From b7e33112809065dafd19bfadfdf6ee9ae7b8110e Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 02:51:06 +0200 Subject: [PATCH 01/35] add florence2 config --- .../florence/configuration_florence2.py | 339 ++++++++++++++++++ 1 file changed, 339 insertions(+) create mode 100644 src/transformers/models/florence/configuration_florence2.py diff --git a/src/transformers/models/florence/configuration_florence2.py b/src/transformers/models/florence/configuration_florence2.py new file mode 100644 index 000000000000..6dd576ee1be5 --- /dev/null +++ b/src/transformers/models/florence/configuration_florence2.py @@ -0,0 +1,339 @@ +# coding=utf-8 +# Copyright 2024 Microsoft and the HuggingFace Inc. 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. +import warnings +""" Florence-2 configuration""" + +from typing import Optional + +from ...configuration_utils import PretrainedConfig +from ...utils import logging + +logger = logging.get_logger(__name__) + +class Florence2VisionConfig(PretrainedConfig): + r""" + This is the configuration class to store the configuration of a [`Florence2VisionModel`]. It is used to instantiate a Florence2VisionModel + according to the specified arguments, defining the model architecture. Instantiating a configuration with the + defaults will yield a similar configuration to that of the Florence2VisionModel architecture. + + Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the + documentation from [`PretrainedConfig`] for more information. + + Args: + drop_path_rate (`float`, *optional*, defaults to 0.1): + The dropout rate of the drop path layer. + patch_size (`List[int]`, *optional*, defaults to [7, 3, 3, 3]): + The patch size of the image. + patch_stride (`List[int]`, *optional*, defaults to [4, 2, 2, 2]): + The patch stride of the image. + patch_padding (`List[int]`, *optional*, defaults to [3, 1, 1, 1]): + The patch padding of the image. + patch_prenorm (`List[bool]`, *optional*, defaults to [false, true, true, true]): + Whether to apply layer normalization before the patch embedding layer. + enable_checkpoint (`bool`, *optional*, defaults to False): + Whether to enable checkpointing. + dim_embed (`List[int]`, *optional*, defaults to [256, 512, 1024, 2048]): + The dimension of the embedding layer. + num_heads (`List[int]`, *optional*, defaults to [8, 16, 32, 64]): + The number of attention heads. + num_groups (`List[int]`, *optional*, defaults to [8, 16, 32, 64]): + The number of groups. + depths (`List[int]`, *optional*, defaults to [1, 1, 9, 1]): + The depth of the model. + window_size (`int`, *optional*, defaults to 12): + The window size of the model. + projection_dim (`int`, *optional*, defaults to 1024): + The dimension of the projection layer. + visual_temporal_embedding (`dict`, *optional*): + The configuration of the visual temporal embedding. + image_pos_embed (`dict`, *optional*): + The configuration of the image position embedding. + image_feature_source (`List[str]`, *optional*, defaults to ["spatial_avg_pool", "temporal_avg_pool"]): + The source of the image feature. + Example: + + ```python + >>> from transformers import Florence2VisionConfig, Florence2VisionModel + + >>> # Initializing a Florence2 Vision style configuration + >>> configuration = Florence2VisionConfig() + + >>> # Initializing a model (with random weights) + >>> model = Florence2VisionModel(configuration) + + >>> # Accessing the model configuration + >>> configuration = model.config + ```""" + + model_type = "florence2_vision" + keys_to_ignore_at_inference = ["past_key_values"] + + def __init__( + self, + drop_path_rate=0.1, + patch_size=[7, 3, 3, 3], + patch_stride=[4, 2, 2, 2], + patch_padding=[3, 1, 1, 1], + patch_prenorm=[False, True, True, True], + enable_checkpoint=False, + dim_embed=[256, 512, 1024, 2048], + num_heads=[8, 16, 32, 64], + num_groups=[8, 16, 32, 64], + depths=[1, 1, 9, 1], + window_size=12, + projection_dim=1024, + visual_temporal_embedding=None, + image_pos_embed=None, + image_feature_source=["spatial_avg_pool", "temporal_avg_pool"], + **kwargs, + ): + self.drop_path_rate = drop_path_rate + self.patch_size = patch_size + self.patch_stride = patch_stride + self.patch_padding = patch_padding + self.patch_prenorm = patch_prenorm + self.enable_checkpoint = enable_checkpoint + self.dim_embed = dim_embed + self.num_heads = num_heads + self.num_groups = num_groups + self.depths = depths + self.window_size = window_size + self.projection_dim = projection_dim + self.visual_temporal_embedding = visual_temporal_embedding + self.image_pos_embed = image_pos_embed + self.image_feature_source = image_feature_source + + super().__init__(**kwargs) + + +class Florence2LanguageConfig(PretrainedConfig): + r""" + This is the configuration class to store the configuration of a [`Florence2LanguagePreTrainedModel`]. It is used to instantiate a BART + 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 BART + [facebook/bart-large](https://huggingface.co/facebook/bart-large) architecture. + + 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 51289): + Vocabulary size of the Florence2Language model. Defines the number of different tokens that can be represented by the + `inputs_ids` passed when calling [`Florence2LanguageModel`]. + d_model (`int`, *optional*, defaults to 1024): + Dimensionality of the layers and the pooler layer. + encoder_layers (`int`, *optional*, defaults to 12): + Number of encoder layers. + decoder_layers (`int`, *optional*, defaults to 12): + Number of decoder layers. + encoder_attention_heads (`int`, *optional*, defaults to 16): + Number of attention heads for each attention layer in the Transformer encoder. + decoder_attention_heads (`int`, *optional*, defaults to 16): + Number of attention heads for each attention layer in the Transformer decoder. + decoder_ffn_dim (`int`, *optional*, defaults to 4096): + Dimensionality of the "intermediate" (often named feed-forward) layer in decoder. + encoder_ffn_dim (`int`, *optional*, defaults to 4096): + Dimensionality of the "intermediate" (often named feed-forward) layer in decoder. + activation_function (`str` or `function`, *optional*, defaults to `"gelu"`): + The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`, + `"relu"`, `"silu"` and `"gelu_new"` are supported. + dropout (`float`, *optional*, defaults to 0.1): + The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. + attention_dropout (`float`, *optional*, defaults to 0.0): + The dropout ratio for the attention probabilities. + activation_dropout (`float`, *optional*, defaults to 0.0): + The dropout ratio for activations inside the fully connected layer. + classifier_dropout (`float`, *optional*, defaults to 0.0): + The dropout ratio for classifier. + max_position_embeddings (`int`, *optional*, defaults to 1024): + The maximum sequence length that this model might ever be used with. Typically set this to something large + just in case (e.g., 512 or 1024 or 2048). + init_std (`float`, *optional*, defaults to 0.02): + The standard deviation of the truncated_normal_initializer for initializing all weight matrices. + encoder_layerdrop (`float`, *optional*, defaults to 0.0): + The LayerDrop probability for the encoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556) + for more details. + decoder_layerdrop (`float`, *optional*, defaults to 0.0): + The LayerDrop probability for the decoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556) + for more details. + scale_embedding (`bool`, *optional*, defaults to `False`): + Scale embeddings by diving by sqrt(d_model). + use_cache (`bool`, *optional*, defaults to `True`): + Whether or not the model should return the last key/values attentions (not used by all models). + num_labels (`int`, *optional*, defaults to 3): + The number of labels to use in [`Florence2LanguageForSequenceClassification`]. + forced_eos_token_id (`int`, *optional*, defaults to 2): + The id of the token to force as the last generated token when `max_length` is reached. Usually set to + `eos_token_id`. + + Example: + + ```python + >>> from transformers import Florence2LanguageConfig, Florence2LanguageModel + + >>> # Initializing a Florence2 Language style configuration + >>> configuration = Florence2LanguageConfig() + + >>> # Initializing a model (with random weights) + >>> model = Florence2LangaugeModel(configuration) + + >>> # Accessing the model configuration + >>> configuration = model.config + ```""" + + model_type = "florence2_language" + keys_to_ignore_at_inference = ["past_key_values"] + attribute_map = {"num_attention_heads": "encoder_attention_heads", "hidden_size": "d_model"} + + def __init__( + self, + vocab_size=51289, + max_position_embeddings=1024, + encoder_layers=12, + encoder_ffn_dim=4096, + encoder_attention_heads=16, + decoder_layers=12, + decoder_ffn_dim=4096, + decoder_attention_heads=16, + encoder_layerdrop=0.0, + decoder_layerdrop=0.0, + activation_function="gelu", + d_model=1024, + dropout=0.1, + attention_dropout=0.0, + activation_dropout=0.0, + init_std=0.02, + classifier_dropout=0.0, + scale_embedding=False, + use_cache=True, + num_labels=3, + pad_token_id=1, + bos_token_id=0, + eos_token_id=2, + is_encoder_decoder=True, + decoder_start_token_id=2, + forced_eos_token_id=2, + **kwargs, + ): + self.vocab_size = vocab_size + self.max_position_embeddings = max_position_embeddings + self.d_model = d_model + self.encoder_ffn_dim = encoder_ffn_dim + self.encoder_layers = encoder_layers + self.encoder_attention_heads = encoder_attention_heads + self.decoder_ffn_dim = decoder_ffn_dim + self.decoder_layers = decoder_layers + self.decoder_attention_heads = decoder_attention_heads + self.dropout = dropout + self.attention_dropout = attention_dropout + self.activation_dropout = activation_dropout + self.activation_function = activation_function + self.init_std = init_std + self.encoder_layerdrop = encoder_layerdrop + self.decoder_layerdrop = decoder_layerdrop + self.classifier_dropout = classifier_dropout + self.use_cache = use_cache + self.num_hidden_layers = encoder_layers + self.scale_embedding = scale_embedding # scale factor will be sqrt(d_model) if True + + super().__init__( + num_labels=num_labels, + pad_token_id=pad_token_id, + bos_token_id=bos_token_id, + eos_token_id=eos_token_id, + is_encoder_decoder=is_encoder_decoder, + decoder_start_token_id=decoder_start_token_id, + forced_eos_token_id=forced_eos_token_id, + **kwargs, + ) + + # ensure backward compatibility for BART CNN models + if self.forced_bos_token_id is None and kwargs.get("force_bos_token_to_be_generated", False): + self.forced_bos_token_id = self.bos_token_id + warnings.warn( + f"Please make sure the config includes `forced_bos_token_id={self.bos_token_id}` in future versions. " + "The config can simply be saved and uploaded again to be fixed." + ) + + +class Florence2Config(PretrainedConfig): + r""" + This is the configuration class to store the configuration of a [`Florence2ForConditionalGeneration`]. It is used to instantiate an + Florence-2 model according to the specified arguments, defining the model architecture. + + 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 (`Florence2VisionConfig`, *optional*): + Custom vision config or dict + text_config (`Union[AutoConfig, dict]`, *optional*): + The config object of the text backbone. + ignore_index (`int`, *optional*, defaults to -100): + The ignore index for the loss function. + vocab_size (`int`, *optional*, defaults to 51289): + Vocabulary size of the Florence2model. Defines the number of different tokens that can be represented by the + `inputs_ids` passed when calling [`~Florence2ForConditionalGeneration`] + projection_dim (`int`, *optional*, defaults to 1024): + Dimension of the multimodal projection space. + + Example: + + ```python + >>> from transformers import Florence2ForConditionalGeneration, Florence2Config, CLIPVisionConfig, BartConfig + + >>> # Initializing a clip-like vision config + >>> vision_config = CLIPVisionConfig() + + >>> # Initializing a Bart config + >>> text_config = BartConfig() + + >>> # Initializing a Florence-2 configuration + >>> configuration = Florence2Config(vision_config, text_config) + + >>> # Initializing a model from the florence-2 configuration + >>> model = Florence2ForConditionalGeneration(configuration) + + >>> # Accessing the model configuration + >>> configuration = model.config + ```""" + + model_type = "florence2" + is_composition = False + + def __init__( + self, + vision_config=None, + text_config=None, + ignore_index=-100, + vocab_size=51289, + projection_dim=1024, + **kwargs, + ): + self.ignore_index = ignore_index + self.vocab_size = vocab_size + self.projection_dim = projection_dim + if vision_config is not None: + vision_config = PretrainedConfig(**vision_config) + self.vision_config = vision_config + self.vocab_size = self.vocab_size + + self.text_config = text_config + if text_config is not None: + self.text_config = Florence2LanguageConfig(**text_config) + + + super().__init__(**kwargs) + From 3adc633745ece2f10ba8a51d7ae9b8023c0c6aca Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 02:54:26 +0200 Subject: [PATCH 02/35] add florence2 processing --- .../models/florence/processing_florence2.py | 1088 +++++++++++++++++ 1 file changed, 1088 insertions(+) create mode 100644 src/transformers/models/florence/processing_florence2.py diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py new file mode 100644 index 000000000000..90797a3a242d --- /dev/null +++ b/src/transformers/models/florence/processing_florence2.py @@ -0,0 +1,1088 @@ +# coding=utf-8 +# Copyright 2024 Microsoft and 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. +""" +Processor class for Florence-2. +""" + +import re +import logging +from typing import List, Optional, Union +import numpy as np + +import torch + +from ...feature_extraction_utils import BatchFeature +from ...image_utils import ImageInput, is_valid_image +from ...processing_utils import ProcessorMixin +from ...tokenization_utils_base import ( + PaddingStrategy, + PreTokenizedInput, + TextInput, + TruncationStrategy, +) +from ...utils import TensorType + + +logger = logging.getLogger(__name__) + +# Copied from transformers.models.idefics2.processing_idefics2.is_url +def is_url(val) -> bool: + return isinstance(val, str) and val.startswith("http") + +# Copied from transformers.models.idefics2.processing_idefics2.is_image_or_image_url +def is_image_or_image_url(elem): + return is_url(elem) or is_valid_image(elem) + + +def _is_str_or_image(elem): + return isinstance(elem, (str)) or is_image_or_image_url(elem) + + +class Florence2Processor(ProcessorMixin): + r""" + Constructs a Florence2 processor which wraps a Florence2 image processor and a Florence2 tokenizer into a single processor. + + [`Florence2Processor`] offers all the functionalities of [`CLIPImageProcessor`] and [`BartTokenizerFast`]. See the + [`~Florence2Processor.__call__`] and [`~Florence2Processor.decode`] for more information. + + Args: + image_processor ([`CLIPImageProcessor`], *optional*): + The image processor is a required input. + tokenizer ([`BartTokenizerFast`], *optional*): + The tokenizer is a required input. + """ + + attributes = ["image_processor", "tokenizer"] + image_processor_class = "CLIPImageProcessor" + tokenizer_class = ("BartTokenizer", "BartTokenizerFast") + + def __init__( + self, + image_processor=None, + tokenizer=None, + ): + if image_processor is None: + raise ValueError("You need to specify an `image_processor`.") + if tokenizer is None: + raise ValueError("You need to specify a `tokenizer`.") + if not hasattr(image_processor, "image_seq_length"): + raise ValueError("Image processor is missing an `image_seq_length` attribute.") + + self.image_seq_length = image_processor.image_seq_length + + tokens_to_add = { + 'additional_special_tokens': \ + tokenizer.additional_special_tokens + \ + ['', '', '', ''] + \ + [f'' for x in range(1000)] + \ + ['', '', '', '','', '', '', '', '', '', '', '', '', '', '', '', '', '', '', ''] + } + tokenizer.add_special_tokens(tokens_to_add) + + self.tasks_answer_post_processing_type = { + '': 'pure_text', + '': 'ocr', + '': 'pure_text', + '': 'pure_text', + '': 'pure_text', + '': 'description_with_bboxes', + '': 'description_with_bboxes', + '': "phrase_grounding", + '': 'polygons', + '': 'polygons', + '': 'description_with_bboxes_or_polygons', + '': 'pure_text', + '': 'pure_text', + '': 'pure_text', + '': 'bboxes' + } + + self.task_prompts_without_inputs = { + '': 'What is the text in the image?', + '': 'What is the text in the image, with regions?', + '': 'What does the image describe?', + '': 'Describe in detail what is shown in the image.', + '': 'Describe with a paragraph what is shown in the image.', + '': 'Locate the objects with category name in the image.', + '': 'Locate the objects in the image, with their descriptions.', + '': 'Locate the region proposals in the image.' + } + + self.task_prompts_with_input = { + '': "Locate the phrases in the caption: {input}", + '': 'Locate {input} in the image with mask', + '': 'What is the polygon mask of region {input}', + '': 'Locate {input} in the image.', + '': 'What is the region {input}?', + '': 'What does the region {input} describe?', + '': 'What text is in the region {input}?', + } + + self.post_processor = Florence2PostProcesser(tokenizer=tokenizer) + + + super().__init__(image_processor, tokenizer) + + def _construct_prompts(self, text): + # replace the task tokens with the task prompts if task token is in the text + prompts = [] + for _text in text: + # 1. fixed task prompts without additional inputs + for task_token, task_prompt in self.task_prompts_without_inputs.items(): + if task_token in _text: + assert _text == task_token, f"Task token {task_token} should be the only token in the text." + _text = task_prompt + break + # 2. task prompts with additional inputs + for task_token, task_prompt in self.task_prompts_with_input.items(): + if task_token in _text: + _text = task_prompt.format(input=_text.replace(task_token, '')) + break + prompts.append(_text) + return prompts + + def __call__( + self, + text: Union[TextInput, PreTokenizedInput, List[TextInput], List[PreTokenizedInput]] = None, + images: ImageInput = None, + tokenize_newline_separately: bool = True, + padding: Union[bool, str, PaddingStrategy] = False, + truncation: Union[bool, str, TruncationStrategy] = None, + max_length=None, + return_tensors: Optional[Union[str, TensorType]] = TensorType.PYTORCH, + do_resize: bool = None, + do_normalize: bool = None, + image_mean: Optional[Union[float, List[float]]] = None, + image_std: Optional[Union[float, List[float]]] = None, + data_format: Optional["ChannelDimension"] = "channels_first", # noqa: F821 + input_data_format: Optional[ + Union[str, "ChannelDimension"] # noqa: F821 + ] = None, + resample: "PILImageResampling" = None, # noqa: F821 + do_convert_rgb: bool = None, + do_thumbnail: bool = None, + do_align_long_axis: bool = None, + do_rescale: bool = None, + ) -> BatchFeature: + """ + Main method to prepare for the model one or several sequences(s) and image(s). This method forwards the `text` + and `kwargs` arguments to BartTokenizerFast's [`~BartTokenizerFast.__call__`] if `text` is not `None` to encode + the text. To prepare the image(s), this method forwards the `images` and `kwrags` arguments to + CLIPImageProcessor's [`~CLIPImageProcessor.__call__`] if `images` is not `None`. Please refer to the doctsring + of the above two methods for more information. + + Args: + text (`str`, `List[str]`, `List[List[str]]`): + The sequence or batch of sequences to be encoded. Each sequence can be a string or a list of strings + (pretokenized string). If the sequences are provided as list of strings (pretokenized), you must set + `is_split_into_words=True` (to lift the ambiguity with a batch of sequences). + images (`PIL.Image.Image`, `np.ndarray`, `torch.Tensor`, `List[PIL.Image.Image]`, `List[np.ndarray]`, `List[torch.Tensor]`): + The image or batch of images to be prepared. Each image can be a PIL image, NumPy array or PyTorch + tensor. In case of a NumPy array/PyTorch tensor, each image should be of shape (C, H, W), where C is a + number of channels, H and W are image height and width. + tokenize_newline_separately (`bool`, defaults to `True`): + Adds a separately tokenized '\n' at the end of the prompt. + padding (`bool`, `str` or [`~utils.PaddingStrategy`], *optional*, defaults to `False`): + Select a strategy to pad the returned sequences (according to the model's padding side and padding + index) among: + - `True` or `'longest'`: Pad to the longest sequence in the batch (or no padding if only a single + sequence if provided). + - `'max_length'`: Pad to a maximum length specified with the argument `max_length` or to the maximum + acceptable input length for the model if that argument is not provided. + - `False` or `'do_not_pad'` (default): No padding (i.e., can output a batch with sequences of different + lengths). + max_length (`int`, *optional*): + Maximum length of the returned list and optionally padding length (see above). + truncation (`bool`, *optional*): + Activates truncation to cut input sequences longer than `max_length` to `max_length`. + return_tensors (`str` or [`~utils.TensorType`], *optional*): + If set, will return tensors of a particular framework. Acceptable values are: + + - `'tf'`: Return TensorFlow `tf.constant` objects. + - `'pt'`: Return PyTorch `torch.Tensor` objects. + - `'np'`: Return NumPy `np.ndarray` objects. + - `'jax'`: Return JAX `jnp.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`. If `suffix` + is provided, the `input_ids` will also contain the suffix input ids. + - **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`. + - **labels** -- Labels compatible with training if `suffix` is not None + """ + + return_token_type_ids = False + + if images is None: + raise ValueError("`images` are expected as arguments to a `Florence2Processor` instance.") + if text is None: + logger.warning_once( + "You are using Florence-2 without a text prompt." + ) + text = "" + + if isinstance(text, List) and isinstance(images, List): + if len(images) < len(text): + raise ValueError( + f"Received {len(images)} images for {len(text)} prompts. Each prompt should be associated with an image." + ) + if _is_str_or_image(text): + text = [text] + elif isinstance(text, list) and _is_str_or_image(text[0]): + pass + + pixel_values = self.image_processor( + images, + do_resize=do_resize, + do_normalize=do_normalize, + return_tensors=return_tensors, + image_mean=image_mean, + image_std=image_std, + input_data_format=input_data_format, + data_format=data_format, + resample=resample, + do_convert_rgb=do_convert_rgb, + )["pixel_values"] + + if max_length is not None: + max_length -= self.image_seq_length # max_length has to account for the image tokens + + text = self._construct_prompts(text) + + inputs = self.tokenizer( + text, + return_tensors=return_tensors, + padding=padding, + max_length=max_length, + truncation=truncation, + return_token_type_ids=return_token_type_ids, + ) + + return_data = {**inputs, "pixel_values": pixel_values} + + if return_token_type_ids: + labels = inputs["input_ids"].masked_fill(inputs["token_type_ids"] == 0, -100) + return_data.update({"labels": labels}) + return BatchFeature(data=return_data) + + # Copied from transformers.models.clip.processing_clip.CLIPProcessor.batch_decode with CLIP->Florence2 + def batch_decode(self, *args, **kwargs): + """ + This method forwards all its arguments to BartTokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please + refer to the docstring of this method for more information. + """ + return self.tokenizer.batch_decode(*args, **kwargs) + + # Copied from transformers.models.clip.processing_clip.CLIPProcessor.decode with CLIP->Florence2 + def decode(self, *args, **kwargs): + """ + This method forwards all its arguments to BartTokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to + the docstring of this method for more information. + """ + return self.tokenizer.decode(*args, **kwargs) + + @property + # Copied from transformers.models.clip.processing_clip.CLIPProcessor.model_input_names with CLIP->Florence2 + def model_input_names(self): + tokenizer_input_names = self.tokenizer.model_input_names + image_processor_input_names = self.image_processor.model_input_names + return list(dict.fromkeys(tokenizer_input_names + image_processor_input_names)) + + def post_process_generation(self, text, task, image_size): + """ + Post-process the output of the model to each of the task outputs. + + Args: + text (`str`): The text to post-process. + task (`str`): The task to post-process the text for. + image_size (`Tuple[int, int]`): The size of the image. height x width. + """ + + task_answer_post_processing_type = self.tasks_answer_post_processing_type.get(task, 'pure_text') + task_answer = self.post_processor( + text=text, + image_size=image_size, + parse_tasks=task_answer_post_processing_type, + )[task_answer_post_processing_type] + + if task_answer_post_processing_type == 'pure_text': + final_answer = task_answer + # remove the special tokens + final_answer = final_answer.replace('', '').replace('', '') + elif task_answer_post_processing_type in ['od', 'description_with_bboxes', 'bboxes']: + od_instances = task_answer + bboxes_od = [_od_instance['bbox'] for _od_instance in od_instances] + labels_od = [str(_od_instance['cat_name']) for _od_instance in od_instances] + final_answer = {'bboxes': bboxes_od, 'labels': labels_od} + elif task_answer_post_processing_type in ['ocr']: + bboxes = [_od_instance['quad_box'] for _od_instance in task_answer] + labels = [str(_od_instance['text']) for _od_instance in task_answer] + final_answer = {'quad_boxes': bboxes, 'labels': labels} + elif task_answer_post_processing_type in ['phrase_grounding']: + bboxes = [] + labels = [] + for _grounded_phrase in task_answer: + for _bbox in _grounded_phrase['bbox']: + bboxes.append(_bbox) + labels.append(_grounded_phrase['cat_name']) + final_answer = {'bboxes': bboxes, 'labels': labels} + elif task_answer_post_processing_type in ['description_with_polygons', 'polygons']: + labels = [] + polygons = [] + for result in task_answer: + label = result['cat_name'] + _polygons = result['polygons'] + labels.append(label) + polygons.append(_polygons) + final_answer = {'polygons': polygons, 'labels': labels} + elif task_answer_post_processing_type in ['description_with_bboxes_or_polygons']: + bboxes = [] + bboxes_labels = [] + polygons = [] + polygons_labels = [] + for result in task_answer: + label = result['cat_name'] + if 'polygons' in result: + _polygons = result['polygons'] + polygons.append(_polygons) + polygons_labels.append(label) + else: + _bbox = result['bbox'] + bboxes.append(_bbox) + bboxes_labels.append(label) + final_answer = {'bboxes': bboxes, 'bboxes_labels': bboxes_labels, 'polygons': polygons, 'polygons_labels': polygons_labels} + else: + raise ValueError('Unknown task answer post processing type: {}'.format(task_answer_post_processing_type)) + + final_answer = { + task: final_answer} + return final_answer + +class BoxQuantizer(object): + def __init__(self, mode, bins): + self.mode = mode + self.bins = bins + + def quantize(self, boxes: torch.Tensor, size): + bins_w, bins_h = self.bins # Quantization bins. + size_w, size_h = size # Original image size. + size_per_bin_w = size_w / bins_w + size_per_bin_h = size_h / bins_h + xmin, ymin, xmax, ymax = boxes.split(1, dim=-1) # Shape: 4 * [N, 1]. + + if self.mode == 'floor': + quantized_xmin = ( + xmin / size_per_bin_w).floor().clamp(0, bins_w - 1) + quantized_ymin = ( + ymin / size_per_bin_h).floor().clamp(0, bins_h - 1) + quantized_xmax = ( + xmax / size_per_bin_w).floor().clamp(0, bins_w - 1) + quantized_ymax = ( + ymax / size_per_bin_h).floor().clamp(0, bins_h - 1) + + elif self.mode == 'round': + raise NotImplementedError() + + else: + raise ValueError('Incorrect quantization type.') + + quantized_boxes = torch.cat( + (quantized_xmin, quantized_ymin, quantized_xmax, quantized_ymax), dim=-1 + ).int() + + return quantized_boxes + + def dequantize(self, boxes: torch.Tensor, size): + bins_w, bins_h = self.bins # Quantization bins. + size_w, size_h = size # Original image size. + size_per_bin_w = size_w / bins_w + size_per_bin_h = size_h / bins_h + xmin, ymin, xmax, ymax = boxes.split(1, dim=-1) # Shape: 4 * [N, 1]. + + if self.mode == 'floor': + # Add 0.5 to use the center position of the bin as the coordinate. + dequantized_xmin = (xmin + 0.5) * size_per_bin_w + dequantized_ymin = (ymin + 0.5) * size_per_bin_h + dequantized_xmax = (xmax + 0.5) * size_per_bin_w + dequantized_ymax = (ymax + 0.5) * size_per_bin_h + + elif self.mode == 'round': + raise NotImplementedError() + + else: + raise ValueError('Incorrect quantization type.') + + dequantized_boxes = torch.cat( + (dequantized_xmin, dequantized_ymin, + dequantized_xmax, dequantized_ymax), dim=-1 + ) + + return dequantized_boxes + + +class CoordinatesQuantizer(object): + """ + Quantize coornidates (Nx2) + """ + + def __init__(self, mode, bins): + self.mode = mode + self.bins = bins + + def quantize(self, coordinates: torch.Tensor, size): + bins_w, bins_h = self.bins # Quantization bins. + size_w, size_h = size # Original image size. + size_per_bin_w = size_w / bins_w + size_per_bin_h = size_h / bins_h + assert coordinates.shape[-1] == 2, 'coordinates should be shape (N, 2)' + x, y = coordinates.split(1, dim=-1) # Shape: 4 * [N, 1]. + + if self.mode == 'floor': + quantized_x = (x / size_per_bin_w).floor().clamp(0, bins_w - 1) + quantized_y = (y / size_per_bin_h).floor().clamp(0, bins_h - 1) + + elif self.mode == 'round': + raise NotImplementedError() + + else: + raise ValueError('Incorrect quantization type.') + + quantized_coordinates = torch.cat( + (quantized_x, quantized_y), dim=-1 + ).int() + + return quantized_coordinates + + def dequantize(self, coordinates: torch.Tensor, size): + bins_w, bins_h = self.bins # Quantization bins. + size_w, size_h = size # Original image size. + size_per_bin_w = size_w / bins_w + size_per_bin_h = size_h / bins_h + assert coordinates.shape[-1] == 2, 'coordinates should be shape (N, 2)' + x, y = coordinates.split(1, dim=-1) # Shape: 4 * [N, 1]. + + if self.mode == 'floor': + # Add 0.5 to use the center position of the bin as the coordinate. + dequantized_x = (x + 0.5) * size_per_bin_w + dequantized_y = (y + 0.5) * size_per_bin_h + + elif self.mode == 'round': + raise NotImplementedError() + + else: + raise ValueError('Incorrect quantization type.') + + dequantized_coordinates = torch.cat( + (dequantized_x, dequantized_y), dim=-1 + ) + + return dequantized_coordinates + + +class Florence2PostProcesser(object): + """ + Florence-2 post process for converting text prediction to various tasks results. + + Args: + config: A dict of configs. + tokenizer: A tokenizer for decoding text to spans. + sample config: + UNIFIED_POST_PROCESS: + # commom configs + NUM_BBOX_HEIGHT_BINS: 1000 + NUM_BBOX_WIDTH_BINS: 1000 + COORDINATES_HEIGHT_BINS: 1000 + COORDINATES_WIDTH_BINS: 1000 + # task specific configs, override the common configs + PRASE_TASKS: + - TASK_NAME: 'video_dense_caption' + PATTERN: 'r([a-zA-Z0-9 ]+)' + SCORE_MODE: 'avg_cat_name_scores' + NUM_BINS: 100 + - TASK_NAME: 'od' + PATTERN: 'r([a-zA-Z0-9 ]+)' + SCORE_MODE: 'avg_cat_name_scores' + + Returns: + parsed_dict (dict): A dict of parsed results. + """ + def __init__( + self, + tokenizer=None + ): + parse_tasks = [] + parse_task_configs = {} + config = self._create_default_config() + for task in config['PARSE_TASKS']: + parse_tasks.append(task['TASK_NAME']) + parse_task_configs[task['TASK_NAME']] = task + + self.config = config + self.parse_tasks = parse_tasks + self.parse_tasks_configs = parse_task_configs + + self.tokenizer = tokenizer + if self.tokenizer is not None: + self.all_special_tokens = set(self.tokenizer.all_special_tokens) + + self.init_quantizers() + self.black_list_of_phrase_grounding = self._create_black_list_of_phrase_grounding() + + def _create_black_list_of_phrase_grounding(self): + black_list = {} + + if 'phrase_grounding' in self.parse_tasks and self.parse_tasks_configs['phrase_grounding']['FILTER_BY_BLACK_LIST']: + black_list = set( + ['it', 'I', 'me', 'mine', + 'you', 'your', 'yours', + 'he', 'him', 'his', + 'she', 'her', 'hers', + 'they', 'them', 'their', 'theirs', + 'one', 'oneself', + 'we', 'us', 'our', 'ours', + 'you', 'your', 'yours', + 'they', 'them', 'their', 'theirs', + 'mine', 'yours', 'his', 'hers', 'its', + 'ours', 'yours', 'theirs', + 'myself', 'yourself', 'himself', 'herself', 'itself', + 'ourselves', 'yourselves', 'themselves', + 'this', 'that', + 'these', 'those', + 'who', 'whom', 'whose', 'which', 'what', + 'who', 'whom', 'whose', 'which', 'that', + 'all', 'another', 'any', 'anybody', 'anyone', 'anything', + 'each', 'everybody', 'everyone', 'everything', + 'few', 'many', 'nobody', 'none', 'one', 'several', + 'some', 'somebody', 'someone', 'something', + 'each other', 'one another', + 'myself', 'yourself', 'himself', 'herself', 'itself', + 'ourselves', 'yourselves', 'themselves', + 'the image', 'image', 'images', 'the', 'a', 'an', 'a group', + 'other objects', 'lots', 'a set', + ] + ) + + return black_list + + def _create_default_config(self): + config = { + 'NUM_BBOX_HEIGHT_BINS': 1000, + 'NUM_BBOX_WIDTH_BINS': 1000, + 'BOX_QUANTIZATION_MODE': 'floor', + 'COORDINATES_HEIGHT_BINS': 1000, + 'COORDINATES_WIDTH_BINS': 1000, + 'COORDINATES_QUANTIZATION_MODE': 'floor', + 'PARSE_TASKS': [ + { + 'TASK_NAME': 'od', + 'PATTERN': r'([a-zA-Z0-9 ]+)' + }, + { + 'TASK_NAME': 'ocr', + 'PATTERN': r'(.