Hugging Face's logo

Transformers documentation

QianfanOCR

You are viewing main version, which requires installation from source. If you'd like regular pip install, checkout the latest stable version (v5.5.4).
Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

This model was released on 2026-03-18 and added to Hugging Face Transformers on 2026-04-16.

QianfanOCR

PyTorch

Overview

Qianfan-OCR is a 4B-parameter end-to-end document intelligence model developed by the Baidu Qianfan Team. It was proposed in Qianfan-OCR: A Unified End-to-End Model for Document Intelligence by Daxiang Dong et al.

Unlike traditional multi-stage OCR pipelines, Qianfan-OCR performs direct image-to-text conversion and supports a broad range of prompt-driven tasks — from structured document parsing and table extraction to chart understanding, document question answering, and key information extraction — all within one model.

The model adopts a multimodal bridging architecture consisting of three components:

  • Vision Encoder: Qianfan-ViT with AnyResolution design (up to 4K), 256 visual tokens per 448×448 tile, max 4,096 tokens per image
  • Language Model: Qwen3-4B with 32K context (extendable to 131K)
  • Cross-Modal Adapter: 2-layer MLP with GELU activation

A key innovation is Layout-as-Thought: an optional thinking phase triggered by <think> tokens, where the model generates structured layout representations (bounding boxes, element types, reading order) before producing final outputs. This is particularly useful for heterogeneous pages with mixed element types (exam papers, technical reports, newspapers).

The model achieves state-of-the-art results on several benchmarks:

  • #1 end-to-end model on OmniDocBench v1.5 with an overall score of 93.12
  • #1 end-to-end model on OlmOCR Bench with a score of 79.8
  • #1 on Key Information Extraction with a mean score of 87.9 across five public KIE benchmarks

This model was contributed by the Baidu Qianfan Team.

Usage example

Document parsing

from transformers import AutoProcessor, AutoModelForImageTextToText

model = AutoModelForImageTextToText.from_pretrained("baidu/Qianfan-OCR", dtype="auto", device_map="auto")
processor = AutoProcessor.from_pretrained("baidu/Qianfan-OCR")

image = "https://huggingface.co/datasets/hf-internal-testing/fixtures_got_ocr/resolve/main/image_ocr.jpg"
messages = [{"role": "user", "content": [{"type": "image", "url": image}, {"type": "text", "text": "Parse this document to Markdown."}]}]

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

generate_ids = model.generate(**inputs, max_new_tokens=64)
processor.decode(generate_ids[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True)

Layout-as-Thought (thinking mode)

For documents with complex layouts, cluttered elements, or non-standard reading orders, enable thinking mode by setting enable_thinking=True in apply_chat_template. The model will first generate structured layout analysis (bounding boxes, element types, reading order), then produce the final output.

from transformers import AutoProcessor, AutoModelForImageTextToText

model = AutoModelForImageTextToText.from_pretrained("baidu/Qianfan-OCR", dtype="auto", device_map="auto")
processor = AutoProcessor.from_pretrained("baidu/Qianfan-OCR")

image = "https://huggingface.co/datasets/hf-internal-testing/fixtures_got_ocr/resolve/main/image_ocr.jpg"
messages = [{"role": "user", "content": [{"type": "image", "url": image}, {"type": "text", "text": "Parse this document to Markdown."}]}]

inputs = processor.apply_chat_template(messages, add_generation_prompt=True, tokenize=True, return_tensors="pt", enable_thinking=True).to(model.device)

generate_ids = model.generate(**inputs, max_new_tokens=128)
processor.decode(generate_ids[0, inputs["input_ids"].shape[1]:], skip_special_tokens=True)

