transformers 5.0.0rc0__py3-none-any.whl → 5.0.0rc1__py3-none-any.whl

transformers/models/owlv2/modeling_owlv2.py

@@ -793,6 +793,7 @@ class Owlv2TextModel(Owlv2PreTrainedModel):
         output_attentions: Optional[bool] = None,
         output_hidden_states: Optional[bool] = None,
         return_dict: Optional[bool] = None,
+        **kwargs,
     ) -> Union[tuple, BaseModelOutputWithPooling]:
         r"""
         input_ids (`torch.LongTensor` of shape `(batch_size * num_max_text_queries, sequence_length)`):
@@ -903,6 +904,7 @@ class Owlv2VisionModel(Owlv2PreTrainedModel):
         output_hidden_states: Optional[bool] = None,
         interpolate_pos_encoding: bool = False,
         return_dict: Optional[bool] = None,
+        **kwargs,
     ) -> Union[tuple, BaseModelOutputWithPooling]:
         r"""
         Examples:
@@ -1052,6 +1054,7 @@ class Owlv2Model(Owlv2PreTrainedModel):
         interpolate_pos_encoding: bool = False,
         return_base_image_embeds: Optional[bool] = None,
         return_dict: Optional[bool] = None,
+        **kwargs,
     ) -> Union[tuple, Owlv2Output]:
         r"""
         return_loss (`bool`, *optional*):
@@ -1602,6 +1605,7 @@ class Owlv2ForObjectDetection(Owlv2PreTrainedModel):
         output_hidden_states: Optional[bool] = None,
         interpolate_pos_encoding: bool = False,
         return_dict: Optional[bool] = None,
+        **kwargs,
     ) -> Owlv2ObjectDetectionOutput:
         r"""
         input_ids (`torch.LongTensor` of shape `(batch_size * num_max_text_queries, sequence_length)`, *optional*):

transformers/models/owlvit/modeling_owlvit.py

@@ -777,6 +777,7 @@ class OwlViTTextModel(OwlViTPreTrainedModel):
         output_attentions: Optional[bool] = None,
         output_hidden_states: Optional[bool] = None,
         return_dict: Optional[bool] = None,
+        **kwargs,
     ) -> Union[tuple, BaseModelOutputWithPooling]:
         r"""
         input_ids (`torch.LongTensor` of shape `(batch_size * num_max_text_queries, sequence_length)`):
@@ -885,6 +886,7 @@ class OwlViTVisionModel(OwlViTPreTrainedModel):
         output_hidden_states: Optional[bool] = None,
         interpolate_pos_encoding: bool = False,
         return_dict: Optional[bool] = None,
+        **kwargs,
     ) -> Union[tuple, BaseModelOutputWithPooling]:
         r"""
         Examples:
@@ -1033,6 +1035,7 @@ class OwlViTModel(OwlViTPreTrainedModel):
         interpolate_pos_encoding: bool = False,
         return_base_image_embeds: Optional[bool] = None,
         return_dict: Optional[bool] = None,
+        **kwargs,
     ) -> Union[tuple, OwlViTOutput]:
         r"""
         return_loss (`bool`, *optional*):
@@ -1543,6 +1546,7 @@ class OwlViTForObjectDetection(OwlViTPreTrainedModel):
         output_hidden_states: Optional[bool] = None,
         interpolate_pos_encoding: bool = False,
         return_dict: Optional[bool] = None,
+        **kwargs,
     ) -> OwlViTObjectDetectionOutput:
         r"""
         input_ids (`torch.LongTensor` of shape `(batch_size * num_max_text_queries, sequence_length)`, *optional*):

transformers/models/paddleocr_vl/__init__.py

@@ -0,0 +1,32 @@
+# coding=utf-8
+# Copyright 2025 the HuggingFace Team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from typing import TYPE_CHECKING
+
+from ...utils import _LazyModule
+from ...utils.import_utils import define_import_structure
+
+
+if TYPE_CHECKING:
+    from .configuration_paddleocr_vl import *
+    from .image_processing_paddleocr_vl import *
+    from .image_processing_paddleocr_vl_fast import *
+    from .modeling_paddleocr_vl import *
+    from .processing_paddleocr_vl import *
+else:
+    import sys
+
+    _file = globals()["__file__"]
+    sys.modules[__name__] = _LazyModule(__name__, _file, define_import_structure(_file), module_spec=__spec__)