+?)', + 'AREA_THRESHOLD': 0.01 + }, + { + 'TASK_NAME': 'phrase_grounding', + 'FILTER_BY_BLACK_LIST': True + }, + { + 'TASK_NAME': 'pure_text', + }, + { + 'TASK_NAME': 'description_with_bboxes', + }, + { + 'TASK_NAME': 'description_with_polygons', + }, + { + 'TASK_NAME': 'polygons', + }, + { + 'TASK_NAME': 'bboxes', + }, + { + 'TASK_NAME': 'description_with_bboxes_or_polygons', + } + ] + } + + return config + + def init_quantizers(self): + # we have box_quantizer (od, grounding) and coordinates_quantizer (ocr, referring_segmentation) + num_bbox_height_bins = self.config.get('NUM_BBOX_HEIGHT_BINS', 1000) + num_bbox_width_bins = self.config.get('NUM_BBOX_WIDTH_BINS', 1000) + box_quantization_mode = self.config.get('BOX_QUANTIZATION_MODE', 'floor') + self.box_quantizer = BoxQuantizer( + box_quantization_mode, + (num_bbox_width_bins, num_bbox_height_bins), + ) + + num_bbox_height_bins = self.config['COORDINATES_HEIGHT_BINS'] if 'COORDINATES_HEIGHT_BINS' in self.config else self.config.get('NUM_BBOX_HEIGHT_BINS', 1000) + num_bbox_width_bins = self.config['COORDINATES_WIDTH_BINS'] if 'COORDINATES_WIDTH_BINS' in self.config else self.config.get('NUM_BBOX_WIDTH_BINS', 1000) + box_quantization_mode = self.config.get('COORDINATES_QUANTIZATION_MODE') if 'COORDINATES_QUANTIZATION_MODE' in self.config else self.config.get('BOX_QUANTIZATION_MODE', 'floor') + self.coordinates_quantizer = CoordinatesQuantizer( + box_quantization_mode, + (num_bbox_width_bins, num_bbox_height_bins), + ) + + def decode_with_spans(self, tokenizer, token_ids): + filtered_tokens = tokenizer.convert_ids_to_tokens( + token_ids, skip_special_tokens=False) + assert len(filtered_tokens) == len(token_ids) + + # To avoid mixing byte-level and unicode for byte-level BPT + # we need to build string separately for added tokens and byte-level tokens + # cf. https://github.com/huggingface/transformers/issues/1133 + sub_texts = [] + for token in filtered_tokens: + if token in self.all_special_tokens: + sub_texts.append(token) + else: + if isinstance(tokenizer, (BartTokenizer, BartTokenizerFast)): + sub_text = tokenizer.convert_tokens_to_string([token]) + elif isinstance(tokenizer, (T5Tokenizer, T5TokenizerFast)): + # Ref: https://github.com/google/sentencepiece#whitespace-is-treated-as-a-basic-symbol + # Note: Do not strip sub_text as it may have functional whitespace + sub_text = token.replace('▁', ' ') + else: + raise ValueError(f'type {type(tokenizer)} not supported') + sub_texts.append(sub_text) + + text = '' + spans = [] + for sub_text in sub_texts: + span = (len(text), len(text) + len(sub_text)) # [start index, end index). + text += sub_text + spans.append(span) + + # Text format: + # 1. T5Tokenizer/T5TokenizerFast: + # " transplanting dog cat" + # Equivalent to t5_tokenizer.decode(input_ids, skip_special_tokens=False, clean_up_tokenization_spaces=False, spaces_between_special_tokens=False) + # 2. BartTokenizer (need to double check): + # "transplanting dogcat" + # Equivalent to bart_tokenizer.decode(input_ids, skip_special_tokens=False, clean_up_tokenization_spaces=False, spaces_between_special_tokens=False) + return text, spans + + def parse_od_from_text_and_spans( + self, + text, + pattern, + image_size, + phrase_centric=False + ): + parsed = list(re.finditer(pattern, text)) + + instances = [] + for i in range(len(parsed)): + # Prepare instance. + instance = {} + + if phrase_centric: + bbox_bins = [int(parsed[i].group(j)) for j in range(2, 6)] + else: + bbox_bins = [int(parsed[i].group(j)) for j in range(1, 5)] + instance['bbox'] = self.box_quantizer.dequantize( + boxes=torch.tensor(bbox_bins), + size=image_size + ).tolist() + + if phrase_centric: + instance['cat_name'] = parsed[i].group(1).lower().strip() + else: + instance['cat_name'] = parsed[i].group(5).lower().strip() + instances.append(instance) + + return instances + + def parse_ocr_from_text_and_spans(self, + text, + pattern, + image_size, + area_threshold=-1.0, + ): + bboxes = [] + labels = [] + text = text.replace('', '') + # ocr with regions + parsed = re.findall(pattern, text) + instances = [] + image_width, image_height = image_size + + for ocr_line in parsed: + ocr_content = ocr_line[0] + quad_box = ocr_line[1:] + quad_box = [int(i) for i in quad_box] + quad_box = self.coordinates_quantizer.dequantize( + torch.tensor(np.array(quad_box).reshape(-1, 2)), + size=image_size + ).reshape(-1).tolist() + + if area_threshold > 0: + x_coords = [i for i in quad_box[0::2]] + y_coords = [i for i in quad_box[1::2]] + + # apply the Shoelace formula + area = 0.5 * abs(sum(x_coords[i] * y_coords[i + 1] - x_coords[i + 1] * y_coords[i] for i in range(4 - 1))) + + if area < (image_width * image_height) * area_threshold: + continue + + bboxes.append(quad_box) + labels.append(ocr_content) + instances.append({ + 'quad_box': quad_box, + 'text': ocr_content, + }) + return instances + + def parse_phrase_grounding_from_text_and_spans(self, text, pattern, image_size): + # ignore and + cur_span = 0 + if text.startswith(''): + cur_span += 3 + + text = text.replace('', '') + text = text.replace('', '') + text = text.replace('', '') + + pattern = r"([^<]+(?:){4,})" + phrases = re.findall(pattern, text) + + # pattern should be text pattern and od pattern + pattern = r'^\s*(.*?)(?=||||||' + + instances = [] + for pharse_text in phrases: + phrase_text_strip = pharse_text.replace('', '', 1) + phrase_text_strip = pharse_text.replace('', '', 1) + + if phrase_text_strip == '': + cur_span += len(pharse_text) + continue + + # Prepare instance. + instance = {} + + # parse phrase, get string + phrase = re.search(pattern, phrase_text_strip) + if phrase is None: + cur_span += len(pharse_text) + continue + + # parse bboxes by box_pattern + bboxes_parsed = list(re.finditer(box_pattern, pharse_text)) + if len(bboxes_parsed) == 0: + cur_span += len(pharse_text) + continue + + phrase = phrase.group() + # remove leading and trailing spaces + phrase = phrase.strip() + + if phrase in self.black_list_of_phrase_grounding: + cur_span += len(pharse_text) + continue + + # a list of list + bbox_bins = [[int(_bboxes_parsed.group(j)) for j in range(1, 5)] for _bboxes_parsed in bboxes_parsed] + instance['bbox'] = self.box_quantizer.dequantize( + boxes=torch.tensor(bbox_bins), + size=image_size + ).tolist() + + # exclude non-ascii characters + phrase = phrase.encode('ascii',errors='ignore').decode('ascii') + instance['cat_name'] = phrase + + instances.append(instance) + + return instances + + def parse_description_with_bboxes_from_text_and_spans(self, text, pattern, image_size, allow_empty_phrase=False): + # temporary parse solution, split by '.' + # ignore and + + text = text.replace('', '') + text = text.replace('', '') + text = text.replace('', '') + + if allow_empty_phrase: + pattern = rf"(?:(?:){{4,}})" + else: + pattern = r"([^<]+(?:){4,})" + phrases = re.findall(pattern, text) + + # pattern should be text pattern and od pattern + pattern = r'^\s*(.*?)(?=||||||' + + instances = [] + for pharse_text in phrases: + phrase_text_strip = pharse_text.replace('', '', 1) + phrase_text_strip = pharse_text.replace('', '', 1) + + if phrase_text_strip == '' and not allow_empty_phrase: + continue + + # parse phrase, get string + phrase = re.search(pattern, phrase_text_strip) + if phrase is None: + continue + + phrase = phrase.group() + # remove leading and trailing spaces + phrase = phrase.strip() + + # parse bboxes by box_pattern + bboxes_parsed = list(re.finditer(box_pattern, pharse_text)) + if len(bboxes_parsed) == 0: + continue + + # a list of list + bbox_bins = [[int(_bboxes_parsed.group(j)) for j in range(1, 5)] for _bboxes_parsed in bboxes_parsed] + + bboxes = self.box_quantizer.dequantize( + boxes=torch.tensor(bbox_bins), + size=image_size + ).tolist() + + phrase = phrase.encode('ascii',errors='ignore').decode('ascii') + for _bboxes in bboxes: + # Prepare instance. + instance = {} + instance['bbox'] = _bboxes + # exclude non-ascii characters + instance['cat_name'] = phrase + instances.append(instance) + + return instances + + def parse_description_with_polygons_from_text_and_spans(self, text, pattern, image_size, + allow_empty_phrase=False, + polygon_sep_token='', + polygon_start_token='', + polygon_end_token='', + with_box_at_start=False, + ): + + # ref_seg format: '<><><><><><>' + # ignore and + + text = text.replace('', '') + text = text.replace('', '') + text = text.replace('', '') + + if allow_empty_phrase: + pattern = rf"(?:(?:|{re.escape(polygon_sep_token)}|{re.escape(polygon_start_token)}|{re.escape(polygon_end_token)}){{4,}})" + else: + # [^<]+: This part matches one or more characters that are not the < symbol. + # The ^ inside the square brackets [] is a negation, meaning it matches anything except <. + # + pattern = rf"([^<]+(?:|{re.escape(polygon_sep_token)}|{re.escape(polygon_start_token)}|{re.escape(polygon_end_token)}){{4,}})" + phrases = re.findall(pattern, text) + + phrase_string_pattern = r'^\s*(.*?)(?=||||||)' + box_pattern = rf'((?:)+)(?:{re.escape(polygon_sep_token)}|$)' + + # one polygons instance is separated by polygon_start_token and polygon_end_token + polygons_instance_pattern = rf'{re.escape(polygon_start_token)}(.*?){re.escape(polygon_end_token)}' + + instances = [] + for phrase_text in phrases: + + # exclude loc_\d+> + # need to get span if want to include category score + phrase_text_strip = re.sub(r'^loc_\d+>', '', phrase_text, count=1) + + # phrase = phrase.replace('', '') + # phrase = phrase.replace('poly>', '') + + if phrase_text_strip == '' and not allow_empty_phrase: + continue + + + # parse phrase, get string + phrase = re.search(phrase_string_pattern, phrase_text_strip) + if phrase is None: + continue + phrase = phrase.group() + # remove leading and trailing spaces + phrase = phrase.strip() + + # parse bboxes by box_pattern + + # split by polygon_start_token and polygon_end_token first using polygons_instance_pattern + if polygon_start_token in phrase_text and polygon_end_token in phrase_text: + polygons_instances_parsed = list(re.finditer(polygons_instance_pattern, phrase_text)) + else: + polygons_instances_parsed = [phrase_text] + + for _polygons_instances_parsed in polygons_instances_parsed: + # Prepare instance. + instance = {} + + # polygons_parsed= list(re.finditer(box_pattern, phrase_text)) + if isinstance(_polygons_instances_parsed, str): + polygons_parsed= list(re.finditer(box_pattern, _polygons_instances_parsed)) + else: + polygons_parsed= list(re.finditer(box_pattern, _polygons_instances_parsed.group(1))) + if len(polygons_parsed) == 0: + continue + + # a list of list (polygon) + bbox = [] + polygons = [] + for _polygon_parsed in polygons_parsed: + # group 1: whole ... + _polygon = _polygon_parsed.group(1) + # parse into list of int + _polygon = [int(_loc_parsed.group(1)) for _loc_parsed in re.finditer(r'', _polygon)] + if with_box_at_start and len(bbox) == 0: + if len(_polygon) > 4: + # no valid bbox prediction + bbox = _polygon[:4] + _polygon = _polygon[4:] + else: + bbox = [0, 0, 0, 0] + # abandon last element if is not paired + if len(_polygon) % 2 == 1: + _polygon = _polygon[:-1] + + # reshape into (n, 2) + _polygon = self.coordinates_quantizer.dequantize( + torch.tensor(np.array(_polygon).reshape(-1, 2)), + size=image_size + ).reshape(-1).tolist() + # reshape back + polygons.append(_polygon) + + instance['cat_name'] = phrase + instance['polygons'] = polygons + if len(bbox) != 0: + instance['bbox'] = self.box_quantizer.dequantize( + boxes=torch.tensor([bbox]), + size=image_size + ).tolist()[0] + + instances.append(instance) + + return instances + + def __call__( + self, + text=None, + image_size=None, + parse_tasks=None, + ): + """ + Args: + text: model outputs + image_size: (width, height) + parse_tasks: a list of tasks to parse, if None, parse all tasks. + + """ + if parse_tasks is not None: + if isinstance(parse_tasks, str): + parse_tasks = [parse_tasks] + for _parse_task in parse_tasks: + assert _parse_task in self.parse_tasks, f'parse task {_parse_task} not supported' + + # sequence or text should be provided + assert text is not None, 'text should be provided' + + parsed_dict = { + 'text': text + } + + for task in self.parse_tasks: + if parse_tasks is not None and task not in parse_tasks: + continue + + pattern = self.parse_tasks_configs[task].get('PATTERN', None) + + if task == 'ocr': + instances = self.parse_ocr_from_text_and_spans( + text, + pattern=pattern, + image_size=image_size, + area_threshold=self.parse_tasks_configs[task].get('AREA_THRESHOLD', 0.01), + ) + parsed_dict['ocr'] = instances + elif task == 'phrase_grounding': + instances = self.parse_phrase_grounding_from_text_and_spans( + text, + pattern=pattern, + image_size=image_size, + ) + parsed_dict['phrase_grounding'] = instances + elif task == 'pure_text': + parsed_dict['pure_text'] = text + elif task == 'description_with_bboxes': + instances = self.parse_description_with_bboxes_from_text_and_spans( + text, + pattern=pattern, + image_size=image_size, + ) + parsed_dict['description_with_bboxes'] = instances + elif task == 'description_with_polygons': + instances = self.parse_description_with_polygons_from_text_and_spans( + text, + pattern=pattern, + image_size=image_size, + ) + parsed_dict['description_with_polygons'] = instances + elif task == 'polygons': + instances = self.parse_description_with_polygons_from_text_and_spans( + text, + pattern=pattern, + image_size=image_size, + allow_empty_phrase=True, + ) + parsed_dict['polygons'] = instances + elif task == 'bboxes': + instances = self.parse_description_with_bboxes_from_text_and_spans( + text, + pattern=pattern, + image_size=image_size, + allow_empty_phrase=True, + ) + parsed_dict['bboxes'] = instances + elif task == 'description_with_bboxes_or_polygons': + if '' in text: + # only support either polygons or bboxes, not both at the same time + instances = self.parse_description_with_polygons_from_text_and_spans( + text, + pattern=pattern, + image_size=image_size, + ) + else: + instances = self.parse_description_with_bboxes_from_text_and_spans( + text, + pattern=pattern, + image_size=image_size, + ) + parsed_dict['description_with_bboxes_or_polygons'] = instances + else: + raise ValueError("task {} is not supported".format(task)) + + return parsed_dict \ No newline at end of file From b4cb3d6e2425dd570c2458af698d25f976f957e8 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 03:01:01 +0200 Subject: [PATCH 03/35] fix florence2 processing, add image_utils.ChannelDimension --- src/transformers/models/florence/processing_florence2.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py index 90797a3a242d..7e985e17a8b7 100644 --- a/src/transformers/models/florence/processing_florence2.py +++ b/src/transformers/models/florence/processing_florence2.py @@ -24,7 +24,7 @@ import torch from ...feature_extraction_utils import BatchFeature -from ...image_utils import ImageInput, is_valid_image +from ...image_utils import ImageInput, is_valid_image, ChannelDimension from ...processing_utils import ProcessorMixin from ...tokenization_utils_base import ( PaddingStrategy, @@ -166,9 +166,9 @@ def __call__( do_normalize: bool = None, image_mean: Optional[Union[float, List[float]]] = None, image_std: Optional[Union[float, List[float]]] = None, - data_format: Optional["ChannelDimension"] = "channels_first", # noqa: F821 + data_format: Optional[ChannelDimension] = "channels_first", # noqa: F821 input_data_format: Optional[ - Union[str, "ChannelDimension"] # noqa: F821 + Union[str, ChannelDimension] # noqa: F821 ] = None, resample: "PILImageResampling" = None, # noqa: F821 do_convert_rgb: bool = None, From 73041bacb3a04297347e7c51674508ac551b18c0 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 03:03:38 +0200 Subject: [PATCH 04/35] fix florence2 processing, add missing tokenizers --- src/transformers/models/florence/processing_florence2.py | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py index 7e985e17a8b7..b97945c47672 100644 --- a/src/transformers/models/florence/processing_florence2.py +++ b/src/transformers/models/florence/processing_florence2.py @@ -23,6 +23,11 @@ import torch +from ...models.bart.tokenization_bart import BartTokenizer +from ...models.bart.tokenization_bart_fast import BartTokenizerFast +from ...models.t5.tokenization_t5 import T5Tokenizer +from ...models.t5.tokenization_t5_fast import T5TokenizerFast + from ...feature_extraction_utils import BatchFeature from ...image_utils import ImageInput, is_valid_image, ChannelDimension from ...processing_utils import ProcessorMixin From 1ce7813f61edda78316239b8e9b42ee91cd06a97 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 03:13:33 +0200 Subject: [PATCH 05/35] add florence2 modeling --- .../models/florence/modeling_florence2.py | 2928 +++++++++++++++++ 1 file changed, 2928 insertions(+) create mode 100644 src/transformers/models/florence/modeling_florence2.py diff --git a/src/transformers/models/florence/modeling_florence2.py b/src/transformers/models/florence/modeling_florence2.py new file mode 100644 index 000000000000..41a71c8b5624 --- /dev/null +++ b/src/transformers/models/florence/modeling_florence2.py @@ -0,0 +1,2928 @@ +# coding=utf-8 +# Copyright 2024 Microsoft and the HuggingFace Inc. 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. + +""" PyTorch Florence-2 model.""" +from dataclasses import dataclass +from typing import List, Optional, Tuple, Union + +import math +import warnings +import torch +import torch.utils.checkpoint +from torch import nn +import torch.nn.functional as F +import torch.utils.checkpoint as checkpoint +from torch.nn import CrossEntropyLoss +from collections import OrderedDict +# from einops import rearrange + +from ...modeling_utils import PreTrainedModel +from ...utils import ( + ModelOutput, + add_start_docstrings, + add_start_docstrings_to_model_forward, + is_flash_attn_2_available, + logging, + replace_return_docstrings, + is_flash_attn_2_available, + is_flash_attn_greater_or_equal_2_10, +) +from .configuration_florence2 import ( + Florence2Config, + Florence2LanguageConfig, + Florence2VisionConfig, +) + + +from ...activations import ACT2FN +from ...modeling_attn_mask_utils import ( + _prepare_4d_attention_mask, + _prepare_4d_attention_mask_for_sdpa, + _prepare_4d_causal_attention_mask, + _prepare_4d_causal_attention_mask_for_sdpa, +) +from ...modeling_outputs import ( + BaseModelOutput, + BaseModelOutputWithPastAndCrossAttentions, + Seq2SeqLMOutput, + Seq2SeqModelOutput, +) + + +if is_flash_attn_2_available(): + from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input # noqa + +logger = logging.get_logger(__name__) + +_CONFIG_FOR_DOC = "Florence2Config" + +class LearnedAbsolutePositionEmbedding2D(nn.Module): + """ + This module learns positional embeddings up to a fixed maximum size. + """ + + def __init__(self, embedding_dim=256, num_pos=50): + super().__init__() + self.row_embeddings = nn.Embedding(num_pos, embedding_dim // 2) + self.column_embeddings = nn.Embedding(num_pos, embedding_dim - (embedding_dim // 2)) + + def forward(self, pixel_values): + """ + pixel_values: (batch_size, height, width, num_channels) + returns: (batch_size, height, width, embedding_dim * 2) + """ + if len(pixel_values.shape) != 4: + raise ValueError('pixel_values must be a 4D tensor') + height, width = pixel_values.shape[1:3] + width_values = torch.arange(width, device=pixel_values.device) + height_values = torch.arange(height, device=pixel_values.device) + x_emb = self.column_embeddings(width_values) + y_emb = self.row_embeddings(height_values) + # (height, width, embedding_dim * 2) + pos = torch.cat([x_emb.unsqueeze(0).repeat(height, 1, 1), y_emb.unsqueeze(1).repeat(1, width, 1)], dim=-1) + # (embedding_dim * 2, height, width) + pos = pos.permute(2, 0, 1) + pos = pos.unsqueeze(0) + # (batch_size, embedding_dim * 2, height, width) + pos = pos.repeat(pixel_values.shape[0], 1, 1, 1) + # (batch_size, height, width, embedding_dim * 2) + pos = pos.permute(0, 2, 3, 1) + return pos + +class PositionalEmbeddingCosine1D(nn.Module): + """ + This class implements a very simple positional encoding. It follows closely + the encoder from the link below: + https://pytorch.org/tutorials/beginner/translation_transformer.html + + Args: + embed_dim: The dimension of the embeddings. + dropout_prob: The dropout probability. + max_seq_len: The maximum length to precompute the positional encodings. + """ + def __init__( + self, + embed_dim: int = 512, + max_seq_len: int = 1024) -> None: + super(PositionalEmbeddingCosine1D, self).__init__() + self.embed_dim = embed_dim + self.max_seq_len = max_seq_len + # Generate the sinusoidal arrays. + factor = math.log(10000) + denominator = torch.exp( + -factor * torch.arange(0, self.embed_dim, 2) / self.embed_dim) + # Matrix where rows correspond to a positional embedding as a function + # of the position index (i.e., the row index). + frequencies = \ + torch.arange(0, self.max_seq_len) \ + .reshape(self.max_seq_len, 1) * denominator + pos_idx_to_embed = torch.zeros((self.max_seq_len, self.embed_dim)) + # Populate uneven entries. + pos_idx_to_embed[:, 0::2] = torch.sin(frequencies) + pos_idx_to_embed[:, 1::2] = torch.cos(frequencies) + # Save the positional embeddings in a constant buffer. + self.register_buffer("pos_idx_to_embed", pos_idx_to_embed) + + def forward(self, seq_embeds: torch.Tensor) -> torch.Tensor: + """ + Args: + seq_embeds: The sequence embeddings in order. Allowed size: + 1. [T, D], where T is the length of the sequence, and D is the + frame embedding dimension. + 2. [B, T, D], where B is the batch size and T and D are the + same as above. + + Returns a tensor of with the same dimensions as the input: i.e., + [1, T, D] or [T, D]. + """ + shape_len = len(seq_embeds.shape) + assert 2 <= shape_len <= 3 + len_seq = seq_embeds.size(-2) + assert len_seq <= self.max_seq_len + pos_embeds = self.pos_idx_to_embed[0:seq_embeds.size(-2), :] + # Adapt pre-computed positional embeddings to the input. + if shape_len == 3: + pos_embeds = pos_embeds.view( + (1, pos_embeds.size(0), pos_embeds.size(1))) + return pos_embeds + + +class LearnedAbsolutePositionEmbedding1D(nn.Module): + """ + Learnable absolute positional embeddings for 1D sequences. + + Args: + embed_dim: The dimension of the embeddings. + max_seq_len: The maximum length to precompute the positional encodings. + """ + def __init__( + self, + embedding_dim: int = 512, + num_pos: int = 1024) -> None: + super(LearnedAbsolutePositionEmbedding1D, self).__init__() + self.embeddings = nn.Embedding(num_pos, embedding_dim) + self.num_pos = num_pos + + def forward(self, seq_embeds: torch.Tensor) -> torch.Tensor: + """ + Args: + seq_embeds: The sequence embeddings in order. Allowed size: + 1. [T, D], where T is the length of the sequence, and D is the + frame embedding dimension. + 2. [B, T, D], where B is the batch size and T and D are the + same as above. + + Returns a tensor of with the same dimensions as the input: i.e., + [1, T, D] or [T, D]. + """ + shape_len = len(seq_embeds.shape) + assert 2 <= shape_len <= 3 + len_seq = seq_embeds.size(-2) + assert len_seq <= self.num_pos + # [T, D] + pos_embeds = self.embeddings(torch.arange(len_seq).to(seq_embeds.device)) + # Adapt pre-computed positional embeddings to the input. + if shape_len == 3: + pos_embeds = pos_embeds.view( + (1, pos_embeds.size(0), pos_embeds.size(1))) + return pos_embeds + + + +class MySequential(nn.Sequential): + def forward(self, *inputs): + for module in self._modules.values(): + if type(inputs) == tuple: + inputs = module(*inputs) + else: + inputs = module(inputs) + return inputs + + +class PreNorm(nn.Module): + def __init__(self, norm, fn, drop_path=None): + super().__init__() + self.norm = norm + self.fn = fn + self.drop_path = drop_path + + def forward(self, x, *args, **kwargs): + shortcut = x + if self.norm != None: + x, size = self.fn(self.norm(x), *args, **kwargs) + else: + x, size = self.fn(x, *args, **kwargs) + + if self.drop_path: + x = self.drop_path(x) + + x = shortcut + x + + return x, size + + +class Mlp(nn.Module): + def __init__( + self, + in_features, + hidden_features=None, + out_features=None, + act_layer=nn.GELU, + ): + super().__init__() + out_features = out_features or in_features + hidden_features = hidden_features or in_features + self.net = nn.Sequential(OrderedDict([ + ("fc1", nn.Linear(in_features, hidden_features)), + ("act", act_layer()), + ("fc2", nn.Linear(hidden_features, out_features)) + ])) + + def forward(self, x, size): + return self.net(x), size + + +class DepthWiseConv2d(nn.Module): + def __init__( + self, + dim_in, + kernel_size, + padding, + stride, + bias=True, + ): + super().__init__() + self.dw = nn.Conv2d( + dim_in, dim_in, + kernel_size=kernel_size, + padding=padding, + groups=dim_in, + stride=stride, + bias=bias + ) + + def forward(self, x, size): + B, N, C = x.shape + H, W = size + assert N == H * W + + x = self.dw(x.transpose(1, 2).view(B, C, H, W)) + size = (x.size(-2), x.size(-1)) + x = x.flatten(2).transpose(1, 2) + return x, size + + +class ConvEmbed(nn.Module): + """ Image to Patch Embedding + """ + + def __init__( + self, + patch_size=7, + in_chans=3, + embed_dim=64, + stride=4, + padding=2, + norm_layer=None, + pre_norm=True + ): + super().