Batched inference

from transformers import AutoProcessor, AutoModelForImageTextToText

model = AutoModelForImageTextToText.from_pretrained("baidu/Qianfan-OCR", dtype="auto", device_map="auto")
processor = AutoProcessor.from_pretrained("baidu/Qianfan-OCR")

image1 = "https://huggingface.co/datasets/hf-internal-testing/fixtures_got_ocr/resolve/main/image_ocr.jpg"
image2 = "https://huggingface.co/datasets/hf-internal-testing/fixtures_got_ocr/resolve/main/multi_box.png"
messages = [
    [{"role": "user", "content": [{"type": "image", "url": image1}, {"type": "text", "text": "Parse this document to Markdown."}]}],
    [{"role": "user", "content": [{"type": "image", "url": image2}, {"type": "text", "text": "OCR the text in the image."}]}],
]

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

generate_ids = model.generate(**inputs, max_new_tokens=64)
processor.batch_decode(generate_ids[:, inputs["input_ids"].shape[1]:], skip_special_tokens=True)

QianfanOCRConfig

class transformers.QianfanOCRConfig

< >

( transformers_version: str | None = None architectures: list[str] | None = None output_hidden_states: bool | None = False return_dict: bool | None = True dtype: typing.Union[str, ForwardRef('torch.dtype'), NoneType] = None chunk_size_feed_forward: int = 0 is_encoder_decoder: bool = False id2label: dict[int, str] | dict[str, str] | None = None label2id: dict[str, int] | dict[str, str] | None = None problem_type: typing.Optional[typing.Literal['regression', 'single_label_classification', 'multi_label_classification']] = None vision_config: dict | transformers.configuration_utils.PreTrainedConfig | None = None text_config: dict | transformers.configuration_utils.PreTrainedConfig | None = None image_token_id: int = 151667 image_seq_length: int = 256 downsample_ratio: float = 0.5 projector_hidden_act: str = 'gelu' vision_feature_layer: int | list[int] = -1 vision_feature_select_strategy: str = 'default' tie_word_embeddings: bool = False )

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_id (int, optional, defaults to 151667) — The image token index used as a placeholder for input images.
  • image_seq_length (int, optional, defaults to 256) — Sequence length of one image embedding.
  • downsample_ratio (float, optional, defaults to 0.5) — Factor by which to downsample the image.
  • projector_hidden_act (str, optional, defaults to gelu) — The activation function used by the multimodal projector.
  • vision_feature_layer (Union[int, list[int]], optional, defaults to -1) — 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.
  • tie_word_embeddings (bool, optional, defaults to False) — Whether to tie weight embeddings according to model’s tied_weights_keys mapping.

This is the configuration class to store the configuration of a QianfanOCRModel. It is used to instantiate a Qianfan Ocr 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 baidu/Qianfan-OCR

Configuration objects inherit from PreTrainedConfig and can be used to control the model outputs. Read the documentation from PreTrainedConfig for more information.

Example:

>>> # Initializing a QianfanOCR style configuration
>>> configuration = QianfanOCRConfig()

>>> # Initializing a model from the configuration
>>> model = QianfanOCRForConditionalGeneration(configuration)

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

QianfanOCRVisionConfig

class transformers.QianfanOCRVisionConfig

< >

( transformers_version: str | None = None architectures: list[str] | None = None output_hidden_states: bool | None = False return_dict: bool | None = True dtype: typing.Union[str, ForwardRef('torch.dtype'), NoneType] = None chunk_size_feed_forward: int = 0 is_encoder_decoder: bool = False id2label: dict[int, str] | dict[str, str] | None = None label2id: dict[str, int] | dict[str, str] | None = None problem_type: typing.Optional[typing.Literal['regression', 'single_label_classification', 'multi_label_classification']] = None hidden_size: int = 1024 num_hidden_layers: int = 24 num_attention_heads: int = 16 attention_bias: bool = True use_qk_norm: bool = False intermediate_size: int = 4096 hidden_act: str = 'gelu' hidden_dropout_prob: float | int = 0.0 attention_dropout: float | int = 0.0 projection_dropout: float | int = 0.0 initializer_range: float = 0.02 norm_type: str = 'layer_norm' layer_norm_eps: float = 1e-06 image_size: int | list[int] | tuple[int, ...] = (448, 448) patch_size: int | list[int] | tuple[int, ...] = (14, 14) num_channels: int = 3 use_mask_token: bool = False use_absolute_position_embeddings: bool = True layer_scale_init_value: float = 0.1 use_mean_pooling: bool = True drop_path_rate: float = 0.1 )