transformers/models/paddleocr_vl/configuration_paddleocr_vl.py

@@ -0,0 +1,336 @@
+#                🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
+#           This file was automatically generated from src/transformers/models/paddleocr_vl/modular_paddleocr_vl.py.
+#               Do NOT edit this file manually as any edits will be overwritten by the generation of
+#             the file from the modular. If any change should be done, please apply the change to the
+#                          modular_paddleocr_vl.py file directly. One of our CI enforces this.
+#                🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨🚨
+# Copyright 2025 The PaddlePaddle Team and The HuggingFace Inc. team. All rights reserved.
+#
+# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
+# and OPT implementations in this library. It has been modified from its
+# original forms to accommodate minor architectural differences compared
+# to GPT-NeoX and OPT used by the Meta AI team that trained the model.
+#
+# 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.
+
+import inspect
+from typing import Optional
+
+from ...configuration_utils import PreTrainedConfig
+from ...modeling_rope_utils import RopeParameters
+
+
+class PaddleOCRVisionConfig(PreTrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`PaddleOCRVisionModel`]. It is used to instantiate a
+    PaddleOCRVL vision encoder according to the specified arguments, defining the model architecture. Instantiating a
+    configuration with the defaults will yield a similar configuration to that of the vision encoder of the PaddleOCRVL
+    [PaddlePaddle/PaddleOCRVL](https://huggingface.co/PaddlePaddle/PaddleOCR-VL) architecture.
+
+    Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PreTrainedConfig`] for more information.
+
+    Args:
+        hidden_size (`int`, *optional*, defaults to 1152):
+            Dimensionality of the encoder layers and the pooler layer.
+        intermediate_size (`int`, *optional*, defaults to 4304):
+            Dimensionality of the "intermediate" (i.e., feed-forward) layer in the Transformer encoder.
+        num_hidden_layers (`int`, *optional*, defaults to 27):
+            Number of hidden layers in the Transformer encoder.
+        num_attention_heads (`int`, *optional*, defaults to 16):
+            Number of attention heads for each attention layer in the Transformer encoder.
+        num_channels (`int`, *optional*, defaults to 3):
+            Number of channels in the input images.
+        image_size (`int`, *optional*, defaults to 384):
+            The size (resolution) of each image.
+        patch_size (`int`, *optional*, defaults to 14):
+            The size (resolution) of each patch.
+        hidden_act (`str` or `function`, *optional*, defaults to `"gelu_pytorch_tanh"`):
+            The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`,
+            `"relu"`, `"selu"` and `"gelu_new"` `"quick_gelu"` are supported.
+        layer_norm_eps (`float`, *optional*, defaults to 1e-06):
+            The epsilon used by the layer normalization layers.
+        attention_dropout (`float`, *optional*, defaults to 0.0):
+            The dropout ratio for the attention probabilities.
+        spatial_merge_size (`int`, *optional*, defaults to 2):
+            The size used for merging spatial dimensions.
+
+    Example:
+
+    ```python
+    >>> from transformers import PaddleOCRVisionConfig, PaddleOCRVisionModel
+
+    >>> # Initializing a PaddleOCRVisionConfig with PaddlePaddle/PaddleOCR-VL style configuration
+    >>> configuration = PaddleOCRVisionConfig()
+
+    >>> # Initializing a PaddleOCRVisionModel (with random weights) from the PaddlePaddle/PaddleOCR-VL style configuration
+    >>> model = PaddleOCRVisionModel(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```
+    """
+
+    model_type = "paddleocr_vl_vision"
+    base_config_key = "vision_config"
+
+    def __init__(
+        self,
+        hidden_size=1152,
+        intermediate_size=4304,
+        num_hidden_layers=27,
+        num_attention_heads=16,
+        num_channels=3,
+        image_size=384,
+        patch_size=14,
+        hidden_act="gelu_pytorch_tanh",
+        layer_norm_eps=1e-6,
+        attention_dropout=0.0,
+        spatial_merge_size=2,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+        self.num_channels = num_channels
+        self.patch_size = patch_size
+        self.image_size = image_size
+        self.attention_dropout = attention_dropout
+        self.layer_norm_eps = layer_norm_eps
+        self.hidden_act = hidden_act
+        self.spatial_merge_size = spatial_merge_size
+
+
+class PaddleOCRTextConfig(PreTrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`PaddleOCRTextModel`]. It is used to instantiate an Ernie 4.5
+    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 Ernie 4.5 0.3B.
+    e.g. [baidu/ERNIE-4.5-0.3B-PT](https://huggingface.co/baidu/ERNIE-4.5-0.3B-PT)
+
+    Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PreTrainedConfig`] for more information.