__init__() + self.patch_size = patch_size + + self.proj = nn.Conv2d( + in_chans, embed_dim, + kernel_size=patch_size, + stride=stride, + padding=padding + ) + + dim_norm = in_chans if pre_norm else embed_dim + self.norm = norm_layer(dim_norm) if norm_layer else None + + self.pre_norm = pre_norm + + def forward(self, x, size): + H, W = size + if len(x.size()) == 3: + if self.norm and self.pre_norm: + x = self.norm(x) + # x = rearrange( + # x, 'b (h w) c -> b c h w', + # h=H, w=W + # ) + x = x.view(-1, H, W, x.size(-1)).permute(0, 3, 1, 2) + + x = self.proj(x) + + _, _, H, W = x.shape + # TODO: check if this is correct + #x = rearrange(x, 'b c h w -> b (h w) c') + x = x.permute(0, 2, 3, 1).contiguous() + B, H, W, C = x.size() + x = x.view(B, -1, C) + + if self.norm and not self.pre_norm: + x = self.norm(x) + + return x, (H, W) + + +class ChannelAttention(nn.Module): + + def __init__(self, dim, groups=8, qkv_bias=True): + super().__init__() + + self.groups = groups + self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias) + self.proj = nn.Linear(dim, dim) + + def forward(self, x, size): + B, N, C = x.shape + + qkv = self.qkv(x).reshape(B, N, 3, self.groups, C // self.groups).permute(2, 0, 3, 1, 4) + q, k, v = qkv[0], qkv[1], qkv[2] + + q = q * (float(N) ** -0.5) + attention = q.transpose(-1, -2) @ k + attention = attention.softmax(dim=-1) + x = (attention @ v.transpose(-1, -2)).transpose(-1, -2) + x = x.transpose(1, 2).reshape(B, N, C) + x = self.proj(x) + return x, size + + +#https://github.com/huggingface/pytorch-image-models/blob/main/timm/layers/drop.py +class DropPath(nn.Module): + """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). + """ + def __init__(self, drop_prob: float = 0., scale_by_keep: bool = True): + super(DropPath, self).__init__() + self.drop_prob = drop_prob + self.scale_by_keep = scale_by_keep + + def forward(self, x): + if self.drop_prob == 0. or not self.training: + return x + keep_prob = 1 - self.drop_prob + shape = (x.shape[0],) + (1,) * (x.ndim - 1) # work with diff dim tensors, not just 2D ConvNets + random_tensor = x.new_empty(shape).bernoulli_(keep_prob) + if keep_prob > 0.0 and self.scale_by_keep: + random_tensor.div_(keep_prob) + return x * random_tensor + + def extra_repr(self): + return f'drop_prob={round(self.drop_prob,3):0.3f}' + + +# TODO: can this be replaced with torch.nn.init.trunc_normal_? +#https://github.com/huggingface/pytorch-image-models/blob/b28945ff056d454b174f0fb8682e362b87150141/timm/layers/weight_init.py +def trunc_normal_(tensor: torch.Tensor, mean=0., std=1., a=-2., b=2.) -> torch.Tensor: + r"""Fills the input Tensor with values drawn from a truncated + normal distribution. The values are effectively drawn from the + normal distribution :math:`\mathcal{N}(\text{mean}, \text{std}^2)` + with values outside :math:`[a, b]` redrawn until they are within + the bounds. The method used for generating the random values works + best when :math:`a \leq \text{mean} \leq b`. + + NOTE: this impl is similar to the PyTorch trunc_normal_, the bounds [a, b] are + applied while sampling the normal with mean/std applied, therefore a, b args + should be adjusted to match the range of mean, std args. + + Args: + tensor: an n-dimensional `torch.Tensor` + mean: the mean of the normal distribution + std: the standard deviation of the normal distribution + a: the minimum cutoff value + b: the maximum cutoff value + """ + # Method based on https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf + def norm_cdf(x): + # Computes standard normal cumulative distribution function + return (1. + math.erf(x / math.sqrt(2.))) / 2. + + with torch.no_grad(): + # Cut & paste from PyTorch official master until it's in a few official releases - RW + if (mean < a - 2 * std) or (mean > b + 2 * std): + warnings.warn("mean is more than 2 std from [a, b] in nn.init.trunc_normal_. " + "The distribution of values may be incorrect.", + stacklevel=2) + + # Values are generated by using a truncated uniform distribution and + # then using the inverse CDF for the normal distribution. + # Get upper and lower cdf values + l = norm_cdf((a - mean) / std) + u = norm_cdf((b - mean) / std) + + # Uniformly fill tensor with values from [l, u], then translate to + # [2l-1, 2u-1]. + tensor.uniform_(2 * l - 1, 2 * u - 1) + + # Use inverse cdf transform for normal distribution to get truncated + # standard normal + tensor.erfinv_() + + # Transform to proper mean, std + tensor.mul_(std * math.sqrt(2.)) + tensor.add_(mean) + + # Clamp to ensure it's in the proper range + tensor.clamp_(min=a, max=b) + return tensor + + +class ChannelBlock(nn.Module): + + def __init__(self, dim, groups, mlp_ratio=4., qkv_bias=True, + drop_path_rate=0., act_layer=nn.GELU, norm_layer=nn.LayerNorm, + conv_at_attn=True, conv_at_ffn=True): + super().__init__() + + drop_path = DropPath(drop_path_rate) if drop_path_rate > 0. else nn.Identity() + + self.conv1 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None + self.channel_attn = PreNorm( + norm_layer(dim), + ChannelAttention(dim, groups=groups, qkv_bias=qkv_bias), + drop_path + ) + self.conv2 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_ffn else None + self.ffn = PreNorm( + norm_layer(dim), + Mlp(in_features=dim, hidden_features=int(dim*mlp_ratio), act_layer=act_layer), + drop_path + ) + + def forward(self, x, size): + if self.conv1: + x, size = self.conv1(x, size) + x, size = self.channel_attn(x, size) + + if self.conv2: + x, size = self.conv2(x, size) + x, size = self.ffn(x, size) + + return x, size + + +def window_partition(x, window_size: int): + B, H, W, C = x.shape + x = x.view(B, H // window_size, window_size, W // window_size, window_size, C) + windows = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(-1, window_size, window_size, C) + return windows + + +def window_reverse(windows, batch_size: int, window_size: int, H: int, W: int): + B = batch_size + # this will cause onnx conversion failed for dynamic axis, because treated as constant + # int(windows.shape[0] / (H * W / window_size / window_size)) + x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1) + x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1) + return x + + +class WindowAttention(nn.Module): + def __init__(self, dim, num_heads, window_size, qkv_bias=True): + + super().__init__() + self.dim = dim + self.window_size = window_size + self.num_heads = num_heads + head_dim = dim // num_heads + self.scale = float(head_dim) ** -0.5 + + self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias) + self.proj = nn.Linear(dim, dim) + + self.softmax = nn.Softmax(dim=-1) + + def forward(self, x, size): + + H, W = size + B, L, C = x.shape + assert L == H * W, "input feature has wrong size" + + x = x.view(B, H, W, C) + + pad_l = pad_t = 0 + pad_r = (self.window_size - W % self.window_size) % self.window_size + pad_b = (self.window_size - H % self.window_size) % self.window_size + x = F.pad(x, (0, 0, pad_l, pad_r, pad_t, pad_b)) + _, Hp, Wp, _ = x.shape + + x = window_partition(x, self.window_size) + x = x.view(-1, self.window_size * self.window_size, C) + + # W-MSA/SW-MSA + # attn_windows = self.attn(x_windows) + + B_, N, C = x.shape + qkv = self.qkv(x).reshape(B_, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4) + q, k, v = qkv[0], qkv[1], qkv[2] + + q = q * self.scale + attn = (q @ k.transpose(-2, -1)) + attn = self.softmax(attn) + + x = (attn @ v).transpose(1, 2).reshape(B_, N, C) + x = self.proj(x) + + # merge windows + x = x.view( + -1, self.window_size, self.window_size, C + ) + x = window_reverse(x, B, self.window_size, Hp, Wp) + + if pad_r > 0 or pad_b > 0: + x = x[:, :H, :W, :].contiguous() + + x = x.view(B, H * W, C) + + return x, size + + +class SpatialBlock(nn.Module): + + def __init__(self, dim, num_heads, window_size, + mlp_ratio=4., qkv_bias=True, drop_path_rate=0., act_layer=nn.GELU, + norm_layer=nn.LayerNorm, conv_at_attn=True, conv_at_ffn=True): + super().__init__() + + drop_path = DropPath(drop_path_rate) if drop_path_rate > 0. else nn.Identity() + + self.conv1 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None + self.window_attn = PreNorm( + norm_layer(dim), + WindowAttention(dim, num_heads, window_size, qkv_bias=qkv_bias), + drop_path + ) + self.conv2 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_ffn else None + self.ffn = PreNorm( + norm_layer(dim), + Mlp(in_features=dim, hidden_features=int(dim*mlp_ratio), act_layer=act_layer), + drop_path + ) + + def forward(self, x, size): + if self.conv1: + x, size = self.conv1(x, size) + x, size = self.window_attn(x, size) + + if self.conv2: + x, size = self.conv2(x, size) + x, size = self.ffn(x, size) + return x, size + + +class DaViT(nn.Module): + """ DaViT: Dual-Attention Transformer + + Args: + in_chans (int): Number of input image channels. Default: 3. + num_classes (int): Number of classes for classification head. Default: 1000. + patch_size (tuple(int)): Patch size of convolution in different stages. Default: (7, 2, 2, 2). + patch_stride (tuple(int)): Patch stride of convolution in different stages. Default: (4, 2, 2, 2). + patch_padding (tuple(int)): Patch padding of convolution in different stages. Default: (3, 0, 0, 0). + patch_prenorm (tuple(bool)): If True, perform norm before convlution layer. Default: (True, False, False, False). + embed_dims (tuple(int)): Patch embedding dimension in different stages. Default: (64, 128, 192, 256). + num_heads (tuple(int)): Number of spatial attention heads in different stages. Default: (4, 8, 12, 16). + num_groups (tuple(int)): Number of channel groups in different stages. Default: (4, 8, 12, 16). + window_size (int): Window size. Default: 7. + mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4. + qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: True. + drop_path_rate (float): Stochastic depth rate. Default: 0.1. + norm_layer (nn.Module): Normalization layer. Default: nn.LayerNorm. + enable_checkpoint (bool): If True, enable checkpointing. Default: False. + conv_at_attn (bool): If True, performe depthwise convolution before attention layer. Default: True. + conv_at_ffn (bool): If True, performe depthwise convolution before ffn layer. Default: True. + """ + + def __init__( + self, + in_chans=3, + num_classes=1000, + depths=(1, 1, 3, 1), + patch_size=(7, 2, 2, 2), + patch_stride=(4, 2, 2, 2), + patch_padding=(3, 0, 0, 0), + patch_prenorm=(False, False, False, False), + embed_dims=(64, 128, 192, 256), + num_heads=(3, 6, 12, 24), + num_groups=(3, 6, 12, 24), + window_size=7, + mlp_ratio=4., + qkv_bias=True, + drop_path_rate=0.1, + norm_layer=nn.LayerNorm, + enable_checkpoint=False, + conv_at_attn=True, + conv_at_ffn=True, + ): + super().__init__() + + self.num_classes = num_classes + self.embed_dims = embed_dims + self.num_heads = num_heads + self.num_groups = num_groups + self.num_stages = len(self.embed_dims) + self.enable_checkpoint = enable_checkpoint + assert self.num_stages == len(self.num_heads) == len(self.num_groups) + + num_stages = len(embed_dims) + dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths)*2)] + + depth_offset = 0 + convs = [] + blocks = [] + for i in range(num_stages): + conv_embed = ConvEmbed( + patch_size=patch_size[i], + stride=patch_stride[i], + padding=patch_padding[i], + in_chans=in_chans if i == 0 else self.embed_dims[i - 1], + embed_dim=self.embed_dims[i], + norm_layer=norm_layer, + pre_norm=patch_prenorm[i] + ) + convs.append(conv_embed) + + block = MySequential( + *[ + MySequential(OrderedDict([ + ( + 'spatial_block', SpatialBlock( + embed_dims[i], + num_heads[i], + window_size, + drop_path_rate=dpr[depth_offset+j*2], + qkv_bias=qkv_bias, + mlp_ratio=mlp_ratio, + conv_at_attn=conv_at_attn, + conv_at_ffn=conv_at_ffn, + ) + ), + ( + 'channel_block', ChannelBlock( + embed_dims[i], + num_groups[i], + drop_path_rate=dpr[depth_offset+j*2+1], + qkv_bias=qkv_bias, + mlp_ratio=mlp_ratio, + conv_at_attn=conv_at_attn, + conv_at_ffn=conv_at_ffn, + ) + ) + ])) for j in range(depths[i]) + ] + ) + blocks.append(block) + depth_offset += depths[i]*2 + + self.convs = nn.ModuleList(convs) + self.blocks = nn.ModuleList(blocks) + + self.norms = norm_layer(self.embed_dims[-1]) + self.avgpool = nn.AdaptiveAvgPool1d(1) + self.head = nn.Linear(self.embed_dims[-1], num_classes) if num_classes > 0 else nn.Identity() + + self.apply(self._init_weights) + + @property + def dim_out(self): + return self.embed_dims[-1] + + def _init_weights(self, m): + if isinstance(m, nn.Linear): + trunc_normal_(m.weight, std=0.02) + if m.bias is not None: + nn.init.constant_(m.bias, 0) + elif isinstance(m, nn.Conv2d): + nn.init.normal_(m.weight, std=0.02) + for name, _ in m.named_parameters(): + if name in ['bias']: + nn.init.constant_(m.bias, 0) + elif isinstance(m, nn.LayerNorm): + nn.init.constant_(m.weight, 1.0) + nn.init.constant_(m.bias, 0) + elif isinstance(m, nn.BatchNorm2d): + nn.init.constant_(m.weight, 1.0) + nn.init.constant_(m.bias, 0) + + def forward_features_unpool(self, x): + """ + forward until avg pooling + Args: + x (_type_): input image tensor + """ + input_size = (x.size(2), x.size(3)) + for conv, block in zip(self.convs, self.blocks): + x, input_size = conv(x, input_size) + if self.enable_checkpoint: + x, input_size = checkpoint.checkpoint(block, x, input_size) + else: + x, input_size = block(x, input_size) + return x + + def forward_features(self, x): + x = self.forward_features_unpool(x) + + # (batch_size, num_tokens, token_dim) + x = self.avgpool(x.transpose(1, 2)) + # (batch_size, 1, num_tokens) + x = torch.flatten(x, 1) + x = self.norms(x) + + return x + + def forward(self, x): + x = self.forward_features(x) + x = self.head(x) + return x + + @classmethod + def from_config(cls, config): + return cls( + depths=config.depths, + embed_dims=config.dim_embed, + num_heads=config.num_heads, + num_groups=config.num_groups, + patch_size=config.patch_size, + patch_stride=config.patch_stride, + patch_padding=config.patch_padding, + patch_prenorm=config.patch_prenorm, + drop_path_rate=config.drop_path_rate, + window_size=config.window_size, + ) + + + + +if is_flash_attn_2_available(): + from flash_attn import flash_attn_func, flash_attn_varlen_func + from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input # noqa + +# Copied from transformers.models.llama.modeling_llama._get_unpad_data +def _get_unpad_data(attention_mask): + seqlens_in_batch = attention_mask.sum(dim=-1, dtype=torch.int32) + indices = torch.nonzero(attention_mask.flatten(), as_tuple=False).flatten() + max_seqlen_in_batch = seqlens_in_batch.max().item() + cu_seqlens = F.pad(torch.cumsum(seqlens_in_batch, dim=0, dtype=torch.int32), (1, 0)) + return ( + indices, + cu_seqlens, + max_seqlen_in_batch, + ) + + +def shift_tokens_right(input_ids: torch.Tensor, pad_token_id: int, decoder_start_token_id: int): + """ + Shift input ids one token to the right. + """ + shifted_input_ids = input_ids.new_zeros(input_ids.shape) + shifted_input_ids[:, 1:] = input_ids[:, :-1].clone() + shifted_input_ids[:, 0] = decoder_start_token_id + + if pad_token_id is None: + raise ValueError("self.model.config.pad_token_id has to be defined.") + # replace possible -100 values in labels by `pad_token_id` + shifted_input_ids.masked_fill_(shifted_input_ids == -100, pad_token_id) + + return shifted_input_ids + + +class Florence2LearnedPositionalEmbedding(nn.Embedding): + """ + This module learns positional embeddings up to a fixed maximum size. + """ + + def __init__(self, num_embeddings: int, embedding_dim: int): + # Florence2 is set up so that if padding_idx is specified then offset the embedding ids by 2 + # and adjust num_embeddings appropriately. Other models don't have this hack + self.offset = 2 + super().__init__(num_embeddings + self.offset, embedding_dim) + + def forward(self, input_ids: torch.Tensor, past_key_values_length: int = 0): + """`input_ids' shape is expected to be [bsz x seqlen].""" + + bsz, seq_len = input_ids.shape[:2] + positions = torch.arange( + past_key_values_length, past_key_values_length + seq_len, dtype=torch.long, device=self.weight.device + ).expand(bsz, -1) + + return super().forward(positions + self.offset) + + +class Florence2ScaledWordEmbedding(nn.Embedding): + """ + This module overrides nn.Embeddings' forward by multiplying with embeddings scale. + """ + + def __init__(self, num_embeddings: int, embedding_dim: int, padding_idx: int, embed_scale: Optional[float] = 1.0): + super().__init__(num_embeddings, embedding_dim, padding_idx) + self.embed_scale = embed_scale + + def forward(self, input_ids: torch.Tensor): + return super().forward(input_ids) * self.embed_scale + + +class Florence2Attention(nn.Module): + """Multi-headed attention from 'Attention Is All You Need' paper""" + + def __init__( + self, + embed_dim: int, + num_heads: int, + dropout: float = 0.0, + is_decoder: bool = False, + bias: bool = True, + is_causal: bool = False, + config: Optional[Florence2LanguageConfig] = None, + ): + super().__init__() + self.embed_dim = embed_dim + self.num_heads = num_heads + self.dropout = dropout + self.head_dim = embed_dim // num_heads + self.config = config + + if (self.head_dim * num_heads) != self.embed_dim: + raise ValueError( + f"embed_dim must be divisible by num_heads (got `embed_dim`: {self.embed_dim}" + f" and `num_heads`: {num_heads})." + ) + self.scaling = self.head_dim**-0.5 + self.is_decoder = is_decoder + self.is_causal = is_causal + + self.k_proj = nn.Linear(embed_dim, embed_dim, bias=bias) + self.v_proj = nn.Linear(embed_dim, embed_dim, bias=bias) + self.q_proj = nn.Linear(embed_dim, embed_dim, bias=bias) + self.out_proj = nn.Linear(embed_dim, embed_dim, bias=bias) + + def _shape(self, tensor: torch.Tensor, seq_len: int, bsz: int): + return tensor.view(bsz, seq_len, self.num_heads, self.head_dim).transpose(1, 2).contiguous() + + def forward( + self, + hidden_states: torch.Tensor, + key_value_states: Optional[torch.Tensor] = None, + past_key_value: Optional[Tuple[torch.Tensor]] = None, + attention_mask: Optional[torch.Tensor] = None, + layer_head_mask: Optional[torch.Tensor] = None, + output_attentions: bool = False, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + """Input shape: Batch x Time x Channel""" + + # if key_value_states are provided this layer is used as a cross-attention layer + # for the decoder + is_cross_attention = key_value_states is not None + + bsz, tgt_len, _ = hidden_states.size() + + # get query proj + query_states = self.q_proj(hidden_states) * self.scaling + # get key, value proj + # `past_key_value[0].shape[2] == key_value_states.shape[1]` + # is checking that the `sequence_length` of the `past_key_value` is the same as + # the provided `key_value_states` to support prefix tuning + if ( + is_cross_attention + and past_key_value is not None + and past_key_value[0].shape[2] == key_value_states.shape[1] + ): + # reuse k,v, cross_attentions + key_states = past_key_value[0] + value_states = past_key_value[1] + elif is_cross_attention: + # cross_attentions + key_states = self._shape(self.k_proj(key_value_states), -1, bsz) + value_states = self._shape(self.v_proj(key_value_states), -1, bsz) + elif past_key_value is not None: + # reuse k, v, self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + key_states = torch.cat([past_key_value[0], key_states], dim=2) + value_states = torch.cat([past_key_value[1], value_states], dim=2) + else: + # self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + + if self.is_decoder: + # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states. + # Further calls to cross_attention layer can then reuse all cross-attention + # key/value_states (first "if" case) + # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of + # all previous decoder key/value_states. Further calls to uni-directional self-attention + # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case) + # if encoder bi-directional self-attention `past_key_value` is always `None` + past_key_value = (key_states, value_states) + + proj_shape = (bsz * self.num_heads, -1, self.head_dim) + query_states = self._shape(query_states, tgt_len, bsz).view(*proj_shape) + key_states = key_states.reshape(*proj_shape) + value_states = value_states.reshape(*proj_shape) + + src_len = key_states.size(1) + attn_weights = torch.bmm(query_states, key_states.transpose(1, 2)) + + if attn_weights.size() != (bsz * self.num_heads, tgt_len, src_len): + raise ValueError( + f"Attention weights should be of size {(bsz * self.num_heads, tgt_len, src_len)}, but is" + f" {attn_weights.size()}" + ) + + if attention_mask is not None: + if attention_mask.size() != (bsz, 1, tgt_len, src_len): + raise ValueError( + f"Attention mask should be of size {(bsz, 1, tgt_len, src_len)}, but is {attention_mask.size()}" + ) + attn_weights = attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + attention_mask + attn_weights = attn_weights.view(bsz * self.num_heads, tgt_len, src_len) + + attn_weights = nn.functional.softmax(attn_weights, dim=-1) + + if layer_head_mask is not None: + if layer_head_mask.size() != (self.num_heads,): + raise ValueError( + f"Head mask for a single layer should be of size {(self.num_heads,)}, but is" + f" {layer_head_mask.size()}" + ) + attn_weights = layer_head_mask.view(1, -1, 1, 1) * attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + attn_weights = attn_weights.view(bsz * self.num_heads, tgt_len, src_len) + + if output_attentions: + # this operation is a bit awkward, but it's required to + # make sure that attn_weights keeps its gradient. + # In order to do so, attn_weights have to be reshaped + # twice and have to be reused in the following + attn_weights_reshaped = attn_weights.view(bsz, self.num_heads, tgt_len, src_len) + attn_weights = attn_weights_reshaped.view(bsz * self.num_heads, tgt_len, src_len) + else: + attn_weights_reshaped = None + + attn_probs = nn.functional.dropout(attn_weights, p=self.dropout, training=self.training) + + attn_output = torch.bmm(attn_probs, value_states) + + if attn_output.size() != (bsz * self.num_heads, tgt_len, self.head_dim): + raise ValueError( + f"`attn_output` should be of size {(bsz * self.num_heads, tgt_len, self.head_dim)}, but is" + f" {attn_output.size()}" + ) + + attn_output = attn_output.view(bsz, self.num_heads, tgt_len, self.head_dim) + attn_output = attn_output.transpose(1, 2) + + # Use the `embed_dim` from the config (stored in the class) rather than `hidden_state` because `attn_output` can be + # partitioned across GPUs when using tensor-parallelism. + attn_output = attn_output.reshape(bsz, tgt_len, self.embed_dim) + + attn_output = self.out_proj(attn_output) + + return attn_output, attn_weights_reshaped, past_key_value + + +class Florence2FlashAttention2(Florence2Attention): + """ + Florence2 flash attention module. This module inherits from `Florence2Attention` as the weights of the module stays + untouched. The only required change would be on the forward pass where it needs to correctly call the public API of + flash attention and deal with padding tokens in case the input contains any of them. + """ + + # Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2.__init__ + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + + # TODO: Should be removed once Flash Attention for RoCm is bumped to 2.1. + # flash_attn<2.1 generates top-left aligned causal mask, while what is needed here is bottom-right alignement, that was made default for flash_attn>=2.1. This attribute is used to handle this difference. Reference: https://github.com/Dao-AILab/flash-attention/releases/tag/v2.1.0. + # Beware that with flash_attn<2.1, using q_seqlen != k_seqlen (except for the case q_seqlen == 1) produces a wrong mask (top-left). + self._flash_attn_uses_top_left_mask = not is_flash_attn_greater_or_equal_2_10() + + def _reshape(self, tensor: torch.Tensor, seq_len: int, bsz: int): + return tensor.view(bsz, seq_len, self.num_heads, self.head_dim) + + def forward( + self, + hidden_states: torch.Tensor, + key_value_states: Optional[torch.Tensor] = None, + past_key_value: Optional[Tuple[torch.Tensor]] = None, + attention_mask: Optional[torch.Tensor] = None, + layer_head_mask: Optional[torch.Tensor] = None, + output_attentions: bool = False, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + # Florence2FlashAttention2 attention does not support output_attentions + if output_attentions: + raise ValueError("Florence2FlashAttention2 attention does not support output_attentions") + + # if key_value_states are provided this layer is used as a cross-attention layer + # for the decoder + is_cross_attention = key_value_states is not None + + bsz, q_len, _ = hidden_states.size() + + # get query proj + query_states = self._reshape(self.q_proj(hidden_states), -1, bsz) + # get key, value proj + # `past_key_value[0].shape[2] == key_value_states.shape[1]` + # is checking that the `sequence_length` of the `past_key_value` is the same as + # the provided `key_value_states` to support prefix tuning + if ( + is_cross_attention + and past_key_value is not None + and past_key_value[0].shape[2] == key_value_states.shape[1] + ): + # reuse k,v, cross_attentions + key_states = past_key_value[0].transpose(1, 2) + value_states = past_key_value[1].transpose(1, 2) + elif is_cross_attention: + # cross_attentions + key_states = self._reshape(self.k_proj(key_value_states), -1, bsz) + value_states = self._reshape(self.v_proj(key_value_states), -1, bsz) + elif past_key_value is not None: + # reuse k, v, self_attention + key_states = self._reshape(self.k_proj(hidden_states), -1, bsz) + value_states = self._reshape(self.v_proj(hidden_states), -1, bsz) + key_states = torch.cat([past_key_value[0].transpose(1, 2), key_states], dim=1) + value_states = torch.cat([past_key_value[1].transpose(1, 2), value_states], dim=1) + else: + # self_attention + key_states = self._reshape(self.k_proj(hidden_states), -1, bsz) + value_states = self._reshape(self.v_proj(hidden_states), -1, bsz) + + if self.is_decoder: + # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states. + # Further calls to cross_attention layer can then reuse all cross-attention + # key/value_states (first "if" case) + # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of + # all previous decoder key/value_states. Further calls to uni-directional self-attention + # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case) + # if encoder bi-directional self-attention `past_key_value` is always `None` + past_key_value = (key_states.transpose(1, 2), value_states.transpose(1, 2)) + + kv_seq_len = key_states.shape[-2] + if past_key_value is not None: + kv_seq_len += past_key_value[0].shape[-2] + + # In PEFT, usually we cast the layer norms in float32 for training stability reasons + # therefore the input hidden states gets silently casted in float32. Hence, we need + # cast them back in the correct dtype just to be sure everything works as expected. + # This might slowdown training & inference so it is recommended to not cast the LayerNorms + # in fp32. (LlamaRMSNorm handles it correctly) + + input_dtype = query_states.dtype + if input_dtype == torch.float32: + if torch.is_autocast_enabled(): + target_dtype = torch.get_autocast_gpu_dtype() + # Handle the case where the model is quantized + elif hasattr(self.config, "_pre_quantization_dtype"): + target_dtype = self.config._pre_quantization_dtype + else: + target_dtype = self.q_proj.weight.dtype + + logger.warning_once( + f"The input hidden states seems to be silently casted in float32, this might be related to" + f" the fact you have upcasted embedding or layer norm layers in float32. We will cast back the input in" + f" {target_dtype}." + ) + + query_states = query_states.to(target_dtype) + key_states = key_states.to(target_dtype) + value_states = value_states.to(target_dtype) + + attn_output = self._flash_attention_forward( + query_states, key_states, value_states, attention_mask, q_len, dropout=self.dropout + ) + + attn_output = attn_output.reshape(bsz, q_len, -1) + attn_output = self.out_proj(attn_output) + + if not output_attentions: + attn_weights = None + + return attn_output, attn_weights, past_key_value + + # Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2._flash_attention_forward + def _flash_attention_forward( + self, query_states, key_states, value_states, attention_mask, query_length, dropout=0.0, softmax_scale=None + ): + """ + Calls the forward method of Flash Attention - if the input hidden states contain at least one padding token + first unpad the input, then computes the attention scores and pad the final attention scores. + + Args: + query_states (`torch.Tensor`): + Input query states to be passed to Flash Attention API + key_states (`torch.Tensor`): + Input key states to be passed to Flash Attention API + value_states (`torch.Tensor`): + Input value states to be passed to Flash Attention API + attention_mask (`torch.Tensor`): + The padding mask - corresponds to a tensor of size `(batch_size, seq_len)` where 0 stands for the + position of padding tokens and 1 for the position of non-padding tokens. + dropout (`float`): + Attention dropout + softmax_scale (`float`, *optional*): + The scaling of QK^T before applying softmax. Default to 1 / sqrt(head_dim) + """ + if not self._flash_attn_uses_top_left_mask: + causal = self.is_causal + else: + # TODO: Remove the `query_length != 1` check once Flash Attention for RoCm is bumped to 2.1. For details, please see the comment in LlamaFlashAttention2 __init__. + causal = self.is_causal and query_length != 1 + + # Contains at least one padding token in the sequence + if attention_mask is not None: + batch_size = query_states.shape[0] + query_states, key_states, value_states, indices_q, cu_seq_lens, max_seq_lens = self._upad_input( + query_states, key_states, value_states, attention_mask, query_length + ) + + cu_seqlens_q, cu_seqlens_k = cu_seq_lens + max_seqlen_in_batch_q, max_seqlen_in_batch_k = max_seq_lens + + attn_output_unpad = flash_attn_varlen_func( + query_states, + key_states, + value_states, + cu_seqlens_q=cu_seqlens_q, + cu_seqlens_k=cu_seqlens_k, + max_seqlen_q=max_seqlen_in_batch_q, + max_seqlen_k=max_seqlen_in_batch_k, + dropout_p=dropout, + softmax_scale=softmax_scale, + causal=causal, + ) + + attn_output = pad_input(attn_output_unpad, indices_q, batch_size, query_length) + else: + attn_output = flash_attn_func( + query_states, key_states, value_states, dropout, softmax_scale=softmax_scale, causal=causal + ) + + return attn_output + + # Copied from transformers.models.llama.modeling_llama.LlamaFlashAttention2._upad_input + def _upad_input(self, query_layer, key_layer, value_layer, attention_mask, query_length): + indices_k, cu_seqlens_k, max_seqlen_in_batch_k = _get_unpad_data(attention_mask) + batch_size, kv_seq_len, num_key_value_heads, head_dim = key_layer.shape + + key_layer = index_first_axis( + key_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k + ) + value_layer = index_first_axis( + value_layer.reshape(batch_size * kv_seq_len, num_key_value_heads, head_dim), indices_k + ) + if query_length == kv_seq_len: + query_layer = index_first_axis( + query_layer.reshape(batch_size * kv_seq_len, self.num_heads, head_dim), indices_k + ) + cu_seqlens_q = cu_seqlens_k + max_seqlen_in_batch_q = max_seqlen_in_batch_k + indices_q = indices_k + elif query_length == 1: + max_seqlen_in_batch_q = 1 + cu_seqlens_q = torch.arange( + batch_size + 1, dtype=torch.int32, device=query_layer.device + ) # There is a memcpy here, that is very bad. + indices_q = cu_seqlens_q[:-1] + query_layer = query_layer.squeeze(1) + else: + # The -q_len: slice assumes left padding. + attention_mask = attention_mask[:, -query_length:] + query_layer, indices_q, cu_seqlens_q, max_seqlen_in_batch_q = unpad_input(query_layer, attention_mask) + + return ( + query_layer, + key_layer, + value_layer, + indices_q, + (cu_seqlens_q, cu_seqlens_k), + (max_seqlen_in_batch_q, max_seqlen_in_batch_k), + ) + + +class Florence2SdpaAttention(Florence2Attention): + def forward( + self, + hidden_states: torch.Tensor, + key_value_states: Optional[torch.Tensor] = None, + past_key_value: Optional[Tuple[torch.Tensor]] = None, + attention_mask: Optional[torch.Tensor] = None, + layer_head_mask: Optional[torch.Tensor] = None, + output_attentions: bool = False, + ) -> Tuple[torch.Tensor, Optional[torch.Tensor], Optional[Tuple[torch.Tensor]]]: + """Input shape: Batch x Time x Channel""" + if output_attentions or layer_head_mask is not None: + # TODO: Improve this warning with e.g. `model.config._attn_implementation = "manual"` once this is implemented. + logger.warning_once( + "Florence2Model is using Florence2SdpaAttention, but `torch.nn.functional.scaled_dot_product_attention` does not support `output_attentions=True` or `layer_head_mask` not None. Falling back to the manual attention" + ' implementation, but specifying the manual implementation will be required from Transformers version v5.0.0 onwards. This warning can be removed using the argument `attn_implementation="eager"` when loading the model.' + ) + return super().forward( + hidden_states, + key_value_states=key_value_states, + past_key_value=past_key_value, + attention_mask=attention_mask, + layer_head_mask=layer_head_mask, + output_attentions=output_attentions, + ) + + # if key_value_states are provided this layer is used as a cross-attention layer + # for the decoder + is_cross_attention = key_value_states is not None + + bsz, tgt_len, _ = hidden_states.size() + + # get query proj + query_states = self.q_proj(hidden_states) + # get key, value proj + # `past_key_value[0].shape[2] == key_value_states.shape[1]` + # is checking that the `sequence_length` of the `past_key_value` is the same as + # the provided `key_value_states` to support prefix tuning + if ( + is_cross_attention + and past_key_value is not None + and past_key_value[0].shape[2] == key_value_states.shape[1] + ): + # reuse k,v, cross_attentions + key_states = past_key_value[0] + value_states = past_key_value[1] + elif is_cross_attention: + # cross_attentions + key_states = self._shape(self.k_proj(key_value_states), -1, bsz) + value_states = self._shape(self.v_proj(key_value_states), -1, bsz) + elif past_key_value is not None: + # reuse k, v, self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + key_states = torch.cat([past_key_value[0], key_states], dim=2) + value_states = torch.cat([past_key_value[1], value_states], dim=2) + else: + # self_attention + key_states = self._shape(self.k_proj(hidden_states), -1, bsz) + value_states = self._shape(self.v_proj(hidden_states), -1, bsz) + + if self.is_decoder: + # if cross_attention save Tuple(torch.Tensor, torch.Tensor) of all cross attention key/value_states. + # Further calls to cross_attention layer can then reuse all cross-attention + # key/value_states (first "if" case) + # if uni-directional self-attention (decoder) save Tuple(torch.Tensor, torch.Tensor) of + # all previous decoder key/value_states. Further calls to uni-directional self-attention + # can concat previous decoder key/value_states to current projected key/value_states (third "elif" case) + # if encoder bi-directional self-attention `past_key_value` is always `None` + past_key_value = (key_states, value_states) + + query_states = self._shape(query_states, tgt_len, bsz) + + # We dispatch to SDPA's Flash Attention or Efficient kernels via this `is_causal` if statement instead of an inline conditional assignment + # in SDPA to support both torch.compile's dynamic shapes and full graph options. An inline conditional prevents dynamic shapes from compiling. + # The tgt_len > 1 is necessary to match with AttentionMaskConverter.to_causal_4d that does not create a causal mask in case tgt_len == 1. + is_causal = True if self.is_causal and attention_mask is None and tgt_len > 1 else False + + # NOTE: SDPA with memory-efficient backend is currently (torch==2.1.2) bugged when using non-contiguous inputs and a custom attn_mask, + # but we are fine here as `_shape` do call `.contiguous()`. Reference: https://github.com/pytorch/pytorch/issues/112577 + attn_output = torch.nn.functional.scaled_dot_product_attention( + query_states, + key_states, + value_states, + attn_mask=attention_mask, + dropout_p=self.dropout if self.training else 0.0, + is_causal=is_causal, + ) + + if attn_output.size() != (bsz, self.num_heads, tgt_len, self.head_dim): + raise ValueError( + f"`attn_output` should be of size {(bsz, self.num_heads, tgt_len, self.head_dim)}, but is" + f" {attn_output.size()}" + ) + + attn_output = attn_output.transpose(1, 2) + + # Use the `embed_dim` from the config (stored in the class) rather than `hidden_state` because `attn_output` can be + # partitioned across GPUs when using tensor-parallelism. + attn_output = attn_output.reshape(bsz, tgt_len, self.embed_dim) + + attn_output = self.out_proj(attn_output) + + return attn_output, None, past_key_value + + +FLORENCE2_ATTENTION_CLASSES = { + "eager": Florence2Attention, + "sdpa": Florence2SdpaAttention, + "flash_attention_2": Florence2FlashAttention2, +} + + +class Florence2EncoderLayer(nn.Module): + def __init__(self, config: Florence2LanguageConfig): + super().__init__() + self.embed_dim = config.d_model + + self.self_attn = FLORENCE2_ATTENTION_CLASSES[config._attn_implementation]( + embed_dim=self.embed_dim, + num_heads=config.encoder_attention_heads, + dropout=config.attention_dropout, + config=config, + ) + self.self_attn_layer_norm = nn.LayerNorm(self.embed_dim) + self.dropout = config.dropout + self.activation_fn = ACT2FN[config.activation_function] + self.activation_dropout = config.activation_dropout + self.fc1 = nn.Linear(self.embed_dim, config.encoder_ffn_dim) + self.fc2 = nn.Linear(config.encoder_ffn_dim, self.embed_dim) + self.final_layer_norm = nn.LayerNorm(self.embed_dim) + + def forward( + self, + hidden_states: torch.FloatTensor, + attention_mask: torch.FloatTensor, + layer_head_mask: torch.FloatTensor, + output_attentions: Optional[bool] = False, + ) -> Tuple[torch.FloatTensor, Optional[torch.FloatTensor]]: + """ + Args: + hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` + attention_mask (`torch.FloatTensor`): attention mask of size + `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values. + layer_head_mask (`torch.FloatTensor`): mask for attention heads in a given layer of size + `(encoder_attention_heads,)`. + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under + returned tensors for more detail. + """ + residual = hidden_states + hidden_states, attn_weights, _ = self.self_attn( + hidden_states=hidden_states, + attention_mask=attention_mask, + layer_head_mask=layer_head_mask, + output_attentions=output_attentions, + ) + hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training) + hidden_states = residual + hidden_states + hidden_states = self.self_attn_layer_norm(hidden_states) + + residual = hidden_states + hidden_states = self.activation_fn(self.fc1(hidden_states)) + hidden_states = nn.functional.dropout(hidden_states, p=self.activation_dropout, training=self.training) + hidden_states = self.fc2(hidden_states) + hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training) + hidden_states = residual + hidden_states + hidden_states = self.final_layer_norm(hidden_states) + + if hidden_states.dtype == torch.float16 and ( + torch.isinf(hidden_states).any() or torch.isnan(hidden_states).any() + ): + clamp_value = torch.finfo(hidden_states.dtype).max - 1000 + hidden_states = torch.clamp(hidden_states, min=-clamp_value, max=clamp_value) + + outputs = (hidden_states,) + + if output_attentions: + outputs += (attn_weights,) + + return outputs + + +class Florence2DecoderLayer(nn.Module): + def __init__(self, config: Florence2LanguageConfig): + super().__init__() + self.embed_dim = config.d_model + + self.self_attn = FLORENCE2_ATTENTION_CLASSES[config._attn_implementation]( + embed_dim=self.embed_dim, + num_heads=config.decoder_attention_heads, + dropout=config.attention_dropout, + is_decoder=True, + is_causal=True, + config=config, + ) + self.dropout = config.dropout + self.activation_fn = ACT2FN[config.activation_function] + self.activation_dropout = config.activation_dropout + + self.self_attn_layer_norm = nn.LayerNorm(self.embed_dim) + self.encoder_attn = FLORENCE2_ATTENTION_CLASSES[config._attn_implementation]( + self.embed_dim, + config.decoder_attention_heads, + dropout=config.attention_dropout, + is_decoder=True, + config=config, + ) + self.encoder_attn_layer_norm = nn.LayerNorm(self.embed_dim) + self.fc1 = nn.Linear(self.embed_dim, config.decoder_ffn_dim) + self.fc2 = nn.Linear(config.decoder_ffn_dim, self.embed_dim) + self.final_layer_norm = nn.LayerNorm(self.embed_dim) + + def forward( + self, + hidden_states: torch.Tensor, + attention_mask: Optional[torch.Tensor] = None, + encoder_hidden_states: Optional[torch.Tensor] = None, + encoder_attention_mask: Optional[torch.Tensor] = None, + layer_head_mask: Optional[torch.Tensor] = None, + cross_attn_layer_head_mask: Optional[torch.Tensor] = None, + past_key_value: Optional[Tuple[torch.Tensor]] = None, + output_attentions: Optional[bool] = False, + use_cache: Optional[bool] = True, + ) -> Tuple[torch.FloatTensor, Optional[Tuple[torch.FloatTensor, torch.FloatTensor]]]: + """ + Args: + hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` + attention_mask (`torch.FloatTensor`): attention mask of size + `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values. + encoder_hidden_states (`torch.FloatTensor`): + cross attention input to the layer of shape `(batch, seq_len, embed_dim)` + encoder_attention_mask (`torch.FloatTensor`): encoder attention mask of size + `(batch, 1, tgt_len, src_len)` where padding elements are indicated by very large negative values. + layer_head_mask (`torch.FloatTensor`): mask for attention heads in a given layer of size + `(encoder_attention_heads,)`. + cross_attn_layer_head_mask (`torch.FloatTensor`): mask for cross-attention heads in a given layer of + size `(decoder_attention_heads,)`. + past_key_value (`Tuple(torch.FloatTensor)`): cached past key and value projection states + output_attentions (`bool`, *optional*): + Whether or not to return the attentions tensors of all attention layers. See `attentions` under + returned tensors for more detail. + """ + residual = hidden_states + + # Self Attention + # decoder uni-directional self-attention cached key/values tuple is at positions 1,2 + self_attn_past_key_value = past_key_value[:2] if past_key_value is not None else None + # add present self-attn cache to positions 1,2 of present_key_value tuple + hidden_states, self_attn_weights, present_key_value = self.self_attn( + hidden_states=hidden_states, + past_key_value=self_attn_past_key_value, + attention_mask=attention_mask, + layer_head_mask=layer_head_mask, + output_attentions=output_attentions, + ) + hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training) + hidden_states = residual + hidden_states + hidden_states = self.self_attn_layer_norm(hidden_states) + + # Cross-Attention Block + cross_attn_present_key_value = None + cross_attn_weights = None + if encoder_hidden_states is not None: + residual = hidden_states + + # cross_attn cached key/values tuple is at positions 3,4 of present_key_value tuple + cross_attn_past_key_value = past_key_value[-2:] if past_key_value is not None else None + hidden_states, cross_attn_weights, cross_attn_present_key_value = self.encoder_attn( + hidden_states=hidden_states, + key_value_states=encoder_hidden_states, + attention_mask=encoder_attention_mask, + layer_head_mask=cross_attn_layer_head_mask, + past_key_value=cross_attn_past_key_value, + output_attentions=output_attentions, + ) + hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training) + hidden_states = residual + hidden_states + hidden_states = self.encoder_attn_layer_norm(hidden_states) + + # add cross-attn to positions 3,4 of present_key_value tuple + present_key_value = present_key_value + cross_attn_present_key_value + + # Fully Connected + residual = hidden_states + hidden_states = self.activation_fn(self.fc1(hidden_states)) + hidden_states = nn.functional.dropout(hidden_states, p=self.activation_dropout, training=self.training) + hidden_states = self.fc2(hidden_states) + hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training) + hidden_states = residual + hidden_states + hidden_states = self.final_layer_norm(hidden_states) + + outputs = (hidden_states,) + + if output_attentions: + outputs += (self_attn_weights, cross_attn_weights) + + if use_cache: + outputs += (present_key_value,) + + return outputs + + + +class Florence2LanguagePreTrainedModel(PreTrainedModel): + config_class = Florence2LanguageConfig + base_model_prefix = "model" + supports_gradient_checkpointing = True + _keys_to_ignore_on_load_unexpected = ["encoder.version", "decoder.version"] + _no_split_modules = [r"Florence2EncoderLayer", r"Florence2DecoderLayer"] + _skip_keys_device_placement = "past_key_values" + _supports_flash_attn_2 = True + _supports_sdpa = True + + def _init_weights(self, module): + std = self.config.init_std + if isinstance(module, nn.Linear): + 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 dummy_inputs(self): + pad_token = self.config.pad_token_id + input_ids = torch.tensor([[0, 6, 10, 4, 2], [0, 8, 12, 2, pad_token]], device=self.device) + dummy_inputs = { + "attention_mask": input_ids.ne(pad_token), + "input_ids": input_ids, + } + return dummy_inputs + + +class Florence2Encoder(Florence2LanguagePreTrainedModel): + """ + Transformer encoder consisting of *config.encoder_layers* self attention layers. Each layer is a + [`Florence2EncoderLayer`]. + + Args: + config: Florence2LanguageConfig + embed_tokens (nn.Embedding): output embedding + """ + + def __init__(self, config: Florence2LanguageConfig, embed_tokens: Optional[nn.Embedding] = None): + super().__init__(config) + + self.dropout = config.dropout + self.layerdrop = config.encoder_layerdrop + + embed_dim = config.d_model + self.padding_idx = config.pad_token_id + self.max_source_positions = config.max_position_embeddings + embed_scale = math.sqrt(embed_dim) if config.scale_embedding else 1.0 + + self.embed_tokens = Florence2ScaledWordEmbedding( + config.vocab_size, embed_dim, self.padding_idx, embed_scale=embed_scale + ) + + if embed_tokens is not None: + self.embed_tokens.weight = embed_tokens.weight + + self.embed_positions = Florence2LearnedPositionalEmbedding( + config.max_position_embeddings, + embed_dim, + ) + self.layers = nn.ModuleList([Florence2EncoderLayer(config) for _ in range(config.encoder_layers)]) + self._use_flash_attention_2 = config._attn_implementation == "flash_attention_2" + self._use_sdpa = config._attn_implementation == "sdpa" + self.layernorm_embedding = nn.LayerNorm(embed_dim) + + self.gradient_checkpointing = False + # Initialize weights and apply final processing + self.post_init() + + def get_input_embeddings(self): + return self.embed_tokens + + def set_input_embeddings(self, value): + self.embed_tokens = value + + def forward( + self, + input_ids: torch.LongTensor = None, + attention_mask: Optional[torch.Tensor] = None, + head_mask: Optional[torch.Tensor] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, BaseModelOutput]: + 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) + 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) + head_mask (`torch.Tensor` of shape `(encoder_layers, encoder_attention_heads)`, *optional*): + Mask to nullify selected heads of the attention modules. Mask values selected in `[0, 1]`: + + - 1 indicates the head is **not masked**, + - 0 indicates the head is **masked**. + + 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. + 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. + """ + 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 + + # retrieve input_ids and inputs_embeds + if input_ids is not None and inputs_embeds is not None: + raise ValueError("You cannot specify both input_ids and inputs_embeds at the same time") + elif input_ids is not None: + input = input_ids + input_ids = input_ids.view(-1, input_ids.shape[-1]) + elif inputs_embeds is not None: + input = inputs_embeds[:, :, -1] + else: + raise ValueError("You have to specify either input_ids or inputs_embeds") + + if inputs_embeds is None: + inputs_embeds = self.embed_tokens(input_ids) + + embed_pos = self.embed_positions(input) + embed_pos = embed_pos.to(inputs_embeds.device) + + hidden_states = inputs_embeds + embed_pos + hidden_states = self.layernorm_embedding(hidden_states) + hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training) + + # expand attention_mask + if attention_mask is not None: + if self._use_flash_attention_2: + attention_mask = attention_mask if 0 in attention_mask else None + elif self._use_sdpa and head_mask is None and not output_attentions: + # output_attentions=True & head_mask can not be supported when using SDPA, fall back to + # the manual implementation that requires a 4D causal mask in all cases. + # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len] + attention_mask = _prepare_4d_attention_mask_for_sdpa(attention_mask, inputs_embeds.dtype) + else: + # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len] + attention_mask = _prepare_4d_attention_mask(attention_mask, inputs_embeds.dtype) + + encoder_states = () if output_hidden_states else None + all_attentions = () if output_attentions else None + + # check if head_mask has a correct number of layers specified if desired + if head_mask is not None: + if head_mask.size()[0] != (len(self.layers)): + raise ValueError( + f"The head_mask should be specified for {len(self.layers)} layers, but it is for" + f" {head_mask.size()[0]}." + ) + + for idx, encoder_layer in enumerate(self.layers): + if output_hidden_states: + encoder_states = encoder_states + (hidden_states,) + # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description) + to_drop = False + if self.training: + dropout_probability = torch.rand([]) + if dropout_probability < self.layerdrop: # skip the layer + to_drop = True + + if to_drop: + layer_outputs = (None, None) + else: + if self.gradient_checkpointing and self.training: + layer_outputs = self._gradient_checkpointing_func( + encoder_layer.__call__, + hidden_states, + attention_mask, + (head_mask[idx] if head_mask is not None else None), + output_attentions, + ) + else: + layer_outputs = encoder_layer( + hidden_states, + attention_mask, + layer_head_mask=(head_mask[idx] if head_mask is not None else None), + output_attentions=output_attentions, + ) + + hidden_states = layer_outputs[0] + + if output_attentions: + all_attentions = all_attentions + (layer_outputs[1],) + + if output_hidden_states: + encoder_states = encoder_states + (hidden_states,) + + if not return_dict: + return tuple(v for v in [hidden_states, encoder_states, all_attentions] if v is not None) + return BaseModelOutput( + last_hidden_state=hidden_states, hidden_states=encoder_states, attentions=all_attentions + ) + + +class Florence2Decoder(Florence2LanguagePreTrainedModel): + """ + Transformer decoder consisting of *config.decoder_layers* layers. Each layer is a [`Florence2DecoderLayer`] + + Args: + config: Florence2LanguageConfig + embed_tokens (nn.Embedding): output embedding + """ + + def __init__(self, config: Florence2LanguageConfig, embed_tokens: Optional[nn.Embedding] = None): + super().__init__(config) + self.dropout = config.dropout + self.layerdrop = config.decoder_layerdrop + self.padding_idx = config.pad_token_id + self.max_target_positions = config.max_position_embeddings + embed_scale = math.sqrt(config.d_model) if config.scale_embedding else 1.0 + + self.embed_tokens = Florence2ScaledWordEmbedding( + config.vocab_size, config.d_model, self.padding_idx, embed_scale=embed_scale + ) + + if embed_tokens is not None: + self.embed_tokens.weight = embed_tokens.weight + + self.embed_positions = Florence2LearnedPositionalEmbedding( + config.max_position_embeddings, + config.d_model, + ) + self.layers = nn.ModuleList([Florence2DecoderLayer(config) for _ in range(config.decoder_layers)]) + self._use_flash_attention_2 = config._attn_implementation == "flash_attention_2" + self._use_sdpa = config._attn_implementation == "sdpa" + + self.layernorm_embedding = nn.LayerNorm(config.d_model) + + self.gradient_checkpointing = False + # Initialize weights and apply final processing + self.post_init() + + def get_input_embeddings(self): + return self.embed_tokens + + def set_input_embeddings(self, value): + self.embed_tokens = value + + def forward( + self, + input_ids: torch.LongTensor = None, + attention_mask: Optional[torch.Tensor] = None, + encoder_hidden_states: Optional[torch.FloatTensor] = None, + encoder_attention_mask: Optional[torch.LongTensor] = None, + head_mask: Optional[torch.Tensor] = None, + cross_attn_head_mask: Optional[torch.Tensor] = None, + past_key_values: Optional[List[torch.FloatTensor]] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + use_cache: Optional[bool] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, BaseModelOutputWithPastAndCrossAttentions]: + 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) + 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) + encoder_hidden_states (`torch.FloatTensor` of shape `(batch_size, encoder_sequence_length, hidden_size)`, *optional*): + Sequence of hidden-states at the output of the last layer of the encoder. Used in the cross-attention + of the decoder. + encoder_attention_mask (`torch.LongTensor` of shape `(batch_size, encoder_sequence_length)`, *optional*): + Mask to avoid performing cross-attention on padding tokens indices of encoder input_ids. 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) + head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*): + Mask to nullify selected heads of the attention modules. Mask values selected in `[0, 1]`: + + - 1 indicates the head is **not masked**, + - 0 indicates the head is **masked**. + + cross_attn_head_mask (`torch.Tensor` of shape `(decoder_layers, decoder_attention_heads)`, *optional*): + Mask to nullify selected heads of the cross-attention modules in the decoder to avoid performing + cross-attention on hidden heads. Mask values selected in `[0, 1]`: + + - 1 indicates the head is **not masked**, + - 0 indicates the head is **masked**. + + 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. + 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. + """ + 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 + ) + use_cache = use_cache if use_cache is not None else self.config.use_cache + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + # retrieve input_ids and inputs_embeds + if input_ids is not None and inputs_embeds is not None: + raise ValueError("You cannot specify both decoder_input_ids and decoder_inputs_embeds at the same time") + elif input_ids is not None: + input = input_ids + input_shape = input.shape + input_ids = input_ids.view(-1, input_shape[-1]) + elif inputs_embeds is not None: + input_shape = inputs_embeds.size()[:-1] + input = inputs_embeds[:, :, -1] + else: + raise ValueError("You have to specify either decoder_input_ids or decoder_inputs_embeds") + + # past_key_values_length + past_key_values_length = past_key_values[0][0].shape[2] if past_key_values is not None else 0 + + if inputs_embeds is None: + inputs_embeds = self.embed_tokens(input) + + if self._use_flash_attention_2: + # 2d mask is passed through the layers + attention_mask = attention_mask if (attention_mask is not None and 0 in attention_mask) else None + elif self._use_sdpa and not output_attentions and cross_attn_head_mask is None: + # output_attentions=True & cross_attn_head_mask can not be supported when using SDPA, and we fall back on + # the manual implementation that requires a 4D causal mask in all cases. + attention_mask = _prepare_4d_causal_attention_mask_for_sdpa( + attention_mask, + input_shape, + inputs_embeds, + past_key_values_length, + ) + else: + # 4d mask is passed through the layers + attention_mask = _prepare_4d_causal_attention_mask( + attention_mask, input_shape, inputs_embeds, past_key_values_length + ) + + # expand encoder attention mask + if encoder_hidden_states is not None and encoder_attention_mask is not None: + if self._use_flash_attention_2: + encoder_attention_mask = encoder_attention_mask if 0 in encoder_attention_mask else None + elif self._use_sdpa and cross_attn_head_mask is None and not output_attentions: + # output_attentions=True & cross_attn_head_mask can not be supported when using SDPA, and we fall back on + # the manual implementation that requires a 4D causal mask in all cases. + # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len] + encoder_attention_mask = _prepare_4d_attention_mask_for_sdpa( + encoder_attention_mask, + inputs_embeds.dtype, + tgt_len=input_shape[-1], + ) + else: + # [bsz, seq_len] -> [bsz, 1, tgt_seq_len, src_seq_len] + encoder_attention_mask = _prepare_4d_attention_mask( + encoder_attention_mask, inputs_embeds.dtype, tgt_len=input_shape[-1] + ) + + # embed positions + positions = self.embed_positions(input, past_key_values_length) + positions = positions.to(inputs_embeds.device) + + hidden_states = inputs_embeds + positions + hidden_states = self.layernorm_embedding(hidden_states) + + hidden_states = nn.functional.dropout(hidden_states, p=self.dropout, training=self.training) + + if self.gradient_checkpointing and self.training: + if use_cache: + logger.warning_once( + "`use_cache=True` is incompatible with gradient checkpointing. Setting `use_cache=False`..." + ) + use_cache = False + + # decoder layers + all_hidden_states = () if output_hidden_states else None + all_self_attns = () if output_attentions else None + all_cross_attentions = () if (output_attentions and encoder_hidden_states is not None) else None + next_decoder_cache = () if use_cache else None + + # check if head_mask/cross_attn_head_mask has a correct number of layers specified if desired + for attn_mask, mask_name in zip([head_mask, cross_attn_head_mask], ["head_mask", "cross_attn_head_mask"]): + if attn_mask is not None: + if attn_mask.size()[0] != (len(self.layers)): + raise ValueError( + f"The `{mask_name}` should be specified for {len(self.layers)} layers, but it is for" + f" {head_mask.size()[0]}." + ) + + for idx, decoder_layer in enumerate(self.layers): + # add LayerDrop (see https://arxiv.org/abs/1909.11556 for description) + if output_hidden_states: + all_hidden_states += (hidden_states,) + if self.training: + dropout_probability = torch.rand([]) + if dropout_probability < self.layerdrop: + continue + + past_key_value = past_key_values[idx] if past_key_values is not None else None + + if self.gradient_checkpointing and self.training: + layer_outputs = self._gradient_checkpointing_func( + decoder_layer.