Parameters

  • hidden_size (int, optional, defaults to 1024) — Dimension of the hidden representations.
  • num_hidden_layers (int, optional, defaults to 24) — Number of hidden layers in the Transformer decoder.
  • num_attention_heads (int, optional, defaults to 16) — Number of attention heads for each attention layer in the Transformer decoder.
  • attention_bias (bool, optional, defaults to True) — Whether to use a bias in the query, key, value and output projection layers during self-attention.
  • use_qk_norm (bool, optional, defaults to False) — Whether to use query-key normalization in the attention.
  • intermediate_size (int, optional, defaults to 4096) — Dimension of the MLP representations.
  • hidden_act (str, optional, defaults to gelu) — The non-linear activation function (function or string) in the decoder. For example, "gelu", "relu", "silu", etc.
  • hidden_dropout_prob (Union[float, int], optional, defaults to 0.0) — The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
  • attention_dropout (Union[float, int], optional, defaults to 0.0) — The dropout ratio for the attention probabilities.
  • projection_dropout (float, optional, defaults to 0.0) — Dropout probability for the projection layer.
  • initializer_range (float, optional, defaults to 0.02) — The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
  • norm_type (str, optional, defaults to "layer_norm") — The type of normalization to use in the encoder. Can be "layer_norm" or "rms_norm".
  • layer_norm_eps (float, optional, defaults to 1e-06) — The epsilon used by the layer normalization layers.
  • image_size (Union[int, list[int], tuple[int, ...]], optional, defaults to (448, 448)) — The size (resolution) of each image.
  • patch_size (Union[int, list[int], tuple[int, ...]], optional, defaults to (14, 14)) — The size (resolution) of each patch.
  • num_channels (int, optional, defaults to 3) — The number of input channels.
  • use_mask_token (bool, optional, defaults to False) — Whether to use a mask token for masked image modeling.
  • use_absolute_position_embeddings (bool, optional, defaults to True) — Whether to use absolute position embeddings.
  • layer_scale_init_value (float, optional, defaults to 0.1) — Scale to use in the self-attention layers. 0.1 for base, 1e-6 for large. Set 0 to disable layer scale.
  • use_mean_pooling (bool, optional, defaults to True) — Whether to mean pool the final hidden states of the patches instead of using the final hidden state of the CLS token, before applying the classification head.
  • drop_path_rate (float, optional, defaults to 0.1) — Dropout rate for stochastic depth.

This is the configuration class to store the configuration of a QianfanOCRModel. It is used to instantiate a Qianfan Ocr 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 baidu/Qianfan-OCR

Configuration objects inherit from PreTrainedConfig and can be used to control the model outputs. Read the documentation from PreTrainedConfig for more information.

Example:

>>> # Initializing a QianfanOCR vision style configuration
>>> configuration = QianfanOCRVisionConfig()

>>> # Initializing a model from the configuration
>>> model = QianfanOCRVisionModel(configuration)

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

QianfanOCRProcessor

class transformers.QianfanOCRProcessor

< >

( image_processor = None tokenizer = None image_seq_length: int = 256 chat_template = None image_placeholder_token: str = '<image>' **kwargs )

Parameters

  • image_processor (GotOcr2ImageProcessor) — The image processor is a required input.
  • tokenizer (Qwen2Tokenizer) — The tokenizer is a required input.
  • image_seq_length (int, optional, defaults to 256) — The number of image tokens to be used for each image in the input. Added for backward compatibility but this should be set as a processor attribute in future models.
  • chat_template (str) — A Jinja template to convert lists of messages in a chat into a tokenizable string.
  • image_placeholder_token (str, optional, defaults to ””) — The token emitted by the chat template to mark image positions. It is replaced by the full <img><IMG_CONTEXT>...<IMG_CONTEXT></img> sequence during processing.

