It is a [Cache](/docs/transformers/v5.8.0/ko/internal/generation_utils#transformers.Cache) instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

  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.

Example:

```python
>>> from PIL import Image
>>> import httpx
>>> from io import BytesIO
>>> from transformers import AutoProcessor, PaliGemmaForConditionalGeneration

>>> model = PaliGemmaForConditionalGeneration.from_pretrained("google/paligemma2-3b-mix-224")
>>> processor = AutoProcessor.from_pretrained("google/paligemma2-3b-mix-224")

>>> prompt = "Where is the cat standing?"
>>> url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/pipeline-cat-chonk.jpeg"
>>> with httpx.stream("GET", url) as response:
...     image = Image.open(BytesIO(response.read()))

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

>>> # Generate
>>> generate_ids = model.generate(**inputs,)
>>> processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
"Where is the cat standing?\nsnow"
```

**Parameters:**

config ([PaliGemmaConfig](/docs/transformers/v5.8.0/ko/model_doc/paligemma#transformers.PaliGemmaConfig)) : 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 [from_pretrained()](/docs/transformers/v5.8.0/ko/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights.

**Returns:**

``PaliGemmaCausalLMOutputWithPast` or `tuple(torch.FloatTensor)``

A `PaliGemmaCausalLMOutputWithPast` or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration ([PaliGemmaConfig](/docs/transformers/v5.8.0/ko/model_doc/paligemma#transformers.PaliGemmaConfig)) and inputs.