__call__, + hidden_states, + attention_mask, + encoder_hidden_states, + encoder_attention_mask, + head_mask[idx] if head_mask is not None else None, + cross_attn_head_mask[idx] if cross_attn_head_mask is not None else None, + None, + output_attentions, + use_cache, + ) + else: + layer_outputs = decoder_layer( + hidden_states, + attention_mask=attention_mask, + encoder_hidden_states=encoder_hidden_states, + encoder_attention_mask=encoder_attention_mask, + layer_head_mask=(head_mask[idx] if head_mask is not None else None), + cross_attn_layer_head_mask=( + cross_attn_head_mask[idx] if cross_attn_head_mask is not None else None + ), + past_key_value=past_key_value, + output_attentions=output_attentions, + use_cache=use_cache, + ) + hidden_states = layer_outputs[0] + + if use_cache: + next_decoder_cache += (layer_outputs[3 if output_attentions else 1],) + + if output_attentions: + all_self_attns += (layer_outputs[1],) + + if encoder_hidden_states is not None: + all_cross_attentions += (layer_outputs[2],) + + # add hidden states from the last decoder layer + if output_hidden_states: + all_hidden_states += (hidden_states,) + + next_cache = next_decoder_cache if use_cache else None + if not return_dict: + return tuple( + v + for v in [hidden_states, next_cache, all_hidden_states, all_self_attns, all_cross_attentions] + if v is not None + ) + return BaseModelOutputWithPastAndCrossAttentions( + last_hidden_state=hidden_states, + past_key_values=next_cache, + hidden_states=all_hidden_states, + attentions=all_self_attns, + cross_attentions=all_cross_attentions, + ) + + +class Florence2LanguageModel(Florence2LanguagePreTrainedModel): + _tied_weights_keys = ["encoder.embed_tokens.weight", "decoder.embed_tokens.weight"] + + def __init__(self, config: Florence2LanguageConfig): + super().__init__(config) + + padding_idx, vocab_size = config.pad_token_id, config.vocab_size + self.shared = nn.Embedding(vocab_size, config.d_model, padding_idx) + + self.encoder = Florence2Encoder(config, self.shared) + self.decoder = Florence2Decoder(config, self.shared) + + # Initialize weights and apply final processing + self.post_init() + + def _tie_weights(self): + if self.config.tie_word_embeddings: + self._tie_or_clone_weights(self.encoder.embed_tokens, self.shared) + self._tie_or_clone_weights(self.decoder.embed_tokens, self.shared) + + def get_input_embeddings(self): + return self.shared + + def set_input_embeddings(self, value): + self.shared = value + self.encoder.embed_tokens = self.shared + self.decoder.embed_tokens = self.shared + + def get_encoder(self): + return self.encoder + + def get_decoder(self): + return self.decoder + + def forward( + self, + input_ids: torch.LongTensor = None, + attention_mask: Optional[torch.Tensor] = None, + decoder_input_ids: Optional[torch.LongTensor] = None, + decoder_attention_mask: Optional[torch.LongTensor] = None, + head_mask: Optional[torch.Tensor] = None, + decoder_head_mask: Optional[torch.Tensor] = None, + cross_attn_head_mask: Optional[torch.Tensor] = None, + encoder_outputs: Optional[List[torch.FloatTensor]] = None, + past_key_values: Optional[List[torch.FloatTensor]] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + decoder_inputs_embeds: Optional[torch.FloatTensor] = None, + use_cache: Optional[bool] = None, + output_attentions: Optional[bool] = None, + output_hidden_states: Optional[bool] = None, + return_dict: Optional[bool] = None, + ) -> Union[Tuple, Seq2SeqModelOutput]: + # different to other models, Florence2 automatically creates decoder_input_ids from + # input_ids if no decoder_input_ids are provided + if decoder_input_ids is None and decoder_inputs_embeds is None: + if input_ids is None: + raise ValueError( + "If no `decoder_input_ids` or `decoder_inputs_embeds` are " + "passed, `input_ids` cannot be `None`. Please pass either " + "`input_ids` or `decoder_input_ids` or `decoder_inputs_embeds`." + ) + + decoder_input_ids = shift_tokens_right( + input_ids, self.config.pad_token_id, self.config.decoder_start_token_id + ) + + 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 + ) + use_cache = use_cache if use_cache is not None else self.config.use_cache + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + if encoder_outputs is None: + encoder_outputs = self.encoder( + input_ids=input_ids, + attention_mask=attention_mask, + head_mask=head_mask, + inputs_embeds=inputs_embeds, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + # If the user passed a tuple for encoder_outputs, we wrap it in a BaseModelOutput when return_dict=True + elif return_dict and not isinstance(encoder_outputs, BaseModelOutput): + encoder_outputs = BaseModelOutput( + last_hidden_state=encoder_outputs[0], + hidden_states=encoder_outputs[1] if len(encoder_outputs) > 1 else None, + attentions=encoder_outputs[2] if len(encoder_outputs) > 2 else None, + ) + + # decoder outputs consists of (dec_features, past_key_value, dec_hidden, dec_attn) + decoder_outputs = self.decoder( + input_ids=decoder_input_ids, + attention_mask=decoder_attention_mask, + encoder_hidden_states=encoder_outputs[0], + encoder_attention_mask=attention_mask, + head_mask=decoder_head_mask, + cross_attn_head_mask=cross_attn_head_mask, + past_key_values=past_key_values, + inputs_embeds=decoder_inputs_embeds, + use_cache=use_cache, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + if not return_dict: + return decoder_outputs + encoder_outputs + + return Seq2SeqModelOutput( + last_hidden_state=decoder_outputs.last_hidden_state, + past_key_values=decoder_outputs.past_key_values, + decoder_hidden_states=decoder_outputs.hidden_states, + decoder_attentions=decoder_outputs.attentions, + cross_attentions=decoder_outputs.cross_attentions, + encoder_last_hidden_state=encoder_outputs.last_hidden_state, + encoder_hidden_states=encoder_outputs.hidden_states, + encoder_attentions=encoder_outputs.attentions, + ) + + +class Florence2LanguageForConditionalGeneration(Florence2LanguagePreTrainedModel): + base_model_prefix = "model" + _tied_weights_keys = ["encoder.embed_tokens.weight", "decoder.embed_tokens.weight", "lm_head.weight"] + _keys_to_ignore_on_load_missing = ["final_logits_bias"] + + def __init__(self, config: Florence2LanguageConfig): + super().__init__(config) + self.model = Florence2LanguageModel(config) + self.register_buffer("final_logits_bias", torch.zeros((1, self.model.shared.num_embeddings))) + self.lm_head = nn.Linear(config.d_model, self.model.shared.num_embeddings, bias=False) + + # Initialize weights and apply final processing + self.post_init() + + def get_encoder(self): + return self.model.get_encoder() + + def get_decoder(self): + return self.model.get_decoder() + + def resize_token_embeddings(self, new_num_tokens: int, pad_to_multiple_of: Optional[int] = None) -> nn.Embedding: + new_embeddings = super().resize_token_embeddings(new_num_tokens, pad_to_multiple_of) + self._resize_final_logits_bias(new_embeddings.weight.shape[0]) + return new_embeddings + + def _resize_final_logits_bias(self, new_num_tokens: int) -> None: + old_num_tokens = self.final_logits_bias.shape[-1] + if new_num_tokens <= old_num_tokens: + new_bias = self.final_logits_bias[:, :new_num_tokens] + else: + extra_bias = torch.zeros((1, new_num_tokens - old_num_tokens), device=self.final_logits_bias.device) + new_bias = torch.cat([self.final_logits_bias, extra_bias], dim=1) + self.register_buffer("final_logits_bias", new_bias) + + def get_output_embeddings(self): + return self.lm_head + + def set_output_embeddings(self, new_embeddings): + self.lm_head = new_embeddings + + def forward( + self, + input_ids: torch.LongTensor = None, + attention_mask: Optional[torch.Tensor] = None, + decoder_input_ids: Optional[torch.LongTensor] = None, + decoder_attention_mask: Optional[torch.LongTensor] = None, + head_mask: Optional[torch.Tensor] = None, + decoder_head_mask: Optional[torch.Tensor] = None, + cross_attn_head_mask: Optional[torch.Tensor] = None, + encoder_outputs: Optional[List[torch.FloatTensor]] = None, + past_key_values: Optional[List[torch.FloatTensor]] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + decoder_inputs_embeds: Optional[torch.FloatTensor] = 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, + ) -> Union[Tuple, Seq2SeqLMOutput]: + r""" + 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: + """ + return_dict = return_dict if return_dict is not None else self.config.use_return_dict + + if labels is not None: + if use_cache: + logger.warning("The `use_cache` argument is changed to `False` since `labels` is provided.") + use_cache = False + if decoder_input_ids is None and decoder_inputs_embeds is None: + decoder_input_ids = shift_tokens_right( + labels, self.config.pad_token_id, self.config.decoder_start_token_id + ) + + outputs = self.model( + input_ids, + attention_mask=attention_mask, + decoder_input_ids=decoder_input_ids, + encoder_outputs=encoder_outputs, + decoder_attention_mask=decoder_attention_mask, + head_mask=head_mask, + decoder_head_mask=decoder_head_mask, + cross_attn_head_mask=cross_attn_head_mask, + past_key_values=past_key_values, + inputs_embeds=inputs_embeds, + decoder_inputs_embeds=decoder_inputs_embeds, + use_cache=use_cache, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + lm_logits = self.lm_head(outputs[0]) + lm_logits = lm_logits + self.final_logits_bias.to(lm_logits.device) + + masked_lm_loss = None + if labels is not None: + labels = labels.to(lm_logits.device) + loss_fct = CrossEntropyLoss() + masked_lm_loss = loss_fct(lm_logits.view(-1, self.config.vocab_size), labels.view(-1)) + + if not return_dict: + output = (lm_logits,) + outputs[1:] + return ((masked_lm_loss,) + output) if masked_lm_loss is not None else output + + return Seq2SeqLMOutput( + loss=masked_lm_loss, + logits=lm_logits, + past_key_values=outputs.past_key_values, + decoder_hidden_states=outputs.decoder_hidden_states, + decoder_attentions=outputs.decoder_attentions, + cross_attentions=outputs.cross_attentions, + encoder_last_hidden_state=outputs.encoder_last_hidden_state, + encoder_hidden_states=outputs.encoder_hidden_states, + encoder_attentions=outputs.encoder_attentions, + ) + + def prepare_inputs_for_generation( + self, + decoder_input_ids, + past_key_values=None, + attention_mask=None, + decoder_attention_mask=None, + head_mask=None, + decoder_head_mask=None, + cross_attn_head_mask=None, + use_cache=None, + encoder_outputs=None, + **kwargs, + ): + # cut decoder_input_ids if past_key_values is used + if past_key_values is not None: + past_length = past_key_values[0][0].shape[2] + + # Some generation methods already pass only the last input ID + if decoder_input_ids.shape[1] > past_length: + remove_prefix_length = past_length + else: + # Default to old behavior: keep only final ID + remove_prefix_length = decoder_input_ids.shape[1] - 1 + + decoder_input_ids = decoder_input_ids[:, remove_prefix_length:] + + return { + "input_ids": None, # encoder_outputs is defined. input_ids not needed + "encoder_outputs": encoder_outputs, + "past_key_values": past_key_values, + "decoder_input_ids": decoder_input_ids, + "attention_mask": attention_mask, + "decoder_attention_mask": decoder_attention_mask, + "head_mask": head_mask, + "decoder_head_mask": decoder_head_mask, + "cross_attn_head_mask": cross_attn_head_mask, + "use_cache": use_cache, # change this to avoid caching (presumably for debugging) + } + + def prepare_decoder_input_ids_from_labels(self, labels: torch.Tensor): + return shift_tokens_right(labels, self.config.pad_token_id, self.config.decoder_start_token_id) + + @staticmethod + def _reorder_cache(past_key_values, beam_idx): + reordered_past = () + for layer_past in past_key_values: + # cached cross_attention states don't have to be reordered -> they are always the same + reordered_past += ( + tuple(past_state.index_select(0, beam_idx.to(past_state.device)) for past_state in layer_past[:2]) + + layer_past[2:], + ) + return reordered_past + +@dataclass +class Florence2Seq2SeqLMOutput(ModelOutput): + """ + Base class for Florence-2 model's outputs that also contains : pre-computed hidden states that can speed up sequential + decoding. + + Args: + last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`): + Sequence of hidden-states at the output of the last layer of the decoder of the model. + + If `past_key_values` is used only the last hidden-state of the sequences of shape `(batch_size, 1, + hidden_size)` is output. + 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. + decoder_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 decoder at the output of each layer plus the optional initial embedding outputs. + decoder_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 of the decoder, after the attention softmax, used to compute the weighted average in the + self-attention heads. + cross_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 of the decoder's cross-attention layer, after the attention softmax, used to compute the + weighted average in the cross-attention heads. + encoder_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 encoder of the model. + encoder_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 encoder at the output of each layer plus the optional initial embedding outputs. + encoder_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 of the encoder, 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_image_tokens, hidden_size)`. + + image_hidden_states of the model produced by the vision encoder + """ + + last_hidden_state: torch.FloatTensor = None + past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None + decoder_hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None + decoder_attentions: Optional[Tuple[torch.FloatTensor, ...]] = None + cross_attentions: Optional[Tuple[torch.FloatTensor, ...]] = None + encoder_last_hidden_state: Optional[torch.FloatTensor] = None + encoder_hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None + encoder_attentions: Optional[Tuple[torch.FloatTensor, ...]] = None + + +FLORENCE2_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 ([`Florence2Config`] or [`Florence2VisionConfig`]): + 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 Florence-2 Model outputting raw hidden-states without any specific head on top.", + FLORENCE2_START_DOCSTRING, +) +class Florence2PreTrainedModel(PreTrainedModel): + config_class = Florence2Config + base_model_prefix = "model" + supports_gradient_checkpointing = True + _skip_keys_device_placement = "past_key_values" + + @property + def _supports_flash_attn_2(self): + """ + Retrieve language_model's attribute to check whether the model supports + Flash Attention 2 or not. + """ + return self.language_model._supports_flash_attn_2 + + @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 + + +FLORENCE2_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 ([]`Florence2Processor`] 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 FLORENCE2 vision model without any head""", + FLORENCE2_START_DOCSTRING, +) +class Florence2VisionModel(Florence2PreTrainedModel): + def __init__(self, config: Florence2VisionConfig): + super().__init__(config) + assert config.model_type == 'davit', 'only DaViT is supported for now' + self.vision_tower = DaViT.from_config(config=config) + + self.post_init() + + def forward(self, pixel_values): + if len(pixel_values.shape) == 4: + x = self.vision_tower.forward_features_unpool(pixel_values) + else: + raise ValueError(f'invalid image shape {pixel_values.shape}') + return x + + +@add_start_docstrings( + """The FLORENCE2 vision model with projection layer""", + FLORENCE2_START_DOCSTRING, +) +class Florence2VisionModelWithProjection(Florence2PreTrainedModel): + def __init__(self, config: Florence2VisionConfig): + super().__init__(config) + assert config.model_type == 'davit', 'only DaViT is supported for now' + self.vision_tower = DaViT.from_config(config=config) + + self._build_image_projection_layers(config) + + self.post_init() + + def _build_image_projection_layers(self, config): + image_dim_out = config.dim_embed[-1] + dim_projection = config.projection_dim + self.image_projection = nn.Parameter( + torch.empty(image_dim_out, dim_projection) + ) + self.image_proj_norm = nn.LayerNorm(dim_projection) + image_pos_embed_config = config.image_pos_embed + if image_pos_embed_config['type'] == 'learned_abs_2d': + self.image_pos_embed = LearnedAbsolutePositionEmbedding2D( + embedding_dim=image_dim_out, + num_pos=image_pos_embed_config['max_pos_embeddings'] + ) + else: + raise NotImplementedError('Not implemented yet') + + self.image_feature_source = config.image_feature_source + + # temporal embedding + visual_temporal_embedding_config = config.visual_temporal_embedding + if visual_temporal_embedding_config['type'] == 'COSINE': + self.visual_temporal_embed = PositionalEmbeddingCosine1D( + embed_dim=image_dim_out, + max_seq_len=visual_temporal_embedding_config['max_temporal_embeddings'] + ) + else: + raise NotImplementedError('Not implemented yet') + + def forward(self, pixel_values): + if len(pixel_values.shape) == 4: + batch_size, C, H, W = pixel_values.shape + T = 1 + x = self.vision_tower.forward_features_unpool(pixel_values) + else: + raise ValueError(f'invalid image shape {pixel_values.shape}') + + if self.image_pos_embed is not None: + x = x.view(batch_size * T, -1, x.shape[-1]) + num_tokens = x.shape[-2] + h, w = int(num_tokens ** 0.5), int(num_tokens ** 0.5) + assert h * w == num_tokens, 'only support square feature maps for now' + x = x.view(batch_size * T, h, w, x.shape[-1]) + pos_embed = self.image_pos_embed(x) + x = x + pos_embed + x = x.view(batch_size, T * h*w, x.shape[-1]) + + if self.visual_temporal_embed is not None: + visual_temporal_embed = self.visual_temporal_embed(x.view(batch_size, T, -1, x.shape[-1])[:, :, 0]) + x = x.view(batch_size, T, -1, x.shape[-1]) + visual_temporal_embed.view(1, T, 1, x.shape[-1]) + + x_feat_dict = {} + + spatial_avg_pool_x = x.view(batch_size, T, -1, x.shape[-1]).mean(dim=2) + x_feat_dict['spatial_avg_pool'] = spatial_avg_pool_x + + temporal_avg_pool_x = x.view(batch_size, T, -1, x.shape[-1]).mean(dim=1) + x_feat_dict['temporal_avg_pool'] = temporal_avg_pool_x + + x = x.view(batch_size, T, -1, x.shape[-1])[:, -1] + x_feat_dict['last_frame'] = x + + new_x = [] + for _image_feature_source in self.image_feature_source: + if _image_feature_source not in x_feat_dict: + raise ValueError('invalid image feature source: {}'.format(_image_feature_source)) + new_x.append(x_feat_dict[_image_feature_source]) + + x = torch.cat(new_x, dim=1) + + x = x @ self.image_projection + x = self.image_proj_norm(x) + + + return x + + + +@add_start_docstrings( + """The FLORENCE2 model which consists of a vision backbone and a language model.""", + FLORENCE2_START_DOCSTRING, +) +class Florence2ForConditionalGeneration(Florence2PreTrainedModel): + def __init__(self, config: Florence2Config): + super().__init__(config) + assert config.vision_config.model_type == 'davit', 'only DaViT is supported for now' + del config.vision_config.model_type + self.vision_tower = DaViT.from_config(config=config.vision_config) + # remove unused layers + del self.vision_tower.head + del self.vision_tower.norms + + self.vocab_size = config.vocab_size + self._attn_implementation = config._attn_implementation + self._build_image_projection_layers(config) + + language_model = Florence2LanguageForConditionalGeneration(config=config.text_config) + + if language_model._tied_weights_keys is not None: + self._tied_weights_keys = [f"language_model.{k}" for k in language_model._tied_weights_keys] + self.language_model = language_model + + self.pad_token_id = self.config.pad_token_id if self.config.pad_token_id is not None else -1 + self.post_init() + + def _build_image_projection_layers(self, config): + image_dim_out = config.vision_config.dim_embed[-1] + dim_projection = config.vision_config.projection_dim + self.image_projection = nn.Parameter( + torch.empty(image_dim_out, dim_projection) + ) + self.image_proj_norm = nn.LayerNorm(dim_projection) + image_pos_embed_config = config.vision_config.image_pos_embed + if image_pos_embed_config['type'] == 'learned_abs_2d': + self.image_pos_embed = LearnedAbsolutePositionEmbedding2D( + embedding_dim=image_dim_out, + num_pos=image_pos_embed_config['max_pos_embeddings'] + ) + else: + raise NotImplementedError('Not implemented yet') + + self.image_feature_source = config.vision_config.image_feature_source + + # temporal embedding + visual_temporal_embedding_config = config.vision_config.visual_temporal_embedding + if visual_temporal_embedding_config['type'] == 'COSINE': + self.visual_temporal_embed = PositionalEmbeddingCosine1D( + embed_dim=image_dim_out, + max_seq_len=visual_temporal_embedding_config['max_temporal_embeddings'] + ) + else: + raise NotImplementedError('Not implemented yet') + + def get_encoder(self): + return self.language_model.get_encoder() + + def get_decoder(self): + return self.language_model.get_decoder() + + def get_input_embeddings(self): + return self.language_model.get_input_embeddings() + + 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 _encode_image(self, pixel_values): + if len(pixel_values.shape) == 4: + batch_size, C, H, W = pixel_values.shape + T = 1 + x = self.vision_tower.forward_features_unpool(pixel_values) + else: + raise ValueError(f'invalid image shape {pixel_values.shape}') + + if self.image_pos_embed is not None: + x = x.view(batch_size * T, -1, x.shape[-1]) + num_tokens = x.shape[-2] + h, w = int(num_tokens ** 0.5), int(num_tokens ** 0.5) + assert h * w == num_tokens, 'only support square feature maps for now' + x = x.view(batch_size * T, h, w, x.shape[-1]) + pos_embed = self.image_pos_embed(x) + x = x + pos_embed + x = x.view(batch_size, T * h*w, x.shape[-1]) + + if self.visual_temporal_embed is not None: + visual_temporal_embed = self.visual_temporal_embed(x.view(batch_size, T, -1, x.shape[-1])[:, :, 0]) + x = x.view(batch_size, T, -1, x.shape[-1]) + visual_temporal_embed.view(1, T, 1, x.shape[-1]) + + x_feat_dict = {} + + spatial_avg_pool_x = x.view(batch_size, T, -1, x.shape[-1]).mean(dim=2) + x_feat_dict['spatial_avg_pool'] = spatial_avg_pool_x + + temporal_avg_pool_x = x.view(batch_size, T, -1, x.shape[-1]).mean(dim=1) + x_feat_dict['temporal_avg_pool'] = temporal_avg_pool_x + + x = x.view(batch_size, T, -1, x.shape[-1])[:, -1] + x_feat_dict['last_frame'] = x + + new_x = [] + for _image_feature_source in self.image_feature_source: + if _image_feature_source not in x_feat_dict: + raise ValueError('invalid image feature source: {}'.format(_image_feature_source)) + new_x.append(x_feat_dict[_image_feature_source]) + + x = torch.cat(new_x, dim=1) + + x = x @ self.image_projection + x = self.image_proj_norm(x) + + return x + + def _merge_input_ids_with_image_features( + self, image_features, inputs_embeds + ): + batch_size, image_token_length = image_features.size()[:-1] + device = image_features.device + image_attention_mask = torch.ones(batch_size, image_token_length, device=device) + + # task_prefix_embeds: [batch_size, padded_context_length, hidden_size] + # task_prefix_attention_mask: [batch_size, context_length] + if inputs_embeds is None: + return image_features, image_attention_mask + + task_prefix_embeds = inputs_embeds + task_prefix_attention_mask = torch.ones(batch_size, task_prefix_embeds.size(1), device=device) + + if len(task_prefix_attention_mask.shape) == 3: + task_prefix_attention_mask = task_prefix_attention_mask[:, 0] + + # concat [image embeds, task prefix embeds] + inputs_embeds = torch.cat([image_features, task_prefix_embeds], dim=1) + attention_mask = torch.cat([image_attention_mask, task_prefix_attention_mask], dim=1) + + return inputs_embeds, attention_mask + + + @add_start_docstrings_to_model_forward(FLORENCE2_INPUTS_DOCSTRING) + @replace_return_docstrings(output_type=Florence2Seq2SeqLMOutput, config_class=_CONFIG_FOR_DOC) + def forward( + self, + input_ids: torch.LongTensor = None, + pixel_values: torch.FloatTensor = None, + attention_mask: Optional[torch.Tensor] = None, + decoder_input_ids: Optional[torch.LongTensor] = None, + decoder_attention_mask: Optional[torch.LongTensor] = None, + head_mask: Optional[torch.Tensor] = None, + decoder_head_mask: Optional[torch.Tensor] = None, + cross_attn_head_mask: Optional[torch.Tensor] = None, + encoder_outputs: Optional[List[torch.FloatTensor]] = None, + past_key_values: Optional[List[torch.FloatTensor]] = None, + inputs_embeds: Optional[torch.FloatTensor] = None, + decoder_inputs_embeds: Optional[torch.FloatTensor] = 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, + ) -> Union[Tuple, Florence2Seq2SeqLMOutput]: + 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, Florence2ForConditionalGeneration + + >>> model = Florence2ForConditionalGeneration.from_pretrained("microsoft/Florence-2-large") + >>> processor = AutoProcessor.from_pretrained("microsoft/Florence-2-large") + + >>> prompt = "" + >>> url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.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=100) + >>> processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0] + "A green car parked in front of a yellow building." + ```""" + 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 + + image_features = None + if inputs_embeds is None: + # 1. Extra the input embeddings + if input_ids is not None: + inputs_embeds = self.get_input_embeddings()(input_ids) + # 2. Merge text and images + if pixel_values is not None: + # (batch_size, num_image_tokens, hidden_size) + image_features = self._encode_image(pixel_values) + inputs_embeds, attention_mask = self._merge_input_ids_with_image_features(image_features, inputs_embeds) + + attention_mask = attention_mask.to(inputs_embeds.dtype) + outputs = self.language_model( + attention_mask=attention_mask, + labels=labels, + inputs_embeds=inputs_embeds, + decoder_input_ids=decoder_input_ids, + encoder_outputs=encoder_outputs, + decoder_attention_mask=decoder_attention_mask, + head_mask=head_mask, + decoder_head_mask=decoder_head_mask, + cross_attn_head_mask=cross_attn_head_mask, + past_key_values=past_key_values, + decoder_inputs_embeds=decoder_inputs_embeds, + use_cache=use_cache, + output_attentions=output_attentions, + output_hidden_states=output_hidden_states, + return_dict=return_dict, + ) + + logits = outputs.logits + logits = logits.float() + loss = outputs.loss + if not return_dict: + output = (logits,) + outputs[1:] + return (loss,) + output if loss is not None else output + + return Florence2Seq2SeqLMOutput( + loss=loss, + logits=logits, + past_key_values=outputs.past_key_values, + decoder_hidden_states=outputs.decoder_hidden_states, + decoder_attentions=outputs.decoder_attentions, + cross_attentions=outputs.cross_attentions, + encoder_last_hidden_state=outputs.encoder_last_hidden_state, + encoder_hidden_states=outputs.encoder_hidden_states, + encoder_attentions=outputs.encoder_attentions, + image_hidden_states=image_features + ) + + def generate( + self, + input_ids, + inputs_embeds=None, + pixel_values=None, + **kwargs + ): + + if inputs_embeds is None: + # 1. Extra the input embeddings + if input_ids is not None: + inputs_embeds = self.get_input_embeddings()(input_ids) + # 2. Merge text and images + if pixel_values is not None: + image_features = self._encode_image(pixel_values) + inputs_embeds, attention_mask = self._merge_input_ids_with_image_features(image_features, inputs_embeds) + + return self.language_model.generate( + input_ids=None, + inputs_embeds=inputs_embeds, + **kwargs + ) + + def prepare_inputs_for_generation( + self, + decoder_input_ids, + past_key_values=None, + attention_mask=None, + pixel_values=None, + decoder_attention_mask=None, + head_mask=None, + decoder_head_mask=None, + cross_attn_head_mask=None, + use_cache=None, + encoder_outputs=None, + **kwargs, + ): + # cut decoder_input_ids if past_key_values is used + if past_key_values is not None: + past_length = past_key_values[0][0].shape[2] + + # Some generation methods already pass only the last input ID + if decoder_input_ids.shape[1] > past_length: + remove_prefix_length = past_length + else: + # Default to old behavior: keep only final ID + remove_prefix_length = decoder_input_ids.shape[1] - 1 + + decoder_input_ids = decoder_input_ids[:, remove_prefix_length:] + + return { + "input_ids": None, # encoder_outputs is defined. input_ids not needed + "encoder_outputs": encoder_outputs, + "past_key_values": past_key_values, + "decoder_input_ids": decoder_input_ids, + "attention_mask": attention_mask, + "pixel_values": pixel_values, + "decoder_attention_mask": decoder_attention_mask, + "head_mask": head_mask, + "decoder_head_mask": decoder_head_mask, + "cross_attn_head_mask": cross_attn_head_mask, + "use_cache": use_cache, # change this to avoid caching (presumably for debugging) + } + + def prepare_decoder_input_ids_from_labels(self, labels: torch.Tensor): + return self.language_model.shift_tokens_right(labels) + + def _reorder_cache(self, *args, **kwargs): + return self.language_model._reorder_cache(*args, **kwargs) \ No newline at end of file From c2221a2175df0a4304d2d85852c4be2dbe3637ce Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 03:14:20 +0200 Subject: [PATCH 06/35] add florence2 __init__.py --- src/transformers/models/florence/__init__.py | 77 ++++++++++++++++++++ 1 file changed, 77 insertions(+) create mode 100644 src/transformers/models/florence/__init__.py diff --git a/src/transformers/models/florence/__init__.py b/src/transformers/models/florence/__init__.py new file mode 100644 index 000000000000..5b532dff1e86 --- /dev/null +++ b/src/transformers/models/florence/__init__.py @@ -0,0 +1,77 @@ +# Copyright 2024 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 ( + OptionalDependencyNotAvailable, + _LazyModule, + is_torch_available, +) + + +_import_structure = { + "configuration_florence2": [ + "Florence2VisionConfig", + "Florence2LanguageConfig", + "Florence2Config", + ], + "processing_florence2": [ + "Florence2Processor", + "Florence2PostProcesser", + ], +} + + +try: + if not is_torch_available(): + raise OptionalDependencyNotAvailable() +except OptionalDependencyNotAvailable: + pass +else: + _import_structure["modeling_florence2"] = [ + "Florence2PreTrainedModel", + "Florence2VisionModel", + "Florence2VisionModelWithProjection", + "Florence2ForConditionalGeneration", + ] + + +if TYPE_CHECKING: + from .configuration_florence2 import ( + Florence2VisionConfig, + Florence2LanguageConfig, + Florence2Config, + ) + from .processing_florence2 import ( + Florence2Processor, + Florence2PostProcesser, + ) + + try: + if not is_torch_available(): + raise OptionalDependencyNotAvailable() + except OptionalDependencyNotAvailable: + pass + else: + from .modeling_florence2 import ( + Florence2PreTrainedModel, + Florence2VisionModel, + Florence2VisionModelWithProjection, + Florence2ForConditionalGeneration, + ) + +else: + import sys + + sys.modules[__name__] = _LazyModule(__name__, globals()["__file__"], _import_structure) From 6d1606ed29209066f6190406d3992824bfc767e7 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 15:47:58 +0200 Subject: [PATCH 07/35] make fixup --- src/transformers/models/florence/__init__.py | 14 +- .../florence/configuration_florence2.py | 13 +- .../models/florence/modeling_florence2.py | 435 +++++----- .../models/florence/processing_florence2.py | 747 ++++++++++-------- 4 files changed, 631 insertions(+), 578 deletions(-) diff --git a/src/transformers/models/florence/__init__.py b/src/transformers/models/florence/__init__.py index 5b532dff1e86..262e7173ca64 100644 --- a/src/transformers/models/florence/__init__.py +++ b/src/transformers/models/florence/__init__.py @@ -22,13 +22,13 @@ _import_structure = { "configuration_florence2": [ - "Florence2VisionConfig", - "Florence2LanguageConfig", "Florence2Config", + "Florence2LanguageConfig", + "Florence2VisionConfig", ], "processing_florence2": [ - "Florence2Processor", "Florence2PostProcesser", + "Florence2Processor", ], } @@ -49,13 +49,13 @@ if TYPE_CHECKING: from .configuration_florence2 import ( - Florence2VisionConfig, - Florence2LanguageConfig, Florence2Config, + Florence2LanguageConfig, + Florence2VisionConfig, ) from .processing_florence2 import ( - Florence2Processor, Florence2PostProcesser, + Florence2Processor, ) try: @@ -65,10 +65,10 @@ pass else: from .modeling_florence2 import ( + Florence2ForConditionalGeneration, Florence2PreTrainedModel, Florence2VisionModel, Florence2VisionModelWithProjection, - Florence2ForConditionalGeneration, ) else: diff --git a/src/transformers/models/florence/configuration_florence2.py b/src/transformers/models/florence/configuration_florence2.py index 6dd576ee1be5..643d6378dfa5 100644 --- a/src/transformers/models/florence/configuration_florence2.py +++ b/src/transformers/models/florence/configuration_florence2.py @@ -12,19 +12,22 @@ # See the License for the specific language governing permissions and # limitations under the License. import warnings + + """ Florence-2 configuration""" -from typing import Optional from ...configuration_utils import PretrainedConfig from ...utils import logging + logger = logging.get_logger(__name__) + class Florence2VisionConfig(PretrainedConfig): r""" This is the configuration class to store the configuration of a [`Florence2VisionModel`]. It is used to instantiate a Florence2VisionModel - according to the specified arguments, defining the model architecture. Instantiating a configuration with the + according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the Florence2VisionModel architecture. Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the @@ -271,7 +274,7 @@ def __init__( class Florence2Config(PretrainedConfig): r""" This is the configuration class to store the configuration of a [`Florence2ForConditionalGeneration`]. It is used to instantiate an - Florence-2 model according to the specified arguments, defining the model architecture. + Florence-2 model according to the specified arguments, defining the model architecture. Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the documentation from [`PretrainedConfig`] for more information. @@ -280,7 +283,7 @@ class Florence2Config(PretrainedConfig): vision_config (`Florence2VisionConfig`, *optional*): Custom vision config or dict text_config (`Union[AutoConfig, dict]`, *optional*): - The config object of the text backbone. + The config object of the text backbone. ignore_index (`int`, *optional*, defaults to -100): The ignore index for the loss function. vocab_size (`int`, *optional*, defaults to 51289): @@ -334,6 +337,4 @@ def __init__( if text_config is not None: self.text_config = Florence2LanguageConfig(**text_config) - super().__init__(**kwargs) - diff --git a/src/transformers/models/florence/modeling_florence2.py b/src/transformers/models/florence/modeling_florence2.py index 41a71c8b5624..c6e09c303ed9 100644 --- a/src/transformers/models/florence/modeling_florence2.py +++ b/src/transformers/models/florence/modeling_florence2.py @@ -13,31 +13,45 @@ # See the License for the specific language governing permissions and # limitations under the License. -""" PyTorch Florence-2 model.""" -from dataclasses import dataclass -from typing import List, Optional, Tuple, Union +"""PyTorch Florence-2 model.""" import math import warnings +from collections import OrderedDict +from dataclasses import dataclass +from typing import List, Optional, Tuple, Union + import torch -import torch.utils.checkpoint -from torch import nn import torch.nn.functional as F +import torch.utils.checkpoint import torch.utils.checkpoint as checkpoint -from torch.nn import CrossEntropyLoss -from collections import OrderedDict -# from einops import rearrange +from torch import nn +from torch.nn import CrossEntropyLoss +from ...activations import ACT2FN +from ...modeling_attn_mask_utils import ( + _prepare_4d_attention_mask, + _prepare_4d_attention_mask_for_sdpa, + _prepare_4d_causal_attention_mask, + _prepare_4d_causal_attention_mask_for_sdpa, +) +from ...modeling_outputs import ( + BaseModelOutput, + BaseModelOutputWithPastAndCrossAttentions, + Seq2SeqLMOutput, + Seq2SeqModelOutput, +) + +# from einops import rearrange from ...modeling_utils import PreTrainedModel from ...utils import ( ModelOutput, add_start_docstrings, add_start_docstrings_to_model_forward, is_flash_attn_2_available, + is_flash_attn_greater_or_equal_2_10, logging, replace_return_docstrings, - is_flash_attn_2_available, - is_flash_attn_greater_or_equal_2_10, ) from .configuration_florence2 import ( Florence2Config, @@ -46,21 +60,6 @@ ) -from ...activations import ACT2FN -from ...modeling_attn_mask_utils import ( - _prepare_4d_attention_mask, - _prepare_4d_attention_mask_for_sdpa, - _prepare_4d_causal_attention_mask, - _prepare_4d_causal_attention_mask_for_sdpa, -) -from ...modeling_outputs import ( - BaseModelOutput, - BaseModelOutputWithPastAndCrossAttentions, - Seq2SeqLMOutput, - Seq2SeqModelOutput, -) - - if is_flash_attn_2_available(): from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input # noqa @@ -68,6 +67,7 @@ _CONFIG_FOR_DOC = "Florence2Config" + class LearnedAbsolutePositionEmbedding2D(nn.Module): """ This module learns positional embeddings up to a fixed maximum size. @@ -80,11 +80,11 @@ def __init__(self, embedding_dim=256, num_pos=50): def forward(self, pixel_values): """ - pixel_values: (batch_size, height, width, num_channels) + pixel_values: (batch_size, height, width, num_channels) returns: (batch_size, height, width, embedding_dim * 2) """ if len(pixel_values.shape) != 4: - raise ValueError('pixel_values must be a 4D tensor') + raise ValueError("pixel_values must be a 4D tensor") height, width = pixel_values.shape[1:3] width_values = torch.arange(width, device=pixel_values.device) height_values = torch.arange(height, device=pixel_values.device) @@ -101,6 +101,7 @@ def forward(self, pixel_values): pos = pos.permute(0, 2, 3, 1) return pos + class PositionalEmbeddingCosine1D(nn.Module): """ This class implements a very simple positional encoding. It follows closely @@ -112,22 +113,17 @@ class PositionalEmbeddingCosine1D(nn.Module): dropout_prob: The dropout probability. max_seq_len: The maximum length to precompute the positional encodings. """ - def __init__( - self, - embed_dim: int = 512, - max_seq_len: int = 1024) -> None: + + def __init__(self, embed_dim: int = 512, max_seq_len: int = 1024) -> None: super(PositionalEmbeddingCosine1D, self).__init__() self.embed_dim = embed_dim self.max_seq_len = max_seq_len # Generate the sinusoidal arrays. factor = math.log(10000) - denominator = torch.exp( - -factor * torch.arange(0, self.embed_dim, 2) / self.embed_dim) + denominator = torch.exp(-factor * torch.arange(0, self.embed_dim, 2) / self.embed_dim) # Matrix where rows correspond to a positional embedding as a function # of the position index (i.e., the row index). - frequencies = \ - torch.arange(0, self.max_seq_len) \ - .reshape(self.max_seq_len, 1) * denominator + frequencies = torch.arange(0, self.max_seq_len).reshape(self.max_seq_len, 1) * denominator pos_idx_to_embed = torch.zeros((self.max_seq_len, self.embed_dim)) # Populate uneven entries. pos_idx_to_embed[:, 0::2] = torch.sin(frequencies) @@ -151,11 +147,10 @@ def forward(self, seq_embeds: torch.Tensor) -> torch.Tensor: assert 2 <= shape_len <= 3 len_seq = seq_embeds.size(-2) assert len_seq <= self.max_seq_len - pos_embeds = self.pos_idx_to_embed[0:seq_embeds.size(-2), :] + pos_embeds = self.pos_idx_to_embed[0 : seq_embeds.size(-2), :] # Adapt pre-computed positional embeddings to the input. if shape_len == 3: - pos_embeds = pos_embeds.view( - (1, pos_embeds.size(0), pos_embeds.size(1))) + pos_embeds = pos_embeds.view((1, pos_embeds.size(0), pos_embeds.size(1))) return pos_embeds @@ -167,10 +162,8 @@ class LearnedAbsolutePositionEmbedding1D(nn.Module): embed_dim: The dimension of the embeddings. max_seq_len: The maximum length to precompute the positional encodings. """ - def __init__( - self, - embedding_dim: int = 512, - num_pos: int = 1024) -> None: + + def __init__(self, embedding_dim: int = 512, num_pos: int = 1024) -> None: super(LearnedAbsolutePositionEmbedding1D, self).__init__() self.embeddings = nn.Embedding(num_pos, embedding_dim) self.num_pos = num_pos @@ -195,12 +188,10 @@ def forward(self, seq_embeds: torch.Tensor) -> torch.Tensor: pos_embeds = self.embeddings(torch.arange(len_seq).to(seq_embeds.device)) # Adapt pre-computed positional embeddings to the input. if shape_len == 3: - pos_embeds = pos_embeds.view( - (1, pos_embeds.size(0), pos_embeds.size(1))) + pos_embeds = pos_embeds.view((1, pos_embeds.size(0), pos_embeds.size(1))) return pos_embeds - class MySequential(nn.Sequential): def forward(self, *inputs): for module in self._modules.values(): @@ -244,11 +235,15 @@ def __init__( super().__init__() out_features = out_features or in_features hidden_features = hidden_features or in_features - self.net = nn.Sequential(OrderedDict([ - ("fc1", nn.Linear(in_features, hidden_features)), - ("act", act_layer()), - ("fc2", nn.Linear(hidden_features, out_features)) - ])) + self.net = nn.Sequential( + OrderedDict( + [ + ("fc1", nn.Linear(in_features, hidden_features)), + ("act", act_layer()), + ("fc2", nn.Linear(hidden_features, out_features)), + ] + ) + ) def forward(self, x, size): return self.net(x), size @@ -265,12 +260,7 @@ def __init__( ): super().__init__() self.dw = nn.Conv2d( - dim_in, dim_in, - kernel_size=kernel_size, - padding=padding, - groups=dim_in, - stride=stride, - bias=bias + dim_in, dim_in, kernel_size=kernel_size, padding=padding, groups=dim_in, stride=stride, bias=bias ) def forward(self, x, size): @@ -285,28 +275,13 @@ def forward(self, x, size): class ConvEmbed(nn.Module): - """ Image to Patch Embedding - """ + """Image to Patch Embedding""" - def __init__( - self, - patch_size=7, - in_chans=3, - embed_dim=64, - stride=4, - padding=2, - norm_layer=None, - pre_norm=True - ): + def __init__(self, patch_size=7, in_chans=3, embed_dim=64, stride=4, padding=2, norm_layer=None, pre_norm=True): super().__init__() self.patch_size = patch_size - self.proj = nn.Conv2d( - in_chans, embed_dim, - kernel_size=patch_size, - stride=stride, - padding=padding - ) + self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=patch_size, stride=stride, padding=padding) dim_norm = in_chans if pre_norm else embed_dim self.norm = norm_layer(dim_norm) if norm_layer else None @@ -328,7 +303,7 @@ def forward(self, x, size): _, _, H, W = x.shape # TODO: check if this is correct - #x = rearrange(x, 'b c h w -> b (h w) c') + # x = rearrange(x, 'b c h w -> b (h w) c') x = x.permute(0, 2, 3, 1).contiguous() B, H, W, C = x.size() x = x.view(B, -1, C) @@ -340,7 +315,6 @@ def forward(self, x, size): class ChannelAttention(nn.Module): - def __init__(self, dim, groups=8, qkv_bias=True): super().__init__() @@ -363,17 +337,17 @@ def forward(self, x, size): return x, size -#https://github.com/huggingface/pytorch-image-models/blob/main/timm/layers/drop.py +# https://github.com/huggingface/pytorch-image-models/blob/main/timm/layers/drop.py class DropPath(nn.Module): - """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). - """ - def __init__(self, drop_prob: float = 0., scale_by_keep: bool = True): + """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).""" + + def __init__(self, drop_prob: float = 0.0, scale_by_keep: bool = True): super(DropPath, self).__init__() self.drop_prob = drop_prob self.scale_by_keep = scale_by_keep def forward(self, x): - if self.drop_prob == 0. or not self.training: + if self.drop_prob == 0.0 or not self.training: return x keep_prob = 1 - self.drop_prob shape = (x.shape[0],) + (1,) * (x.ndim - 1) # work with diff dim tensors, not just 2D ConvNets @@ -383,12 +357,12 @@ def forward(self, x): return x * random_tensor def extra_repr(self): - return f'drop_prob={round(self.drop_prob,3):0.3f}' + return f"drop_prob={round(self.drop_prob,3):0.3f}" # TODO: can this be replaced with torch.nn.init.trunc_normal_? -#https://github.com/huggingface/pytorch-image-models/blob/b28945ff056d454b174f0fb8682e362b87150141/timm/layers/weight_init.py -def trunc_normal_(tensor: torch.Tensor, mean=0., std=1., a=-2., b=2.) -> torch.Tensor: +# https://github.com/huggingface/pytorch-image-models/blob/b28945ff056d454b174f0fb8682e362b87150141/timm/layers/weight_init.py +def trunc_normal_(tensor: torch.Tensor, mean=0.0, std=1.0, a=-2.0, b=2.0) -> torch.Tensor: r"""Fills the input Tensor with values drawn from a truncated normal distribution. The values are effectively drawn from the normal distribution :math:`\mathcal{N}(\text{mean}, \text{std}^2)` @@ -407,17 +381,20 @@ def trunc_normal_(tensor: torch.Tensor, mean=0., std=1., a=-2., b=2.) -> torch.T a: the minimum cutoff value b: the maximum cutoff value """ + # Method based on https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf def norm_cdf(x): # Computes standard normal cumulative distribution function - return (1. + math.erf(x / math.sqrt(2.))) / 2. - + return (1.0 + math.erf(x / math.sqrt(2.0))) / 2.0 + with torch.no_grad(): # Cut & paste from PyTorch official master until it's in a few official releases - RW if (mean < a - 2 * std) or (mean > b + 2 * std): - warnings.warn("mean is more than 2 std from [a, b] in nn.init.trunc_normal_. " - "The distribution of values may be incorrect.", - stacklevel=2) + warnings.warn( + "mean is more than 2 std from [a, b] in nn.init.trunc_normal_. " + "The distribution of values may be incorrect.", + stacklevel=2, + ) # Values are generated by using a truncated uniform distribution and # then using the inverse CDF for the normal distribution. @@ -434,7 +411,7 @@ def norm_cdf(x): tensor.erfinv_() # Transform to proper mean, std - tensor.mul_(std * math.sqrt(2.)) + tensor.mul_(std * math.sqrt(2.0)) tensor.add_(mean) # Clamp to ensure it's in the proper range @@ -443,25 +420,29 @@ def norm_cdf(x): class ChannelBlock(nn.Module): - - def __init__(self, dim, groups, mlp_ratio=4., qkv_bias=True, - drop_path_rate=0., act_layer=nn.GELU, norm_layer=nn.LayerNorm, - conv_at_attn=True, conv_at_ffn=True): + def __init__( + self, + dim, + groups, + mlp_ratio=4.0, + qkv_bias=True, + drop_path_rate=0.0, + act_layer=nn.GELU, + norm_layer=nn.LayerNorm, + conv_at_attn=True, + conv_at_ffn=True, + ): super().__init__() - drop_path = DropPath(drop_path_rate) if drop_path_rate > 0. else nn.Identity() + drop_path = DropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() self.conv1 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None self.channel_attn = PreNorm( - norm_layer(dim), - ChannelAttention(dim, groups=groups, qkv_bias=qkv_bias), - drop_path + norm_layer(dim), ChannelAttention(dim, groups=groups, qkv_bias=qkv_bias), drop_path ) self.conv2 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_ffn else None self.ffn = PreNorm( - norm_layer(dim), - Mlp(in_features=dim, hidden_features=int(dim*mlp_ratio), act_layer=act_layer), - drop_path + norm_layer(dim), Mlp(in_features=dim, hidden_features=int(dim * mlp_ratio), act_layer=act_layer), drop_path ) def forward(self, x, size): @@ -484,9 +465,9 @@ def window_partition(x, window_size: int): def window_reverse(windows, batch_size: int, window_size: int, H: int, W: int): - B = batch_size + B = batch_size # this will cause onnx conversion failed for dynamic axis, because treated as constant - # int(windows.shape[0] / (H * W / window_size / window_size)) + # int(windows.shape[0] / (H * W / window_size / window_size)) x = windows.view(B, H // window_size, W // window_size, window_size, window_size, -1) x = x.permute(0, 1, 3, 2, 4, 5).contiguous().view(B, H, W, -1) return x @@ -494,7 +475,6 @@ def window_reverse(windows, batch_size: int, window_size: int, H: int, W: int): class WindowAttention(nn.Module): def __init__(self, dim, num_heads, window_size, qkv_bias=True): - super().__init__() self.dim = dim self.window_size = window_size @@ -508,7 +488,6 @@ def __init__(self, dim, num_heads, window_size, qkv_bias=True): self.softmax = nn.Softmax(dim=-1) def forward(self, x, size): - H, W = size B, L, C = x.shape assert L == H * W, "input feature has wrong size" @@ -532,16 +511,14 @@ def forward(self, x, size): q, k, v = qkv[0], qkv[1], qkv[2] q = q * self.scale - attn = (q @ k.transpose(-2, -1)) + attn = q @ k.transpose(-2, -1) attn = self.softmax(attn) x = (attn @ v).transpose(1, 2).reshape(B_, N, C) x = self.proj(x) # merge windows - x = x.view( - -1, self.window_size, self.window_size, C - ) + x = x.view(-1, self.window_size, self.window_size, C) x = window_reverse(x, B, self.window_size, Hp, Wp) if pad_r > 0 or pad_b > 0: @@ -553,25 +530,30 @@ def forward(self, x, size): class SpatialBlock(nn.Module): - - def __init__(self, dim, num_heads, window_size, - mlp_ratio=4., qkv_bias=True, drop_path_rate=0., act_layer=nn.GELU, - norm_layer=nn.LayerNorm, conv_at_attn=True, conv_at_ffn=True): + def __init__( + self, + dim, + num_heads, + window_size, + mlp_ratio=4.0, + qkv_bias=True, + drop_path_rate=0.0, + act_layer=nn.GELU, + norm_layer=nn.LayerNorm, + conv_at_attn=True, + conv_at_ffn=True, + ): super().__init__() - drop_path = DropPath(drop_path_rate) if drop_path_rate > 0. else nn.Identity() + drop_path = DropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() self.conv1 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None self.window_attn = PreNorm( - norm_layer(dim), - WindowAttention(dim, num_heads, window_size, qkv_bias=qkv_bias), - drop_path + norm_layer(dim), WindowAttention(dim, num_heads, window_size, qkv_bias=qkv_bias), drop_path ) self.conv2 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_ffn else None self.ffn = PreNorm( - norm_layer(dim), - Mlp(in_features=dim, hidden_features=int(dim*mlp_ratio), act_layer=act_layer), - drop_path + norm_layer(dim), Mlp(in_features=dim, hidden_features=int(dim * mlp_ratio), act_layer=act_layer), drop_path ) def forward(self, x, size): @@ -586,7 +568,7 @@ def forward(self, x, size): class DaViT(nn.Module): - """ DaViT: Dual-Attention Transformer + """DaViT: Dual-Attention Transformer Args: in_chans (int): Number of input image channels. Default: 3. @@ -621,14 +603,14 @@ def __init__( num_heads=(3, 6, 12, 24), num_groups=(3, 6, 12, 24), window_size=7, - mlp_ratio=4., + mlp_ratio=4.0, qkv_bias=True, drop_path_rate=0.1, norm_layer=nn.LayerNorm, enable_checkpoint=False, conv_at_attn=True, conv_at_ffn=True, - ): + ): super().__init__() self.num_classes = num_classes @@ -640,7 +622,7 @@ def __init__( assert self.num_stages == len(self.num_heads) == len(self.num_groups) num_stages = len(embed_dims) - dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths)*2)] + dpr = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths) * 2)] depth_offset = 0 convs = [] @@ -653,41 +635,48 @@ def __init__( in_chans=in_chans if i == 0 else self.embed_dims[i - 1], embed_dim=self.embed_dims[i], norm_layer=norm_layer, - pre_norm=patch_prenorm[i] + pre_norm=patch_prenorm[i], ) convs.append(conv_embed) block = MySequential( *[ - MySequential(OrderedDict([ - ( - 'spatial_block', SpatialBlock( - embed_dims[i], - num_heads[i], - window_size, - drop_path_rate=dpr[depth_offset+j*2], - qkv_bias=qkv_bias, - mlp_ratio=mlp_ratio, - conv_at_attn=conv_at_attn, - conv_at_ffn=conv_at_ffn, - ) - ), - ( - 'channel_block', ChannelBlock( - embed_dims[i], - num_groups[i], - drop_path_rate=dpr[depth_offset+j*2+1], - qkv_bias=qkv_bias, - mlp_ratio=mlp_ratio, - conv_at_attn=conv_at_attn, - conv_at_ffn=conv_at_ffn, - ) + MySequential( + OrderedDict( + [ + ( + "spatial_block", + SpatialBlock( + embed_dims[i], + num_heads[i], + window_size, + drop_path_rate=dpr[depth_offset + j * 2], + qkv_bias=qkv_bias, + mlp_ratio=mlp_ratio, + conv_at_attn=conv_at_attn, + conv_at_ffn=conv_at_ffn, + ), + ), + ( + "channel_block", + ChannelBlock( + embed_dims[i], + num_groups[i], + drop_path_rate=dpr[depth_offset + j * 2 + 1], + qkv_bias=qkv_bias, + mlp_ratio=mlp_ratio, + conv_at_attn=conv_at_attn, + conv_at_ffn=conv_at_ffn, + ), + ), + ] ) - ])) for j in range(depths[i]) + ) + for j in range(depths[i]) ] ) blocks.append(block) - depth_offset += depths[i]*2 + depth_offset += depths[i] * 2 self.convs = nn.ModuleList(convs) self.blocks = nn.ModuleList(blocks) @@ -710,7 +699,7 @@ def _init_weights(self, m): elif isinstance(m, nn.Conv2d): nn.init.normal_(m.weight, std=0.02) for name, _ in m.named_parameters(): - if name in ['bias']: + if name in ["bias"]: nn.init.constant_(m.bias, 0) elif isinstance(m, nn.LayerNorm): nn.init.constant_(m.weight, 1.0) @@ -721,7 +710,7 @@ def _init_weights(self, m): def forward_features_unpool(self, x): """ - forward until avg pooling + forward until avg pooling Args: x (_type_): input image tensor """ @@ -749,7 +738,7 @@ def forward(self, x): x = self.forward_features(x) x = self.head(x) return x - + @classmethod def from_config(cls, config): return cls( @@ -766,12 +755,11 @@ def from_config(cls, config): ) - - if is_flash_attn_2_available(): from flash_attn import flash_attn_func, flash_attn_varlen_func from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input # noqa + # Copied from transformers.models.llama.modeling_llama._get_unpad_data def _get_unpad_data(attention_mask): seqlens_in_batch = attention_mask.sum(dim=-1, dtype=torch.int32) @@ -1516,7 +1504,6 @@ def forward( return outputs - class Florence2LanguagePreTrainedModel(PreTrainedModel): config_class = Florence2LanguageConfig base_model_prefix = "model" @@ -2320,6 +2307,7 @@ def _reorder_cache(past_key_values, beam_idx): ) return reordered_past + @dataclass class Florence2Seq2SeqLMOutput(ModelOutput): """ @@ -2495,6 +2483,7 @@ def _supports_sdpa(self): Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. """ + @add_start_docstrings( """The FLORENCE2 vision model without any head""", FLORENCE2_START_DOCSTRING, @@ -2502,16 +2491,16 @@ def _supports_sdpa(self): class Florence2VisionModel(Florence2PreTrainedModel): def __init__(self, config: Florence2VisionConfig): super().__init__(config) - assert config.model_type == 'davit', 'only DaViT is supported for now' + assert config.model_type == "davit", "only DaViT is supported for now" self.vision_tower = DaViT.from_config(config=config) self.post_init() - + def forward(self, pixel_values): if len(pixel_values.shape) == 4: x = self.vision_tower.forward_features_unpool(pixel_values) else: - raise ValueError(f'invalid image shape {pixel_values.shape}') + raise ValueError(f"invalid image shape {pixel_values.shape}") return x @@ -2522,40 +2511,36 @@ def forward(self, pixel_values): class Florence2VisionModelWithProjection(Florence2PreTrainedModel): def __init__(self, config: Florence2VisionConfig): super().__init__(config) - assert config.model_type == 'davit', 'only DaViT is supported for now' + assert config.model_type == "davit", "only DaViT is supported for now" self.vision_tower = DaViT.from_config(config=config) self._build_image_projection_layers(config) self.post_init() - + def _build_image_projection_layers(self, config): image_dim_out = config.dim_embed[-1] dim_projection = config.projection_dim - self.image_projection = nn.Parameter( - torch.empty(image_dim_out, dim_projection) - ) + self.image_projection = nn.Parameter(torch.empty(image_dim_out, dim_projection)) self.image_proj_norm = nn.LayerNorm(dim_projection) image_pos_embed_config = config.image_pos_embed - if image_pos_embed_config['type'] == 'learned_abs_2d': + if image_pos_embed_config["type"] == "learned_abs_2d": self.image_pos_embed = LearnedAbsolutePositionEmbedding2D( - embedding_dim=image_dim_out, - num_pos=image_pos_embed_config['max_pos_embeddings'] + embedding_dim=image_dim_out, num_pos=image_pos_embed_config["max_pos_embeddings"] ) else: - raise NotImplementedError('Not implemented yet') + raise NotImplementedError("Not implemented yet") self.image_feature_source = config.image_feature_source # temporal embedding visual_temporal_embedding_config = config.visual_temporal_embedding - if visual_temporal_embedding_config['type'] == 'COSINE': + if visual_temporal_embedding_config["type"] == "COSINE": self.visual_temporal_embed = PositionalEmbeddingCosine1D( - embed_dim=image_dim_out, - max_seq_len=visual_temporal_embedding_config['max_temporal_embeddings'] + embed_dim=image_dim_out, max_seq_len=visual_temporal_embedding_config["max_temporal_embeddings"] ) else: - raise NotImplementedError('Not implemented yet') + raise NotImplementedError("Not implemented yet") def forward(self, pixel_values): if len(pixel_values.shape) == 4: @@ -2563,17 +2548,17 @@ def forward(self, pixel_values): T = 1 x = self.vision_tower.forward_features_unpool(pixel_values) else: - raise ValueError(f'invalid image shape {pixel_values.shape}') - + raise ValueError(f"invalid image shape {pixel_values.shape}") + if self.image_pos_embed is not None: x = x.view(batch_size * T, -1, x.shape[-1]) num_tokens = x.shape[-2] - h, w = int(num_tokens ** 0.5), int(num_tokens ** 0.5) - assert h * w == num_tokens, 'only support square feature maps for now' + h, w = int(num_tokens**0.5), int(num_tokens**0.5) + assert h * w == num_tokens, "only support square feature maps for now" x = x.view(batch_size * T, h, w, x.shape[-1]) pos_embed = self.image_pos_embed(x) x = x + pos_embed - x = x.view(batch_size, T * h*w, x.shape[-1]) + x = x.view(batch_size, T * h * w, x.shape[-1]) if self.visual_temporal_embed is not None: visual_temporal_embed = self.visual_temporal_embed(x.view(batch_size, T, -1, x.shape[-1])[:, :, 0]) @@ -2582,18 +2567,18 @@ def forward(self, pixel_values): x_feat_dict = {} spatial_avg_pool_x = x.view(batch_size, T, -1, x.shape[-1]).mean(dim=2) - x_feat_dict['spatial_avg_pool'] = spatial_avg_pool_x + x_feat_dict["spatial_avg_pool"] = spatial_avg_pool_x temporal_avg_pool_x = x.view(batch_size, T, -1, x.shape[-1]).mean(dim=1) - x_feat_dict['temporal_avg_pool'] = temporal_avg_pool_x + x_feat_dict["temporal_avg_pool"] = temporal_avg_pool_x x = x.view(batch_size, T, -1, x.shape[-1])[:, -1] - x_feat_dict['last_frame'] = x + x_feat_dict["last_frame"] = x new_x = [] for _image_feature_source in self.image_feature_source: if _image_feature_source not in x_feat_dict: - raise ValueError('invalid image feature source: {}'.format(_image_feature_source)) + raise ValueError("invalid image feature source: {}".format(_image_feature_source)) new_x.append(x_feat_dict[_image_feature_source]) x = torch.cat(new_x, dim=1) @@ -2601,11 +2586,9 @@ def forward(self, pixel_values): x = x @ self.image_projection x = self.image_proj_norm(x) - return x - @add_start_docstrings( """The FLORENCE2 model which consists of a vision backbone and a language model.""", FLORENCE2_START_DOCSTRING, @@ -2613,10 +2596,10 @@ def forward(self, pixel_values): class Florence2ForConditionalGeneration(Florence2PreTrainedModel): def __init__(self, config: Florence2Config): super().