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

[QianfanOCRProcessor] offers all the functionalities of [GotOcr2ImageProcessor] and [Qwen2Tokenizer]. See the [~GotOcr2ImageProcessor] and [~Qwen2Tokenizer] for more information.

__call__

< >

( images: typing.Union[ForwardRef('PIL.Image.Image'), numpy.ndarray, ForwardRef('torch.Tensor'), list['PIL.Image.Image'], list[numpy.ndarray], list['torch.Tensor'], NoneType] = None text: str | list[str] | list[list[str]] | None = None videos = None **kwargs ) → BatchFeature

Parameters

  • images (Union[PIL.Image.Image, numpy.ndarray, torch.Tensor, list[PIL.Image.Image], list[numpy.ndarray], list[torch.Tensor]], optional) — Image to preprocess. Expects a single or batch of images with pixel values ranging from 0 to 255. If passing in images with pixel values between 0 and 1, set do_rescale=False.
  • text (Union[str, list[str], list[list[str]]], optional) — The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings (pretokenized string). If you pass a pretokenized input, set is_split_into_words=True to avoid ambiguity with batched inputs.
  • videos (`) -- Video to preprocess. Expects a single or batch of videos with pixel values ranging from 0 to 255. If passing in videos with pixel values between 0 and 1, set do_rescale=False`.
  • return_tensors (str or TensorType, optional) — If set, will return tensors of a particular framework. Acceptable values are:

    • 'pt': Return PyTorch torch.Tensor objects.
    • 'np': Return NumPy np.ndarray objects.

Returns

BatchFeature

A BatchFeature with the following fields:

  • input_ids — List of token ids to be fed to a model. Returned when text is not None.
  • attention_mask — List of indices specifying which tokens should be attended to by the model (when return_attention_mask=True or if “attention_mask” is in self.model_input_names and if text is not None).
  • pixel_values — Pixel values to be fed to a model. Returned when images is not None.

QianfanOCRVisionModel

class transformers.QianfanOCRVisionModel

< >

( config: QianfanOCRVisionConfig )

Parameters

  • config (QianfanOCRVisionConfig) — 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() method to load the model weights.

The bare Qianfan Ocr Model outputting raw hidden-states without any specific head on top.

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 subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

forward

< >

( pixel_values: Tensor bool_masked_pos: torch.BoolTensor | None = None **kwargs: typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs] ) → QianfanOCRVisionModelOutputWithPooling or tuple(torch.FloatTensor)

Parameters

  • pixel_values (torch.Tensor of shape (batch_size, num_channels, image_size, image_size)) — The tensors corresponding to the input images. Pixel values can be obtained using GotOcr2ImageProcessor. See GotOcr2ImageProcessor.__call__() for details (QianfanOCRProcessor uses GotOcr2ImageProcessor for processing images).
  • bool_masked_pos (torch.BoolTensor of shape (batch_size, num_patches), optional) — Boolean masked positions. Indicates which patches are masked (1) and which aren’t (0).

Returns

QianfanOCRVisionModelOutputWithPooling or tuple(torch.FloatTensor)

A QianfanOCRVisionModelOutputWithPooling 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 (QianfanOCRConfig) and inputs.

The QianfanOCRVisionModel 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.

  • pooler_output (torch.FloatTensor of shape (batch_size, hidden_size)) — Average of the last layer hidden states of the patch tokens (excluding the [CLS] token) if config.use_mean_pooling is set to True. If set to False, then the final hidden state of the [CLS] token will be returned.

QianfanOCRModel

class transformers.QianfanOCRModel

< >

( config: QianfanOCRConfig )

Parameters

  • config (QianfanOCRConfig) — 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() method to load the model weights.

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

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 subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

forward

< >

( input_ids: torch.LongTensor | None = None pixel_values: torch.FloatTensor | None = None attention_mask: torch.Tensor | None = None position_ids: torch.LongTensor | None = None past_key_values: transformers.cache_utils.Cache | None = None inputs_embeds: torch.FloatTensor | None = None vision_feature_layer: int | list[int] | None = None vision_feature_select_strategy: str | None = None **kwargs: typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs] ) → QianfanOCRModelOutputWithPast or tuple(torch.FloatTensor)

Parameters

  • 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. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.