+
+
+    Args:
+        vocab_size (`int`, *optional*, defaults to 103424):
+            Vocabulary size of the Ernie 4.5 model. Defines the number of different tokens that can be represented by the
+            `inputs_ids` passed when calling [`PaddleOCRTextModel`]
+        hidden_size (`int`, *optional*, defaults to 1024):
+            Dimension of the hidden representations.
+        intermediate_size (`int`, *optional*, defaults to 3072):
+            Dimension of the MLP representations.
+        num_hidden_layers (`int`, *optional*, defaults to 18):
+            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.
+        num_key_value_heads (`int`, *optional*, defaults to 2):
+            This is the number of key_value heads that should be used to implement Grouped Query Attention. If
+            `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if
+            `num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When
+            converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed
+            by meanpooling all the original heads within that group. For more details, check out [this
+            paper](https://huggingface.co/papers/2305.13245). If it is not specified, will default to
+            `num_attention_heads`.
+        hidden_act (`str` or `function`, *optional*, defaults to `"silu"`):
+            The non-linear activation function (function or string) in the decoder.
+        max_position_embeddings (`int`, *optional*, defaults to 131072):
+            The maximum sequence length that this model might ever be used with.
+        initializer_range (`float`, *optional*, defaults to 0.02):
+            The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
+        rms_norm_eps (`float`, *optional*, defaults to 1e-05):
+            The epsilon used by the rms normalization layers.
+        use_cache (`bool`, *optional*, defaults to `True`):
+            Whether or not the model should return the last key/values attentions.
+        pad_token_id (`int`, *optional*, defaults to 0):
+            Padding token id.
+        bos_token_id (`int`, *optional*, defaults to 1):
+            Beginning of stream token id.
+        eos_token_id (`int`, *optional*, defaults to 2):
+            End of stream token id.
+        tie_word_embeddings (`bool`, *optional*, defaults to `True`):
+            Whether to tie weight embeddings
+        rope_parameters (`RopeParameters`, *optional*):
+            Dictionary containing the configuration parameters for the RoPE embeddings. The dictionary should contain
+            a value for `rope_theta` and optionally parameters used for scaling in case you want to use RoPE
+            with longer `max_position_embeddings`.
+        use_bias (`bool`, *optional*, defaults to `False`):
+            Whether to use a bias in any of the projections including mlp and attention for example.
+        head_dim (`int`, *optional*, defaults to 128):
+            The attention head dimension. If None, it will default to hidden_size // num_attention_heads
+
+    ```python
+    >>> from transformers import PaddleOCRTextModel, PaddleOCRTextConfig
+
+    >>> # Initializing a PaddleOCRText 0.3B style configuration
+    >>> configuration = PaddleOCRTextConfig()
+
+    >>> # Initializing a model from the 0.3B style configuration
+    >>> model = PaddleOCRTextModel(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "paddleocr_vl_text"
+    keys_to_ignore_at_inference = ["past_key_values"]
+    default_theta = 500000.0
+    # Default tensor parallel plan for base model `PaddleOCRTextModel`
+    base_model_tp_plan = {
+        "layers.*.self_attn.q_proj": "colwise",
+        "layers.*.self_attn.k_proj": "colwise",
+        "layers.*.self_attn.v_proj": "colwise",
+        "layers.*.self_attn.o_proj": "rowwise",
+        "layers.*.mlp.gate_proj": "colwise",
+        "layers.*.mlp.up_proj": "colwise",
+        "layers.*.mlp.down_proj": "rowwise",
+    }
+    base_model_pp_plan = {
+        "embed_tokens": (["input_ids"], ["inputs_embeds"]),
+        "layers": (["hidden_states", "attention_mask"], ["hidden_states"]),
+        "norm": (["hidden_states"], ["hidden_states"]),
+    }
+
+    def __init__(
+        self,
+        vocab_size: Optional[int] = 103424,
+        hidden_size: Optional[int] = 1024,
+        intermediate_size: Optional[int] = 3072,
+        num_hidden_layers: Optional[int] = 18,
+        num_attention_heads: Optional[int] = 16,
+        num_key_value_heads: Optional[int] = 2,
+        hidden_act: Optional[str] = "silu",
+        max_position_embeddings: Optional[int] = 131072,
+        initializer_range: Optional[float] = 0.02,
+        rms_norm_eps: Optional[int] = 1e-05,
+        use_cache: Optional[int] = True,
+        pad_token_id: Optional[int] = 0,
+        bos_token_id: Optional[int] = 1,
+        eos_token_id: Optional[int] = 2,
+        tie_word_embeddings: Optional[bool] = True,
+        rope_parameters: Optional[RopeParameters | dict[str, RopeParameters]] = None,