__init__(config) - assert config.vision_config.model_type == 'davit', 'only DaViT is supported for now' + assert config.vision_config.model_type == "davit", "only DaViT is supported for now" del config.vision_config.model_type self.vision_tower = DaViT.from_config(config=config.vision_config) - # remove unused layers + # remove unused layers del self.vision_tower.head del self.vision_tower.norms @@ -2632,34 +2615,30 @@ def __init__(self, config: Florence2Config): self.pad_token_id = self.config.pad_token_id if self.config.pad_token_id is not None else -1 self.post_init() - + def _build_image_projection_layers(self, config): image_dim_out = config.vision_config.dim_embed[-1] dim_projection = config.vision_config.projection_dim - self.image_projection = nn.Parameter( - torch.empty(image_dim_out, dim_projection) - ) + self.image_projection = nn.Parameter(torch.empty(image_dim_out, dim_projection)) self.image_proj_norm = nn.LayerNorm(dim_projection) image_pos_embed_config = config.vision_config.image_pos_embed - if image_pos_embed_config['type'] == 'learned_abs_2d': + if image_pos_embed_config["type"] == "learned_abs_2d": self.image_pos_embed = LearnedAbsolutePositionEmbedding2D( - embedding_dim=image_dim_out, - num_pos=image_pos_embed_config['max_pos_embeddings'] + embedding_dim=image_dim_out, num_pos=image_pos_embed_config["max_pos_embeddings"] ) else: - raise NotImplementedError('Not implemented yet') + raise NotImplementedError("Not implemented yet") self.image_feature_source = config.vision_config.image_feature_source # temporal embedding visual_temporal_embedding_config = config.vision_config.visual_temporal_embedding - if visual_temporal_embedding_config['type'] == 'COSINE': + if visual_temporal_embedding_config["type"] == "COSINE": self.visual_temporal_embed = PositionalEmbeddingCosine1D( - embed_dim=image_dim_out, - max_seq_len=visual_temporal_embedding_config['max_temporal_embeddings'] + embed_dim=image_dim_out, max_seq_len=visual_temporal_embedding_config["max_temporal_embeddings"] ) else: - raise NotImplementedError('Not implemented yet') + raise NotImplementedError("Not implemented yet") def get_encoder(self): return self.language_model.get_encoder() @@ -2677,24 +2656,24 @@ def resize_token_embeddings(self, new_num_tokens: Optional[int] = None, pad_to_m self.config.vocab_size = model_embeds.num_embeddings self.vocab_size = model_embeds.num_embeddings return model_embeds - + def _encode_image(self, pixel_values): if len(pixel_values.shape) == 4: batch_size, C, H, W = pixel_values.shape T = 1 x = self.vision_tower.forward_features_unpool(pixel_values) else: - raise ValueError(f'invalid image shape {pixel_values.shape}') - + raise ValueError(f"invalid image shape {pixel_values.shape}") + if self.image_pos_embed is not None: x = x.view(batch_size * T, -1, x.shape[-1]) num_tokens = x.shape[-2] - h, w = int(num_tokens ** 0.5), int(num_tokens ** 0.5) - assert h * w == num_tokens, 'only support square feature maps for now' + h, w = int(num_tokens**0.5), int(num_tokens**0.5) + assert h * w == num_tokens, "only support square feature maps for now" x = x.view(batch_size * T, h, w, x.shape[-1]) pos_embed = self.image_pos_embed(x) x = x + pos_embed - x = x.view(batch_size, T * h*w, x.shape[-1]) + x = x.view(batch_size, T * h * w, x.shape[-1]) if self.visual_temporal_embed is not None: visual_temporal_embed = self.visual_temporal_embed(x.view(batch_size, T, -1, x.shape[-1])[:, :, 0]) @@ -2703,18 +2682,18 @@ def _encode_image(self, pixel_values): x_feat_dict = {} spatial_avg_pool_x = x.view(batch_size, T, -1, x.shape[-1]).mean(dim=2) - x_feat_dict['spatial_avg_pool'] = spatial_avg_pool_x + x_feat_dict["spatial_avg_pool"] = spatial_avg_pool_x temporal_avg_pool_x = x.view(batch_size, T, -1, x.shape[-1]).mean(dim=1) - x_feat_dict['temporal_avg_pool'] = temporal_avg_pool_x + x_feat_dict["temporal_avg_pool"] = temporal_avg_pool_x x = x.view(batch_size, T, -1, x.shape[-1])[:, -1] - x_feat_dict['last_frame'] = x + x_feat_dict["last_frame"] = x new_x = [] for _image_feature_source in self.image_feature_source: if _image_feature_source not in x_feat_dict: - raise ValueError('invalid image feature source: {}'.format(_image_feature_source)) + raise ValueError("invalid image feature source: {}".format(_image_feature_source)) new_x.append(x_feat_dict[_image_feature_source]) x = torch.cat(new_x, dim=1) @@ -2722,11 +2701,9 @@ def _encode_image(self, pixel_values): x = x @ self.image_projection x = self.image_proj_norm(x) - return x + return x - def _merge_input_ids_with_image_features( - self, image_features, inputs_embeds - ): + def _merge_input_ids_with_image_features(self, image_features, inputs_embeds): batch_size, image_token_length = image_features.size()[:-1] device = image_features.device image_attention_mask = torch.ones(batch_size, image_token_length, device=device) @@ -2748,7 +2725,6 @@ def _merge_input_ids_with_image_features( return inputs_embeds, attention_mask - @add_start_docstrings_to_model_forward(FLORENCE2_INPUTS_DOCSTRING) @replace_return_docstrings(output_type=Florence2Seq2SeqLMOutput, config_class=_CONFIG_FOR_DOC) def forward( @@ -2816,7 +2792,9 @@ def forward( if pixel_values is not None: # (batch_size, num_image_tokens, hidden_size) image_features = self._encode_image(pixel_values) - inputs_embeds, attention_mask = self._merge_input_ids_with_image_features(image_features, inputs_embeds) + inputs_embeds, attention_mask = self._merge_input_ids_with_image_features( + image_features, inputs_embeds + ) attention_mask = attention_mask.to(inputs_embeds.dtype) outputs = self.language_model( @@ -2854,17 +2832,10 @@ def forward( encoder_last_hidden_state=outputs.encoder_last_hidden_state, encoder_hidden_states=outputs.encoder_hidden_states, encoder_attentions=outputs.encoder_attentions, - image_hidden_states=image_features + image_hidden_states=image_features, ) - def generate( - self, - input_ids, - inputs_embeds=None, - pixel_values=None, - **kwargs - ): - + def generate(self, input_ids, inputs_embeds=None, pixel_values=None, **kwargs): if inputs_embeds is None: # 1. Extra the input embeddings if input_ids is not None: @@ -2872,13 +2843,11 @@ def generate( # 2. Merge text and images if pixel_values is not None: image_features = self._encode_image(pixel_values) - inputs_embeds, attention_mask = self._merge_input_ids_with_image_features(image_features, inputs_embeds) - - return self.language_model.generate( - input_ids=None, - inputs_embeds=inputs_embeds, - **kwargs - ) + inputs_embeds, attention_mask = self._merge_input_ids_with_image_features( + image_features, inputs_embeds + ) + + return self.language_model.generate(input_ids=None, inputs_embeds=inputs_embeds, **kwargs) def prepare_inputs_for_generation( self, @@ -2906,7 +2875,7 @@ def prepare_inputs_for_generation( remove_prefix_length = decoder_input_ids.shape[1] - 1 decoder_input_ids = decoder_input_ids[:, remove_prefix_length:] - + return { "input_ids": None, # encoder_outputs is defined. input_ids not needed "encoder_outputs": encoder_outputs, @@ -2920,9 +2889,9 @@ def prepare_inputs_for_generation( "cross_attn_head_mask": cross_attn_head_mask, "use_cache": use_cache, # change this to avoid caching (presumably for debugging) } - + def prepare_decoder_input_ids_from_labels(self, labels: torch.Tensor): return self.language_model.shift_tokens_right(labels) def _reorder_cache(self, *args, **kwargs): - return self.language_model._reorder_cache(*args, **kwargs) \ No newline at end of file + return self.language_model._reorder_cache(*args, **kwargs) diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py index b97945c47672..b6531de62945 100644 --- a/src/transformers/models/florence/processing_florence2.py +++ b/src/transformers/models/florence/processing_florence2.py @@ -16,20 +16,19 @@ Processor class for Florence-2. """ -import re import logging +import re from typing import List, Optional, Union -import numpy as np +import numpy as np import torch +from ...feature_extraction_utils import BatchFeature +from ...image_utils import ChannelDimension, ImageInput, is_valid_image from ...models.bart.tokenization_bart import BartTokenizer from ...models.bart.tokenization_bart_fast import BartTokenizerFast from ...models.t5.tokenization_t5 import T5Tokenizer from ...models.t5.tokenization_t5_fast import T5TokenizerFast - -from ...feature_extraction_utils import BatchFeature -from ...image_utils import ImageInput, is_valid_image, ChannelDimension from ...processing_utils import ProcessorMixin from ...tokenization_utils_base import ( PaddingStrategy, @@ -42,10 +41,12 @@ logger = logging.getLogger(__name__) + # Copied from transformers.models.idefics2.processing_idefics2.is_url def is_url(val) -> bool: return isinstance(val, str) and val.startswith("http") + # Copied from transformers.models.idefics2.processing_idefics2.is_image_or_image_url def is_image_or_image_url(elem): return is_url(elem) or is_valid_image(elem) @@ -88,58 +89,77 @@ def __init__( self.image_seq_length = image_processor.image_seq_length tokens_to_add = { - 'additional_special_tokens': \ - tokenizer.additional_special_tokens + \ - ['', '', '', ''] + \ - [f'' for x in range(1000)] + \ - ['', '', '', '','', '', '', '', '', '', '', '', '', '', '', '', '', '', '', ''] - } + "additional_special_tokens": tokenizer.additional_special_tokens + + ["", "", "", ""] + + [f"" for x in range(1000)] + + [ + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + "", + ] + } tokenizer.add_special_tokens(tokens_to_add) self.tasks_answer_post_processing_type = { - '': 'pure_text', - '': 'ocr', - '': 'pure_text', - '': 'pure_text', - '': 'pure_text', - '': 'description_with_bboxes', - '': 'description_with_bboxes', - '': "phrase_grounding", - '': 'polygons', - '': 'polygons', - '': 'description_with_bboxes_or_polygons', - '': 'pure_text', - '': 'pure_text', - '': 'pure_text', - '': 'bboxes' + "": "pure_text", + "": "ocr", + "": "pure_text", + "": "pure_text", + "": "pure_text", + "": "description_with_bboxes", + "": "description_with_bboxes", + "": "phrase_grounding", + "": "polygons", + "": "polygons", + "": "description_with_bboxes_or_polygons", + "": "pure_text", + "": "pure_text", + "": "pure_text", + "": "bboxes", } self.task_prompts_without_inputs = { - '': 'What is the text in the image?', - '': 'What is the text in the image, with regions?', - '': 'What does the image describe?', - '': 'Describe in detail what is shown in the image.', - '': 'Describe with a paragraph what is shown in the image.', - '': 'Locate the objects with category name in the image.', - '': 'Locate the objects in the image, with their descriptions.', - '': 'Locate the region proposals in the image.' + "": "What is the text in the image?", + "": "What is the text in the image, with regions?", + "": "What does the image describe?", + "": "Describe in detail what is shown in the image.", + "": "Describe with a paragraph what is shown in the image.", + "": "Locate the objects with category name in the image.", + "": "Locate the objects in the image, with their descriptions.", + "": "Locate the region proposals in the image.", } self.task_prompts_with_input = { - '': "Locate the phrases in the caption: {input}", - '': 'Locate {input} in the image with mask', - '': 'What is the polygon mask of region {input}', - '': 'Locate {input} in the image.', - '': 'What is the region {input}?', - '': 'What does the region {input} describe?', - '': 'What text is in the region {input}?', + "": "Locate the phrases in the caption: {input}", + "": "Locate {input} in the image with mask", + "": "What is the polygon mask of region {input}", + "": "Locate {input} in the image.", + "": "What is the region {input}?", + "": "What does the region {input} describe?", + "": "What text is in the region {input}?", } self.post_processor = Florence2PostProcesser(tokenizer=tokenizer) - super().__init__(image_processor, tokenizer) - + def _construct_prompts(self, text): # replace the task tokens with the task prompts if task token is in the text prompts = [] @@ -150,10 +170,10 @@ def _construct_prompts(self, text): assert _text == task_token, f"Task token {task_token} should be the only token in the text." _text = task_prompt break - # 2. task prompts with additional inputs + # 2. task prompts with additional inputs for task_token, task_prompt in self.task_prompts_with_input.items(): if task_token in _text: - _text = task_prompt.format(input=_text.replace(task_token, '')) + _text = task_prompt.format(input=_text.replace(task_token, "")) break prompts.append(_text) return prompts @@ -237,9 +257,7 @@ def __call__( if images is None: raise ValueError("`images` are expected as arguments to a `Florence2Processor` instance.") if text is None: - logger.warning_once( - "You are using Florence-2 without a text prompt." - ) + logger.warning_once("You are using Florence-2 without a text prompt.") text = "" if isinstance(text, List) and isinstance(images, List): @@ -319,65 +337,70 @@ def post_process_generation(self, text, task, image_size): image_size (`Tuple[int, int]`): The size of the image. height x width. """ - task_answer_post_processing_type = self.tasks_answer_post_processing_type.get(task, 'pure_text') + task_answer_post_processing_type = self.tasks_answer_post_processing_type.get(task, "pure_text") task_answer = self.post_processor( text=text, image_size=image_size, parse_tasks=task_answer_post_processing_type, )[task_answer_post_processing_type] - if task_answer_post_processing_type == 'pure_text': + if task_answer_post_processing_type == "pure_text": final_answer = task_answer # remove the special tokens - final_answer = final_answer.replace('', '').replace('', '') - elif task_answer_post_processing_type in ['od', 'description_with_bboxes', 'bboxes']: + final_answer = final_answer.replace("", "").replace("", "") + elif task_answer_post_processing_type in ["od", "description_with_bboxes", "bboxes"]: od_instances = task_answer - bboxes_od = [_od_instance['bbox'] for _od_instance in od_instances] - labels_od = [str(_od_instance['cat_name']) for _od_instance in od_instances] - final_answer = {'bboxes': bboxes_od, 'labels': labels_od} - elif task_answer_post_processing_type in ['ocr']: - bboxes = [_od_instance['quad_box'] for _od_instance in task_answer] - labels = [str(_od_instance['text']) for _od_instance in task_answer] - final_answer = {'quad_boxes': bboxes, 'labels': labels} - elif task_answer_post_processing_type in ['phrase_grounding']: + bboxes_od = [_od_instance["bbox"] for _od_instance in od_instances] + labels_od = [str(_od_instance["cat_name"]) for _od_instance in od_instances] + final_answer = {"bboxes": bboxes_od, "labels": labels_od} + elif task_answer_post_processing_type in ["ocr"]: + bboxes = [_od_instance["quad_box"] for _od_instance in task_answer] + labels = [str(_od_instance["text"]) for _od_instance in task_answer] + final_answer = {"quad_boxes": bboxes, "labels": labels} + elif task_answer_post_processing_type in ["phrase_grounding"]: bboxes = [] labels = [] for _grounded_phrase in task_answer: - for _bbox in _grounded_phrase['bbox']: + for _bbox in _grounded_phrase["bbox"]: bboxes.append(_bbox) - labels.append(_grounded_phrase['cat_name']) - final_answer = {'bboxes': bboxes, 'labels': labels} - elif task_answer_post_processing_type in ['description_with_polygons', 'polygons']: + labels.append(_grounded_phrase["cat_name"]) + final_answer = {"bboxes": bboxes, "labels": labels} + elif task_answer_post_processing_type in ["description_with_polygons", "polygons"]: labels = [] polygons = [] for result in task_answer: - label = result['cat_name'] - _polygons = result['polygons'] + label = result["cat_name"] + _polygons = result["polygons"] labels.append(label) polygons.append(_polygons) - final_answer = {'polygons': polygons, 'labels': labels} - elif task_answer_post_processing_type in ['description_with_bboxes_or_polygons']: + final_answer = {"polygons": polygons, "labels": labels} + elif task_answer_post_processing_type in ["description_with_bboxes_or_polygons"]: bboxes = [] bboxes_labels = [] polygons = [] polygons_labels = [] for result in task_answer: - label = result['cat_name'] - if 'polygons' in result: - _polygons = result['polygons'] + label = result["cat_name"] + if "polygons" in result: + _polygons = result["polygons"] polygons.append(_polygons) polygons_labels.append(label) else: - _bbox = result['bbox'] + _bbox = result["bbox"] bboxes.append(_bbox) bboxes_labels.append(label) - final_answer = {'bboxes': bboxes, 'bboxes_labels': bboxes_labels, 'polygons': polygons, 'polygons_labels': polygons_labels} + final_answer = { + "bboxes": bboxes, + "bboxes_labels": bboxes_labels, + "polygons": polygons, + "polygons_labels": polygons_labels, + } else: - raise ValueError('Unknown task answer post processing type: {}'.format(task_answer_post_processing_type)) + raise ValueError("Unknown task answer post processing type: {}".format(task_answer_post_processing_type)) + + final_answer = {task: final_answer} + return final_answer - final_answer = { - task: final_answer} - return final_answer class BoxQuantizer(object): def __init__(self, mode, bins): @@ -386,57 +409,48 @@ def __init__(self, mode, bins): def quantize(self, boxes: torch.Tensor, size): bins_w, bins_h = self.bins # Quantization bins. - size_w, size_h = size # Original image size. + size_w, size_h = size # Original image size. size_per_bin_w = size_w / bins_w size_per_bin_h = size_h / bins_h xmin, ymin, xmax, ymax = boxes.split(1, dim=-1) # Shape: 4 * [N, 1]. - if self.mode == 'floor': - quantized_xmin = ( - xmin / size_per_bin_w).floor().clamp(0, bins_w - 1) - quantized_ymin = ( - ymin / size_per_bin_h).floor().clamp(0, bins_h - 1) - quantized_xmax = ( - xmax / size_per_bin_w).floor().clamp(0, bins_w - 1) - quantized_ymax = ( - ymax / size_per_bin_h).floor().clamp(0, bins_h - 1) - - elif self.mode == 'round': + if self.mode == "floor": + quantized_xmin = (xmin / size_per_bin_w).floor().clamp(0, bins_w - 1) + quantized_ymin = (ymin / size_per_bin_h).floor().clamp(0, bins_h - 1) + quantized_xmax = (xmax / size_per_bin_w).floor().clamp(0, bins_w - 1) + quantized_ymax = (ymax / size_per_bin_h).floor().clamp(0, bins_h - 1) + + elif self.mode == "round": raise NotImplementedError() else: - raise ValueError('Incorrect quantization type.') + raise ValueError("Incorrect quantization type.") - quantized_boxes = torch.cat( - (quantized_xmin, quantized_ymin, quantized_xmax, quantized_ymax), dim=-1 - ).int() + quantized_boxes = torch.cat((quantized_xmin, quantized_ymin, quantized_xmax, quantized_ymax), dim=-1).int() return quantized_boxes def dequantize(self, boxes: torch.Tensor, size): bins_w, bins_h = self.bins # Quantization bins. - size_w, size_h = size # Original image size. + size_w, size_h = size # Original image size. size_per_bin_w = size_w / bins_w size_per_bin_h = size_h / bins_h xmin, ymin, xmax, ymax = boxes.split(1, dim=-1) # Shape: 4 * [N, 1]. - if self.mode == 'floor': + if self.mode == "floor": # Add 0.5 to use the center position of the bin as the coordinate. dequantized_xmin = (xmin + 0.5) * size_per_bin_w dequantized_ymin = (ymin + 0.5) * size_per_bin_h dequantized_xmax = (xmax + 0.5) * size_per_bin_w dequantized_ymax = (ymax + 0.5) * size_per_bin_h - elif self.mode == 'round': + elif self.mode == "round": raise NotImplementedError() else: - raise ValueError('Incorrect quantization type.') + raise ValueError("Incorrect quantization type.") - dequantized_boxes = torch.cat( - (dequantized_xmin, dequantized_ymin, - dequantized_xmax, dequantized_ymax), dim=-1 - ) + dequantized_boxes = torch.cat((dequantized_xmin, dequantized_ymin, dequantized_xmax, dequantized_ymax), dim=-1) return dequantized_boxes @@ -452,57 +466,53 @@ def __init__(self, mode, bins): def quantize(self, coordinates: torch.Tensor, size): bins_w, bins_h = self.bins # Quantization bins. - size_w, size_h = size # Original image size. + size_w, size_h = size # Original image size. size_per_bin_w = size_w / bins_w size_per_bin_h = size_h / bins_h - assert coordinates.shape[-1] == 2, 'coordinates should be shape (N, 2)' + assert coordinates.shape[-1] == 2, "coordinates should be shape (N, 2)" x, y = coordinates.split(1, dim=-1) # Shape: 4 * [N, 1]. - if self.mode == 'floor': + if self.mode == "floor": quantized_x = (x / size_per_bin_w).floor().clamp(0, bins_w - 1) quantized_y = (y / size_per_bin_h).floor().clamp(0, bins_h - 1) - elif self.mode == 'round': + elif self.mode == "round": raise NotImplementedError() else: - raise ValueError('Incorrect quantization type.') + raise ValueError("Incorrect quantization type.") - quantized_coordinates = torch.cat( - (quantized_x, quantized_y), dim=-1 - ).int() + quantized_coordinates = torch.cat((quantized_x, quantized_y), dim=-1).int() return quantized_coordinates def dequantize(self, coordinates: torch.Tensor, size): bins_w, bins_h = self.bins # Quantization bins. - size_w, size_h = size # Original image size. + size_w, size_h = size # Original image size. size_per_bin_w = size_w / bins_w size_per_bin_h = size_h / bins_h - assert coordinates.shape[-1] == 2, 'coordinates should be shape (N, 2)' + assert coordinates.shape[-1] == 2, "coordinates should be shape (N, 2)" x, y = coordinates.split(1, dim=-1) # Shape: 4 * [N, 1]. - if self.mode == 'floor': + if self.mode == "floor": # Add 0.5 to use the center position of the bin as the coordinate. dequantized_x = (x + 0.5) * size_per_bin_w dequantized_y = (y + 0.5) * size_per_bin_h - elif self.mode == 'round': + elif self.mode == "round": raise NotImplementedError() else: - raise ValueError('Incorrect quantization type.') + raise ValueError("Incorrect quantization type.") - dequantized_coordinates = torch.cat( - (dequantized_x, dequantized_y), dim=-1 - ) + dequantized_coordinates = torch.cat((dequantized_x, dequantized_y), dim=-1) return dequantized_coordinates class Florence2PostProcesser(object): - """ - Florence-2 post process for converting text prediction to various tasks results. + r""" + Florence-2 post process for converting text prediction to various tasks results. Args: config: A dict of configs. @@ -527,22 +537,20 @@ class Florence2PostProcesser(object): Returns: parsed_dict (dict): A dict of parsed results. """ - def __init__( - self, - tokenizer=None - ): + + def __init__(self, tokenizer=None): parse_tasks = [] parse_task_configs = {} config = self._create_default_config() - for task in config['PARSE_TASKS']: - parse_tasks.append(task['TASK_NAME']) - parse_task_configs[task['TASK_NAME']] = task + for task in config["PARSE_TASKS"]: + parse_tasks.append(task["TASK_NAME"]) + parse_task_configs[task["TASK_NAME"]] = task self.config = config self.parse_tasks = parse_tasks self.parse_tasks_configs = parse_task_configs - self.tokenizer = tokenizer + self.tokenizer = tokenizer if self.tokenizer is not None: self.all_special_tokens = set(self.tokenizer.all_special_tokens) @@ -552,105 +560,188 @@ def __init__( def _create_black_list_of_phrase_grounding(self): black_list = {} - if 'phrase_grounding' in self.parse_tasks and self.parse_tasks_configs['phrase_grounding']['FILTER_BY_BLACK_LIST']: - black_list = set( - ['it', 'I', 'me', 'mine', - 'you', 'your', 'yours', - 'he', 'him', 'his', - 'she', 'her', 'hers', - 'they', 'them', 'their', 'theirs', - 'one', 'oneself', - 'we', 'us', 'our', 'ours', - 'you', 'your', 'yours', - 'they', 'them', 'their', 'theirs', - 'mine', 'yours', 'his', 'hers', 'its', - 'ours', 'yours', 'theirs', - 'myself', 'yourself', 'himself', 'herself', 'itself', - 'ourselves', 'yourselves', 'themselves', - 'this', 'that', - 'these', 'those', - 'who', 'whom', 'whose', 'which', 'what', - 'who', 'whom', 'whose', 'which', 'that', - 'all', 'another', 'any', 'anybody', 'anyone', 'anything', - 'each', 'everybody', 'everyone', 'everything', - 'few', 'many', 'nobody', 'none', 'one', 'several', - 'some', 'somebody', 'someone', 'something', - 'each other', 'one another', - 'myself', 'yourself', 'himself', 'herself', 'itself', - 'ourselves', 'yourselves', 'themselves', - 'the image', 'image', 'images', 'the', 'a', 'an', 'a group', - 'other objects', 'lots', 'a set', - ] + if ( + "phrase_grounding" in self.parse_tasks + and self.parse_tasks_configs["phrase_grounding"]["FILTER_BY_BLACK_LIST"] + ): + black_list = set( + [ + "it", + "I", + "me", + "mine", + "you", + "your", + "yours", + "he", + "him", + "his", + "she", + "her", + "hers", + "they", + "them", + "their", + "theirs", + "one", + "oneself", + "we", + "us", + "our", + "ours", + "you", + "your", + "yours", + "they", + "them", + "their", + "theirs", + "mine", + "yours", + "his", + "hers", + "its", + "ours", + "yours", + "theirs", + "myself", + "yourself", + "himself", + "herself", + "itself", + "ourselves", + "yourselves", + "themselves", + "this", + "that", + "these", + "those", + "who", + "whom", + "whose", + "which", + "what", + "who", + "whom", + "whose", + "which", + "that", + "all", + "another", + "any", + "anybody", + "anyone", + "anything", + "each", + "everybody", + "everyone", + "everything", + "few", + "many", + "nobody", + "none", + "one", + "several", + "some", + "somebody", + "someone", + "something", + "each other", + "one another", + "myself", + "yourself", + "himself", + "herself", + "itself", + "ourselves", + "yourselves", + "themselves", + "the image", + "image", + "images", + "the", + "a", + "an", + "a group", + "other objects", + "lots", + "a set", + ] ) return black_list - + def _create_default_config(self): config = { - 'NUM_BBOX_HEIGHT_BINS': 1000, - 'NUM_BBOX_WIDTH_BINS': 1000, - 'BOX_QUANTIZATION_MODE': 'floor', - 'COORDINATES_HEIGHT_BINS': 1000, - 'COORDINATES_WIDTH_BINS': 1000, - 'COORDINATES_QUANTIZATION_MODE': 'floor', - 'PARSE_TASKS': [ - { - 'TASK_NAME': 'od', - 'PATTERN': r'([a-zA-Z0-9 ]+)' - }, + "NUM_BBOX_HEIGHT_BINS": 1000, + "NUM_BBOX_WIDTH_BINS": 1000, + "BOX_QUANTIZATION_MODE": "floor", + "COORDINATES_HEIGHT_BINS": 1000, + "COORDINATES_WIDTH_BINS": 1000, + "COORDINATES_QUANTIZATION_MODE": "floor", + "PARSE_TASKS": [ + {"TASK_NAME": "od", "PATTERN": r"([a-zA-Z0-9 ]+)"}, { - 'TASK_NAME': 'ocr', - 'PATTERN': r'(.+?)', - 'AREA_THRESHOLD': 0.01 + "TASK_NAME": "ocr", + "PATTERN": r"(.+?)", + "AREA_THRESHOLD": 0.01, }, + {"TASK_NAME": "phrase_grounding", "FILTER_BY_BLACK_LIST": True}, { - 'TASK_NAME': 'phrase_grounding', - 'FILTER_BY_BLACK_LIST': True + "TASK_NAME": "pure_text", }, { - 'TASK_NAME': 'pure_text', + "TASK_NAME": "description_with_bboxes", }, { - 'TASK_NAME': 'description_with_bboxes', + "TASK_NAME": "description_with_polygons", }, { - 'TASK_NAME': 'description_with_polygons', + "TASK_NAME": "polygons", }, { - 'TASK_NAME': 'polygons', + "TASK_NAME": "bboxes", }, { - 'TASK_NAME': 'bboxes', + "TASK_NAME": "description_with_bboxes_or_polygons", }, - { - 'TASK_NAME': 'description_with_bboxes_or_polygons', - } - ] + ], } return config def init_quantizers(self): # we have box_quantizer (od, grounding) and coordinates_quantizer (ocr, referring_segmentation) - num_bbox_height_bins = self.config.get('NUM_BBOX_HEIGHT_BINS', 1000) - num_bbox_width_bins = self.config.get('NUM_BBOX_WIDTH_BINS', 1000) - box_quantization_mode = self.config.get('BOX_QUANTIZATION_MODE', 'floor') + num_bbox_height_bins = self.config.get("NUM_BBOX_HEIGHT_BINS", 1000) + num_bbox_width_bins = self.config.get("NUM_BBOX_WIDTH_BINS", 1000) + box_quantization_mode = self.config.get("BOX_QUANTIZATION_MODE", "floor") self.box_quantizer = BoxQuantizer( box_quantization_mode, (num_bbox_width_bins, num_bbox_height_bins), ) - - num_bbox_height_bins = self.config['COORDINATES_HEIGHT_BINS'] if 'COORDINATES_HEIGHT_BINS' in self.config else self.config.get('NUM_BBOX_HEIGHT_BINS', 1000) - num_bbox_width_bins = self.config['COORDINATES_WIDTH_BINS'] if 'COORDINATES_WIDTH_BINS' in self.config else self.config.get('NUM_BBOX_WIDTH_BINS', 1000) - box_quantization_mode = self.config.get('COORDINATES_QUANTIZATION_MODE') if 'COORDINATES_QUANTIZATION_MODE' in self.config else self.config.get('BOX_QUANTIZATION_MODE', 'floor') + + num_bbox_height_bins = ( + self.config["COORDINATES_HEIGHT_BINS"] + if "COORDINATES_HEIGHT_BINS" in self.config + else self.config.get("NUM_BBOX_HEIGHT_BINS", 1000) + ) + num_bbox_width_bins = ( + self.config["COORDINATES_WIDTH_BINS"] + if "COORDINATES_WIDTH_BINS" in self.config + else self.config.get("NUM_BBOX_WIDTH_BINS", 1000) + ) + box_quantization_mode = ( + self.config.get("COORDINATES_QUANTIZATION_MODE") + if "COORDINATES_QUANTIZATION_MODE" in self.config + else self.config.get("BOX_QUANTIZATION_MODE", "floor") + ) self.coordinates_quantizer = CoordinatesQuantizer( box_quantization_mode, (num_bbox_width_bins, num_bbox_height_bins), ) def decode_with_spans(self, tokenizer, token_ids): - filtered_tokens = tokenizer.convert_ids_to_tokens( - token_ids, skip_special_tokens=False) + filtered_tokens = tokenizer.convert_ids_to_tokens(token_ids, skip_special_tokens=False) assert len(filtered_tokens) == len(token_ids) # To avoid mixing byte-level and unicode for byte-level BPT @@ -666,12 +757,12 @@ def decode_with_spans(self, tokenizer, token_ids): elif isinstance(tokenizer, (T5Tokenizer, T5TokenizerFast)): # Ref: https://github.com/google/sentencepiece#whitespace-is-treated-as-a-basic-symbol # Note: Do not strip sub_text as it may have functional whitespace - sub_text = token.replace('▁', ' ') + sub_text = token.replace("▁", " ") else: - raise ValueError(f'type {type(tokenizer)} not supported') + raise ValueError(f"type {type(tokenizer)} not supported") sub_texts.append(sub_text) - text = '' + text = "" spans = [] for sub_text in sub_texts: span = (len(text), len(text) + len(sub_text)) # [start index, end index). @@ -679,7 +770,7 @@ def decode_with_spans(self, tokenizer, token_ids): spans.append(span) # Text format: - # 1. T5Tokenizer/T5TokenizerFast: + # 1. T5Tokenizer/T5TokenizerFast: # " transplanting dog cat" # Equivalent to t5_tokenizer.decode(input_ids, skip_special_tokens=False, clean_up_tokenization_spaces=False, spaces_between_special_tokens=False) # 2. BartTokenizer (need to double check): @@ -687,13 +778,7 @@ def decode_with_spans(self, tokenizer, token_ids): # Equivalent to bart_tokenizer.decode(input_ids, skip_special_tokens=False, clean_up_tokenization_spaces=False, spaces_between_special_tokens=False) return text, spans - def parse_od_from_text_and_spans( - self, - text, - pattern, - image_size, - phrase_centric=False - ): + def parse_od_from_text_and_spans(self, text, pattern, image_size, phrase_centric=False): parsed = list(re.finditer(pattern, text)) instances = [] @@ -705,28 +790,26 @@ def parse_od_from_text_and_spans( bbox_bins = [int(parsed[i].group(j)) for j in range(2, 6)] else: bbox_bins = [int(parsed[i].group(j)) for j in range(1, 5)] - instance['bbox'] = self.box_quantizer.dequantize( - boxes=torch.tensor(bbox_bins), - size=image_size - ).tolist() + instance["bbox"] = self.box_quantizer.dequantize(boxes=torch.tensor(bbox_bins), size=image_size).tolist() if phrase_centric: - instance['cat_name'] = parsed[i].group(1).lower().strip() + instance["cat_name"] = parsed[i].group(1).lower().strip() else: - instance['cat_name'] = parsed[i].group(5).lower().strip() + instance["cat_name"] = parsed[i].group(5).lower().strip() instances.append(instance) return instances - def parse_ocr_from_text_and_spans(self, - text, - pattern, - image_size, - area_threshold=-1.0, - ): + def parse_ocr_from_text_and_spans( + self, + text, + pattern, + image_size, + area_threshold=-1.0, + ): bboxes = [] labels = [] - text = text.replace('', '') + text = text.replace("", "") # ocr with regions parsed = re.findall(pattern, text) instances = [] @@ -736,59 +819,64 @@ def parse_ocr_from_text_and_spans(self, ocr_content = ocr_line[0] quad_box = ocr_line[1:] quad_box = [int(i) for i in quad_box] - quad_box = self.coordinates_quantizer.dequantize( - torch.tensor(np.array(quad_box).reshape(-1, 2)), - size=image_size - ).reshape(-1).tolist() + quad_box = ( + self.coordinates_quantizer.dequantize(torch.tensor(np.array(quad_box).reshape(-1, 2)), size=image_size) + .reshape(-1) + .tolist() + ) if area_threshold > 0: x_coords = [i for i in quad_box[0::2]] y_coords = [i for i in quad_box[1::2]] # apply the Shoelace formula - area = 0.5 * abs(sum(x_coords[i] * y_coords[i + 1] - x_coords[i + 1] * y_coords[i] for i in range(4 - 1))) + area = 0.5 * abs( + sum(x_coords[i] * y_coords[i + 1] - x_coords[i + 1] * y_coords[i] for i in range(4 - 1)) + ) if area < (image_width * image_height) * area_threshold: continue bboxes.append(quad_box) labels.append(ocr_content) - instances.append({ - 'quad_box': quad_box, - 'text': ocr_content, - }) + instances.append( + { + "quad_box": quad_box, + "text": ocr_content, + } + ) return instances def parse_phrase_grounding_from_text_and_spans(self, text, pattern, image_size): # ignore and cur_span = 0 - if text.startswith(''): + if text.startswith(""): cur_span += 3 - text = text.replace('', '') - text = text.replace('', '') - text = text.replace('', '') + text = text.replace("", "") + text = text.replace("", "") + text = text.replace("", "") pattern = r"([^<]+(?:){4,})" phrases = re.findall(pattern, text) - + # pattern should be text pattern and od pattern - pattern = r'^\s*(.