    What are 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 GotOcr2ImageProcessor. See GotOcr2ImageProcessor.__call__() for details (QianfanOCRProcessor 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?

  • 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?

  • 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 instance is allowed as input, see our kv cache guide. If no past_key_values are passed, 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).

  • 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) — 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) — The feature selection strategy used to select the vision feature from the vision backbone. Can be one of "default" or "full".

Returns

QianfanOCRModelOutputWithPast or tuple(torch.FloatTensor)

A QianfanOCRModelOutputWithPast 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 (QianfanOCRConfig) and inputs.

The QianfanOCRModel 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.

  • last_hidden_state (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size), optional) — Sequence of hidden-states at the output of the last layer of the model.

  • past_key_values (Cache, optional, returned when use_cache=True is passed or when config.use_cache=True) — It is a Cache instance. For more details, see our kv cache guide.

    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 and after projecting the last hidden state.

QianfanOCRForConditionalGeneration

class transformers.QianfanOCRForConditionalGeneration

< >

( config: QianfanOCRConfig )

Parameters

  • config (QianfanOCRConfig) — 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() method to load the model weights.

The QIANFAN_OCR model which consists of a vision backbone and a language model.

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 subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

forward

< >

( input_ids: torch.LongTensor | None = None pixel_values: torch.FloatTensor | None = None attention_mask: torch.Tensor | None = None position_ids: torch.LongTensor | None = None past_key_values: transformers.cache_utils.Cache | None = None inputs_embeds: torch.FloatTensor | None = None vision_feature_layer: int | list[int] | None = None vision_feature_select_strategy: str | None = None labels: torch.LongTensor | None = None logits_to_keep: int | torch.Tensor = 0 image_sizes: torch.Tensor | None = None **kwargs: typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs] ) → QianfanOCRCausalLMOutputWithPast or tuple(torch.FloatTensor)

Parameters

  • 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. See PreTrainedTokenizer.encode() and PreTrainedTokenizer.call() for details.

    What are 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 GotOcr2ImageProcessor. See GotOcr2ImageProcessor.__call__() for details (QianfanOCRProcessor 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?

  • 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?

  • 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 instance is allowed as input, see our kv cache guide. If no past_key_values are passed, 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).

  • 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) — 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) — The feature selection strategy used to select the vision feature from the vision backbone. Can be one of "default" or "full".
  • 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 (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).
  • image_sizes (torch.Tensor of shape (batch_size, 2), optional) — The sizes of the images in the batch, being (height, width) for each image.

Returns

QianfanOCRCausalLMOutputWithPast or tuple(torch.FloatTensor)

A QianfanOCRCausalLMOutputWithPast 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 (QianfanOCRConfig) and inputs.

The QianfanOCRForConditionalGeneration 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.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 instance. For more details, see our kv cache guide.

    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 and after projecting the last hidden state.

Example:

>>> import torch
>>> from transformers import AutoProcessor, AutoModelForImageTextToText
>>> torch_device = "cuda"
>>> processor = AutoProcessor.from_pretrained("baidu/Qianfan-OCR")
>>> model = AutoModelForImageTextToText.from_pretrained(
...     "baidu/Qianfan-OCR", dtype=torch.bfloat16, device_map=torch_device
... )
>>> messages = [
...     {
...         "role": "user",
...         "content": [
...             {"type": "image", "url": "https://example.com/image.jpg"},
...             {"type": "text", "text": "Describe this image."},
...         ],
...     },
... ]
>>> inputs = processor.apply_chat_template(messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt").to(torch_device)
>>> generate_ids = model.generate(**inputs, max_new_tokens=200)
>>> print(processor.decode(generate_ids[0, inputs["input_ids"].shape[1] :], skip_special_tokens=True))
Update on GitHub

←PPLCNetV3 Qwen2.5-Omni→