+        use_bias: Optional[bool] = False,
+        head_dim: Optional[int] = 128,
+        **kwargs,
+    ):
+        self.vocab_size = vocab_size
+        self.max_position_embeddings = max_position_embeddings
+        self.hidden_size = hidden_size
+        self.intermediate_size = intermediate_size
+        self.num_hidden_layers = num_hidden_layers
+        self.num_attention_heads = num_attention_heads
+
+        # for backward compatibility
+        if num_key_value_heads is None:
+            num_key_value_heads = num_attention_heads
+
+        self.num_key_value_heads = num_key_value_heads
+        self.hidden_act = hidden_act
+        self.initializer_range = initializer_range
+        self.rms_norm_eps = rms_norm_eps
+        self.use_cache = use_cache
+        self.use_bias = use_bias
+        self.head_dim = head_dim if head_dim is not None else self.hidden_size // self.num_attention_heads
+        self.rope_parameters = rope_parameters
+
+        super().__init__(
+            pad_token_id=pad_token_id,
+            bos_token_id=bos_token_id,
+            eos_token_id=eos_token_id,
+            tie_word_embeddings=tie_word_embeddings,
+            **kwargs,
+        )
+
+
+class PaddleOCRVLConfig(PreTrainedConfig):
+    r"""
+    This is the configuration class to store the configuration of a [`PaddleOCRVLForConditionalGeneration`]. It is used to instantiate a
+    PaddleOCRVL model according to the specified arguments, defining the model architecture. Instantiating a configuration
+    with the defaults will yield a similar configuration to that of
+    PaddleOCRVL [PaddlePaddle/PaddleOCR-VL](https://huggingface.co/PaddlePaddle/PaddleOCR-VL).
+
+    Configuration objects inherit from [`PreTrainedConfig`] and can be used to control the model outputs. Read the
+    documentation from [`PreTrainedConfig`] for more information.
+
+
+    Args:
+        text_config (`Union[PreTrainedConfig, dict]`, *optional*, defaults to `PaddleOCRTextConfig`):
+            The config object or dictionary of the text backbone.
+        vision_config (`Union[PreTrainedConfig, dict]`,  *optional*, defaults to `PaddleOCRVisionConfig`):
+            The config object or dictionary of the vision backbone.
+        image_token_id (`int`, *optional*, defaults to 100295):
+            The image token index to encode the image prompt.
+        video_token_id (`int`, *optional*, defaults to 100296):
+            The video token index to encode the image prompt.
+        vision_start_token_id (`int`, *optional*, defaults to 101305):
+            The token index to denote start of vision input.
+        vision_end_token_id (`int`, *optional*, defaults to 101306):
+            The token index to denote end of vision input.
+
+    ```python
+    >>> from transformers import PaddleOCRVLForConditionalGeneration, PaddleOCRVLConfig
+
+    >>> # Initializing a PaddleOCRVL style configuration
+    >>> configuration = PaddleOCRVLConfig()
+
+    >>> # Initializing a model from the PaddleOCRVL style configuration
+    >>> model = PaddleOCRVLForConditionalGeneration(configuration)
+
+    >>> # Accessing the model configuration
+    >>> configuration = model.config
+    ```"""
+
+    model_type = "paddleocr_vl"
+
+    sub_configs = {"vision_config": PaddleOCRVisionConfig, "text_config": PaddleOCRTextConfig}
+    keys_to_ignore_at_inference = ["past_key_values"]
+
+    def __init__(
+        self,
+        text_config=None,
+        vision_config=None,
+        image_token_id=100295,
+        video_token_id=100296,
+        vision_start_token_id=101305,
+        vision_end_token_id=101306,
+        **kwargs,
+    ):
+        if isinstance(vision_config, dict):
+            self.vision_config = self.sub_configs["vision_config"](**vision_config)
+        elif vision_config is None:
+            self.vision_config = self.sub_configs["vision_config"]()
+
+        if isinstance(text_config, dict):
+            self.text_config = self.sub_configs["text_config"](**text_config)
+        elif text_config is None:
+            # Hub configs are saved as flat dicts so we pop some of kwargs to init `TextConfig`
+            text_params = inspect.signature(self.sub_configs["text_config"].__init__).parameters.keys()
+            text_params = list(text_params) + ["rope_scaling", "rope_theta"]
+            text_config = {key: kwargs.pop(key) for key in text_params if key in kwargs}
+            text_config["dtype"] = kwargs.get("torch_dtype", kwargs.get("dtype"))  # don't pop the dtype
+            self.text_config = self.sub_configs["text_config"](**text_config)
+
+        self.image_token_id = image_token_id
+        self.video_token_id = video_token_id
+        self.vision_start_token_id = vision_start_token_id
+        self.vision_end_token_id = vision_end_token_id
+
+        # FIXME: arthur/cyril - tying has to be used from the text config
+        kwargs["tie_word_embeddings"] = self.text_config.tie_word_embeddings
+        super().__init__(**kwargs)
+
+
+__all__ = ["PaddleOCRVLConfig", "PaddleOCRVisionConfig", "PaddleOCRTextConfig"]