*?)(?=||||||' + pattern = r"^\s*(.*?)(?=||||||" instances = [] for pharse_text in phrases: - phrase_text_strip = pharse_text.replace('', '', 1) - phrase_text_strip = pharse_text.replace('', '', 1) + phrase_text_strip = pharse_text.replace("", "", 1) + phrase_text_strip = pharse_text.replace("", "", 1) - if phrase_text_strip == '': + if phrase_text_strip == "": cur_span += len(pharse_text) continue # Prepare instance. instance = {} - # parse phrase, get string + # parse phrase, get string phrase = re.search(pattern, phrase_text_strip) if phrase is None: cur_span += len(pharse_text) @@ -808,16 +896,13 @@ def parse_phrase_grounding_from_text_and_spans(self, text, pattern, image_size): cur_span += len(pharse_text) continue - # a list of list + # a list of list bbox_bins = [[int(_bboxes_parsed.group(j)) for j in range(1, 5)] for _bboxes_parsed in bboxes_parsed] - instance['bbox'] = self.box_quantizer.dequantize( - boxes=torch.tensor(bbox_bins), - size=image_size - ).tolist() + instance["bbox"] = self.box_quantizer.dequantize(boxes=torch.tensor(bbox_bins), size=image_size).tolist() # exclude non-ascii characters - phrase = phrase.encode('ascii',errors='ignore').decode('ascii') - instance['cat_name'] = phrase + phrase = phrase.encode("ascii", errors="ignore").decode("ascii") + instance["cat_name"] = phrase instances.append(instance) @@ -827,29 +912,29 @@ def parse_description_with_bboxes_from_text_and_spans(self, text, pattern, image # temporary parse solution, split by '.' # ignore and - text = text.replace('', '') - text = text.replace('', '') - text = text.replace('', '') + text = text.replace("", "") + text = text.replace("", "") + text = text.replace("", "") if allow_empty_phrase: - pattern = rf"(?:(?:){{4,}})" + pattern = r"(?:(?:){4,})" else: pattern = r"([^<]+(?:){4,})" phrases = re.findall(pattern, text) - + # pattern should be text pattern and od pattern - pattern = r'^\s*(.*?)(?=||||||' + pattern = r"^\s*(.*?)(?=||||||" instances = [] for pharse_text in phrases: - phrase_text_strip = pharse_text.replace('', '', 1) - phrase_text_strip = pharse_text.replace('', '', 1) + phrase_text_strip = pharse_text.replace("", "", 1) + phrase_text_strip = pharse_text.replace("", "", 1) - if phrase_text_strip == '' and not allow_empty_phrase: + if phrase_text_strip == "" and not allow_empty_phrase: continue - # parse phrase, get string + # parse phrase, get string phrase = re.search(pattern, phrase_text_strip) if phrase is None: continue @@ -863,70 +948,68 @@ def parse_description_with_bboxes_from_text_and_spans(self, text, pattern, image if len(bboxes_parsed) == 0: continue - # a list of list + # a list of list bbox_bins = [[int(_bboxes_parsed.group(j)) for j in range(1, 5)] for _bboxes_parsed in bboxes_parsed] - bboxes = self.box_quantizer.dequantize( - boxes=torch.tensor(bbox_bins), - size=image_size - ).tolist() + bboxes = self.box_quantizer.dequantize(boxes=torch.tensor(bbox_bins), size=image_size).tolist() - phrase = phrase.encode('ascii',errors='ignore').decode('ascii') + phrase = phrase.encode("ascii", errors="ignore").decode("ascii") for _bboxes in bboxes: # Prepare instance. instance = {} - instance['bbox'] = _bboxes + instance["bbox"] = _bboxes # exclude non-ascii characters - instance['cat_name'] = phrase + instance["cat_name"] = phrase instances.append(instance) return instances - def parse_description_with_polygons_from_text_and_spans(self, text, pattern, image_size, - allow_empty_phrase=False, - polygon_sep_token='', - polygon_start_token='', - polygon_end_token='', - with_box_at_start=False, - ): - + def parse_description_with_polygons_from_text_and_spans( + self, + text, + pattern, + image_size, + allow_empty_phrase=False, + polygon_sep_token="", + polygon_start_token="", + polygon_end_token="", + with_box_at_start=False, + ): # ref_seg format: '<><><><><><>' # ignore and - text = text.replace('', '') - text = text.replace('', '') - text = text.replace('', '') + text = text.replace("", "") + text = text.replace("", "") + text = text.replace("", "") if allow_empty_phrase: pattern = rf"(?:(?:|{re.escape(polygon_sep_token)}|{re.escape(polygon_start_token)}|{re.escape(polygon_end_token)}){{4,}})" else: - # [^<]+: This part matches one or more characters that are not the < symbol. + # [^<]+: This part matches one or more characters that are not the < symbol. # The ^ inside the square brackets [] is a negation, meaning it matches anything except <. # pattern = rf"([^<]+(?:|{re.escape(polygon_sep_token)}|{re.escape(polygon_start_token)}|{re.escape(polygon_end_token)}){{4,}})" phrases = re.findall(pattern, text) - phrase_string_pattern = r'^\s*(.*?)(?=||||||)' - box_pattern = rf'((?:)+)(?:{re.escape(polygon_sep_token)}|$)' + phrase_string_pattern = r"^\s*(.*?)(?=||||||)" + box_pattern = rf"((?:)+)(?:{re.escape(polygon_sep_token)}|$)" # one polygons instance is separated by polygon_start_token and polygon_end_token - polygons_instance_pattern = rf'{re.escape(polygon_start_token)}(.*?){re.escape(polygon_end_token)}' - + polygons_instance_pattern = rf"{re.escape(polygon_start_token)}(.*?){re.escape(polygon_end_token)}" + instances = [] for phrase_text in phrases: - # exclude loc_\d+> # need to get span if want to include category score - phrase_text_strip = re.sub(r'^loc_\d+>', '', phrase_text, count=1) + phrase_text_strip = re.sub(r"^loc_\d+>", "", phrase_text, count=1) # phrase = phrase.replace('', '') # phrase = phrase.replace('poly>', '') - if phrase_text_strip == '' and not allow_empty_phrase: + if phrase_text_strip == "" and not allow_empty_phrase: continue - - # parse phrase, get string + # parse phrase, get string phrase = re.search(phrase_string_pattern, phrase_text_strip) if phrase is None: continue @@ -947,10 +1030,10 @@ def parse_description_with_polygons_from_text_and_spans(self, text, pattern, ima instance = {} # polygons_parsed= list(re.finditer(box_pattern, phrase_text)) - if isinstance(_polygons_instances_parsed, str): - polygons_parsed= list(re.finditer(box_pattern, _polygons_instances_parsed)) + if isinstance(_polygons_instances_parsed, str): + polygons_parsed = list(re.finditer(box_pattern, _polygons_instances_parsed)) else: - polygons_parsed= list(re.finditer(box_pattern, _polygons_instances_parsed.group(1))) + polygons_parsed = list(re.finditer(box_pattern, _polygons_instances_parsed.group(1))) if len(polygons_parsed) == 0: continue @@ -961,7 +1044,7 @@ def parse_description_with_polygons_from_text_and_spans(self, text, pattern, ima # group 1: whole ... _polygon = _polygon_parsed.group(1) # parse into list of int - _polygon = [int(_loc_parsed.group(1)) for _loc_parsed in re.finditer(r'', _polygon)] + _polygon = [int(_loc_parsed.group(1)) for _loc_parsed in re.finditer(r"", _polygon)] if with_box_at_start and len(bbox) == 0: if len(_polygon) > 4: # no valid bbox prediction @@ -969,25 +1052,27 @@ def parse_description_with_polygons_from_text_and_spans(self, text, pattern, ima _polygon = _polygon[4:] else: bbox = [0, 0, 0, 0] - # abandon last element if is not paired + # abandon last element if is not paired if len(_polygon) % 2 == 1: _polygon = _polygon[:-1] - + # reshape into (n, 2) - _polygon = self.coordinates_quantizer.dequantize( - torch.tensor(np.array(_polygon).reshape(-1, 2)), - size=image_size - ).reshape(-1).tolist() + _polygon = ( + self.coordinates_quantizer.dequantize( + torch.tensor(np.array(_polygon).reshape(-1, 2)), size=image_size + ) + .reshape(-1) + .tolist() + ) # reshape back polygons.append(_polygon) - instance['cat_name'] = phrase - instance['polygons'] = polygons + instance["cat_name"] = phrase + instance["polygons"] = polygons if len(bbox) != 0: - instance['bbox'] = self.box_quantizer.dequantize( - boxes=torch.tensor([bbox]), - size=image_size - ).tolist()[0] + instance["bbox"] = self.box_quantizer.dequantize( + boxes=torch.tensor([bbox]), size=image_size + ).tolist()[0] instances.append(instance) @@ -1010,84 +1095,82 @@ def __call__( if isinstance(parse_tasks, str): parse_tasks = [parse_tasks] for _parse_task in parse_tasks: - assert _parse_task in self.parse_tasks, f'parse task {_parse_task} not supported' - - # sequence or text should be provided - assert text is not None, 'text should be provided' + assert _parse_task in self.parse_tasks, f"parse task {_parse_task} not supported" - parsed_dict = { - 'text': text - } + # sequence or text should be provided + assert text is not None, "text should be provided" + + parsed_dict = {"text": text} for task in self.parse_tasks: if parse_tasks is not None and task not in parse_tasks: continue - pattern = self.parse_tasks_configs[task].get('PATTERN', None) + pattern = self.parse_tasks_configs[task].get("PATTERN", None) - if task == 'ocr': + if task == "ocr": instances = self.parse_ocr_from_text_and_spans( text, pattern=pattern, image_size=image_size, - area_threshold=self.parse_tasks_configs[task].get('AREA_THRESHOLD', 0.01), + area_threshold=self.parse_tasks_configs[task].get("AREA_THRESHOLD", 0.01), ) - parsed_dict['ocr'] = instances - elif task == 'phrase_grounding': - instances = self.parse_phrase_grounding_from_text_and_spans( + parsed_dict["ocr"] = instances + elif task == "phrase_grounding": + instances = self.parse_phrase_grounding_from_text_and_spans( text, pattern=pattern, image_size=image_size, ) - parsed_dict['phrase_grounding'] = instances - elif task == 'pure_text': - parsed_dict['pure_text'] = text - elif task == 'description_with_bboxes': - instances = self.parse_description_with_bboxes_from_text_and_spans( + parsed_dict["phrase_grounding"] = instances + elif task == "pure_text": + parsed_dict["pure_text"] = text + elif task == "description_with_bboxes": + instances = self.parse_description_with_bboxes_from_text_and_spans( text, pattern=pattern, image_size=image_size, ) - parsed_dict['description_with_bboxes'] = instances - elif task == 'description_with_polygons': - instances = self.parse_description_with_polygons_from_text_and_spans( + parsed_dict["description_with_bboxes"] = instances + elif task == "description_with_polygons": + instances = self.parse_description_with_polygons_from_text_and_spans( text, pattern=pattern, image_size=image_size, ) - parsed_dict['description_with_polygons'] = instances - elif task == 'polygons': - instances = self.parse_description_with_polygons_from_text_and_spans( + parsed_dict["description_with_polygons"] = instances + elif task == "polygons": + instances = self.parse_description_with_polygons_from_text_and_spans( text, pattern=pattern, image_size=image_size, allow_empty_phrase=True, ) - parsed_dict['polygons'] = instances - elif task == 'bboxes': - instances = self.parse_description_with_bboxes_from_text_and_spans( + parsed_dict["polygons"] = instances + elif task == "bboxes": + instances = self.parse_description_with_bboxes_from_text_and_spans( text, pattern=pattern, image_size=image_size, allow_empty_phrase=True, ) - parsed_dict['bboxes'] = instances - elif task == 'description_with_bboxes_or_polygons': - if '' in text: + parsed_dict["bboxes"] = instances + elif task == "description_with_bboxes_or_polygons": + if "" in text: # only support either polygons or bboxes, not both at the same time - instances = self.parse_description_with_polygons_from_text_and_spans( + instances = self.parse_description_with_polygons_from_text_and_spans( text, pattern=pattern, image_size=image_size, ) else: - instances = self.parse_description_with_bboxes_from_text_and_spans( + instances = self.parse_description_with_bboxes_from_text_and_spans( text, pattern=pattern, image_size=image_size, ) - parsed_dict['description_with_bboxes_or_polygons'] = instances + parsed_dict["description_with_bboxes_or_polygons"] = instances else: raise ValueError("task {} is not supported".format(task)) - return parsed_dict \ No newline at end of file + return parsed_dict From eaa2e7b6b976e8ad93a352643dc467ff620a76c8 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 16:15:44 +0200 Subject: [PATCH 08/35] fix florence2 modeling, rm einops.rearrange dep --- src/transformers/models/florence/modeling_florence2.py | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/src/transformers/models/florence/modeling_florence2.py b/src/transformers/models/florence/modeling_florence2.py index c6e09c303ed9..a4a7a5d0aafb 100644 --- a/src/transformers/models/florence/modeling_florence2.py +++ b/src/transformers/models/florence/modeling_florence2.py @@ -41,8 +41,6 @@ Seq2SeqLMOutput, Seq2SeqModelOutput, ) - -# from einops import rearrange from ...modeling_utils import PreTrainedModel from ...utils import ( ModelOutput, @@ -211,7 +209,7 @@ def __init__(self, norm, fn, drop_path=None): def forward(self, x, *args, **kwargs): shortcut = x - if self.norm != None: + if self.norm is not None: x, size = self.fn(self.norm(x), *args, **kwargs) else: x, size = self.fn(x, *args, **kwargs) @@ -293,17 +291,11 @@ def forward(self, x, size): if len(x.size()) == 3: if self.norm and self.pre_norm: x = self.norm(x) - # x = rearrange( - # x, 'b (h w) c -> b c h w', - # h=H, w=W - # ) x = x.view(-1, H, W, x.size(-1)).permute(0, 3, 1, 2) x = self.proj(x) _, _, H, W = x.shape - # TODO: check if this is correct - # x = rearrange(x, 'b c h w -> b (h w) c') x = x.permute(0, 2, 3, 1).contiguous() B, H, W, C = x.size() x = x.view(B, -1, C) From 492d30f2373469e8d1d4bb4d039f46fc13183926 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 16:25:54 +0200 Subject: [PATCH 09/35] fix florence2 config, make fixup --- src/transformers/models/florence/configuration_florence2.py | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/src/transformers/models/florence/configuration_florence2.py b/src/transformers/models/florence/configuration_florence2.py index 643d6378dfa5..3b8467187571 100644 --- a/src/transformers/models/florence/configuration_florence2.py +++ b/src/transformers/models/florence/configuration_florence2.py @@ -11,11 +11,10 @@ # 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 warnings - -""" Florence-2 configuration""" +"""Florence-2 configuration""" +import warnings from ...configuration_utils import PretrainedConfig from ...utils import logging From 5b2737b3cafe96ff82561ceb798827ae1adb54aa Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 16:34:55 +0200 Subject: [PATCH 10/35] fix florence2 processing, make fixup --- .../models/florence/processing_florence2.py | 210 +++++++++--------- 1 file changed, 104 insertions(+), 106 deletions(-) diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py index b6531de62945..51ce3be8f6dc 100644 --- a/src/transformers/models/florence/processing_florence2.py +++ b/src/transformers/models/florence/processing_florence2.py @@ -564,110 +564,108 @@ def _create_black_list_of_phrase_grounding(self): "phrase_grounding" in self.parse_tasks and self.parse_tasks_configs["phrase_grounding"]["FILTER_BY_BLACK_LIST"] ): - black_list = set( - [ - "it", - "I", - "me", - "mine", - "you", - "your", - "yours", - "he", - "him", - "his", - "she", - "her", - "hers", - "they", - "them", - "their", - "theirs", - "one", - "oneself", - "we", - "us", - "our", - "ours", - "you", - "your", - "yours", - "they", - "them", - "their", - "theirs", - "mine", - "yours", - "his", - "hers", - "its", - "ours", - "yours", - "theirs", - "myself", - "yourself", - "himself", - "herself", - "itself", - "ourselves", - "yourselves", - "themselves", - "this", - "that", - "these", - "those", - "who", - "whom", - "whose", - "which", - "what", - "who", - "whom", - "whose", - "which", - "that", - "all", - "another", - "any", - "anybody", - "anyone", - "anything", - "each", - "everybody", - "everyone", - "everything", - "few", - "many", - "nobody", - "none", - "one", - "several", - "some", - "somebody", - "someone", - "something", - "each other", - "one another", - "myself", - "yourself", - "himself", - "herself", - "itself", - "ourselves", - "yourselves", - "themselves", - "the image", - "image", - "images", - "the", - "a", - "an", - "a group", - "other objects", - "lots", - "a set", - ] - ) + black_list = { + "it", + "I", + "me", + "mine", + "you", + "your", + "yours", + "he", + "him", + "his", + "she", + "her", + "hers", + "they", + "them", + "their", + "theirs", + "one", + "oneself", + "we", + "us", + "our", + "ours", + "you", + "your", + "yours", + "they", + "them", + "their", + "theirs", + "mine", + "yours", + "his", + "hers", + "its", + "ours", + "yours", + "theirs", + "myself", + "yourself", + "himself", + "herself", + "itself", + "ourselves", + "yourselves", + "themselves", + "this", + "that", + "these", + "those", + "who", + "whom", + "whose", + "which", + "what", + "who", + "whom", + "whose", + "which", + "that", + "all", + "another", + "any", + "anybody", + "anyone", + "anything", + "each", + "everybody", + "everyone", + "everything", + "few", + "many", + "nobody", + "none", + "one", + "several", + "some", + "somebody", + "someone", + "something", + "each other", + "one another", + "myself", + "yourself", + "himself", + "herself", + "itself", + "ourselves", + "yourselves", + "themselves", + "the image", + "image", + "images", + "the", + "a", + "an", + "a group", + "other objects", + "lots", + "a set", + } return black_list @@ -826,8 +824,8 @@ def parse_ocr_from_text_and_spans( ) if area_threshold > 0: - x_coords = [i for i in quad_box[0::2]] - y_coords = [i for i in quad_box[1::2]] + x_coords = list(quad_box[0::2]) + y_coords = list(quad_box[1::2]) # apply the Shoelace formula area = 0.5 * abs( From 7cef9d9cb15d4363eb705ce5f6f15e6e95a66382 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 17:27:03 +0200 Subject: [PATCH 11/35] fix florence2 processing, make repo-consistency --- src/transformers/models/florence/processing_florence2.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py index 51ce3be8f6dc..60b9c39bd049 100644 --- a/src/transformers/models/florence/processing_florence2.py +++ b/src/transformers/models/florence/processing_florence2.py @@ -307,7 +307,7 @@ def __call__( # Copied from transformers.models.clip.processing_clip.CLIPProcessor.batch_decode with CLIP->Florence2 def batch_decode(self, *args, **kwargs): """ - This method forwards all its arguments to BartTokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please + This method forwards all its arguments to Florence2TokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please refer to the docstring of this method for more information. """ return self.tokenizer.batch_decode(*args, **kwargs) @@ -315,7 +315,7 @@ def batch_decode(self, *args, **kwargs): # Copied from transformers.models.clip.processing_clip.CLIPProcessor.decode with CLIP->Florence2 def decode(self, *args, **kwargs): """ - This method forwards all its arguments to BartTokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to + This method forwards all its arguments to Florence2TokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to the docstring of this method for more information. """ return self.tokenizer.decode(*args, **kwargs) From 8875899b5cb848938715ee00d31807c520238673 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 17:49:28 +0200 Subject: [PATCH 12/35] fix processing, add PILImageResampling --- src/transformers/models/florence/processing_florence2.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py index 60b9c39bd049..f37fb9d3c89e 100644 --- a/src/transformers/models/florence/processing_florence2.py +++ b/src/transformers/models/florence/processing_florence2.py @@ -24,7 +24,7 @@ import torch from ...feature_extraction_utils import BatchFeature -from ...image_utils import ChannelDimension, ImageInput, is_valid_image +from ...image_utils import ChannelDimension, ImageInput, PILImageResampling, is_valid_image from ...models.bart.tokenization_bart import BartTokenizer from ...models.bart.tokenization_bart_fast import BartTokenizerFast from ...models.t5.tokenization_t5 import T5Tokenizer @@ -191,11 +191,11 @@ def __call__( do_normalize: bool = None, image_mean: Optional[Union[float, List[float]]] = None, image_std: Optional[Union[float, List[float]]] = None, - data_format: Optional[ChannelDimension] = "channels_first", # noqa: F821 + data_format: Optional[ChannelDimension] = "channels_first", input_data_format: Optional[ - Union[str, ChannelDimension] # noqa: F821 + Union[str, ChannelDimension] ] = None, - resample: "PILImageResampling" = None, # noqa: F821 + resample: PILImageResampling = None, # noqa: F821 # type: ignore do_convert_rgb: bool = None, do_thumbnail: bool = None, do_align_long_axis: bool = None, From 86d0862043fa89b838fa435dd84d4c12b75f9e8f Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 18:01:01 +0200 Subject: [PATCH 13/35] fix import of optional dependencies --- src/transformers/models/florence/__init__.py | 21 ++++++++++++++++--- .../models/florence/processing_florence2.py | 17 +++++++++------ 2 files changed, 29 insertions(+), 9 deletions(-) diff --git a/src/transformers/models/florence/__init__.py b/src/transformers/models/florence/__init__.py index 262e7173ca64..693c2700e0e5 100644 --- a/src/transformers/models/florence/__init__.py +++ b/src/transformers/models/florence/__init__.py @@ -17,6 +17,7 @@ OptionalDependencyNotAvailable, _LazyModule, is_torch_available, + is_vision_available, ) @@ -27,11 +28,18 @@ "Florence2VisionConfig", ], "processing_florence2": [ - "Florence2PostProcesser", - "Florence2Processor", + "Florence2PostProcesser" ], } +try: + if not is_vision_available(): + raise OptionalDependencyNotAvailable() +except OptionalDependencyNotAvailable: + pass +else: + _import_structure["processing_florence2"].append("Florence2Processor") + try: if not is_torch_available(): @@ -55,9 +63,16 @@ ) from .processing_florence2 import ( Florence2PostProcesser, - Florence2Processor, ) + try: + if not is_vision_available(): + raise OptionalDependencyNotAvailable() + except OptionalDependencyNotAvailable: + pass + else: + from .processing_florence2 import Florence2Processor + try: if not is_torch_available(): raise OptionalDependencyNotAvailable() diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py index f37fb9d3c89e..1d04201b324a 100644 --- a/src/transformers/models/florence/processing_florence2.py +++ b/src/transformers/models/florence/processing_florence2.py @@ -24,7 +24,7 @@ import torch from ...feature_extraction_utils import BatchFeature -from ...image_utils import ChannelDimension, ImageInput, PILImageResampling, is_valid_image +from ...image_utils import ChannelDimension, ImageInput, is_valid_image from ...models.bart.tokenization_bart import BartTokenizer from ...models.bart.tokenization_bart_fast import BartTokenizerFast from ...models.t5.tokenization_t5 import T5Tokenizer @@ -36,7 +36,14 @@ TextInput, TruncationStrategy, ) -from ...utils import TensorType +from ...utils import ( + TensorType, + is_vision_available, +) + + +if is_vision_available(): + from ...image_utils import PILImageResampling logger = logging.getLogger(__name__) @@ -192,10 +199,8 @@ def __call__( image_mean: Optional[Union[float, List[float]]] = None, image_std: Optional[Union[float, List[float]]] = None, data_format: Optional[ChannelDimension] = "channels_first", - input_data_format: Optional[ - Union[str, ChannelDimension] - ] = None, - resample: PILImageResampling = None, # noqa: F821 # type: ignore + input_data_format: Optional[Union[str, ChannelDimension]] = None, + resample: PILImageResampling = None, # noqa: F821 do_convert_rgb: bool = None, do_thumbnail: bool = None, do_align_long_axis: bool = None, From 3765e739cb9e8493e99f456d7501aca0835ff508 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 18:18:15 +0200 Subject: [PATCH 14/35] add florence2 to models.__init__ --- src/transformers/models/__init__.py | 1 + 1 file changed, 1 insertion(+) diff --git a/src/transformers/models/__init__.py b/src/transformers/models/__init__.py index 24b602f18c8f..83f6427d8710 100644 --- a/src/transformers/models/__init__.py +++ b/src/transformers/models/__init__.py @@ -86,6 +86,7 @@ fastspeech2_conformer, flaubert, flava, + florence, fnet, focalnet, fsmt, From 3f915bdafa4eecc8824484bb859a5a2f4298636f Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 18:18:35 +0200 Subject: [PATCH 15/35] add florence2 to transformers.__init__ --- src/transformers/__init__.py | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py index 4976a4a1b90e..df678da14048 100755 --- a/src/transformers/__init__.py +++ b/src/transformers/__init__.py @@ -423,6 +423,11 @@ "FlavaMultimodalConfig", "FlavaTextConfig", ], + "models.florence": [ + "Florence2Config", + "Florence2LanguageConfig", + "Florence2VisionConfig", + ], "models.fnet": ["FNetConfig"], "models.focalnet": ["FocalNetConfig"], "models.fsmt": [ @@ -1130,6 +1135,7 @@ _import_structure["models.dpt"].extend(["DPTFeatureExtractor", "DPTImageProcessor"]) _import_structure["models.efficientnet"].append("EfficientNetImageProcessor") _import_structure["models.flava"].extend(["FlavaFeatureExtractor", "FlavaImageProcessor", "FlavaProcessor"]) + _import_structure["models.florence"].extend(["Florence2PostProcesser", "Florence2Processor"]) _import_structure["models.fuyu"].extend(["FuyuImageProcessor", "FuyuProcessor"]) _import_structure["models.glpn"].extend(["GLPNFeatureExtractor", "GLPNImageProcessor"]) _import_structure["models.grounding_dino"].extend(["GroundingDinoImageProcessor"]) @@ -2119,6 +2125,14 @@ "FlavaTextModel", ] ) + _import_structure["models.florence"].extend( + [ + "Florence2PreTrainedModel", + "Florence2VisionModel", + "Florence2VisionModelWithProjection", + "Florence2ForConditionalGeneration", + ] + ) _import_structure["models.fnet"].extend( [ "FNetForMaskedLM", @@ -5013,6 +5027,11 @@ FlavaMultimodalConfig, FlavaTextConfig, ) + from .models.florence import ( + Florence2Config, + Florence2LanguageConfig, + Florence2VisionConfig, + ) from .models.fnet import FNetConfig from .models.focalnet import FocalNetConfig from .models.fsmt import ( @@ -5751,6 +5770,7 @@ FlavaImageProcessor, FlavaProcessor, ) + from .models.florence import Florence2Processor, Florence2PostProcesser from .models.fuyu import FuyuImageProcessor, FuyuProcessor from .models.glpn import GLPNFeatureExtractor, GLPNImageProcessor from .models.grounding_dino import GroundingDinoImageProcessor @@ -6593,6 +6613,12 @@ FlavaPreTrainedModel, FlavaTextModel, ) + from .models.florence import ( + Florence2ForConditionalGeneration, + Florence2PreTrainedModel, + Florence2VisionModel, + Florence2VisionModelWithProjection, + ) from .models.fnet import ( FNetForMaskedLM, FNetForMultipleChoice, From bdb3706b8fc656807a80e72ceb9a2259f42dcadf Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 18:21:11 +0200 Subject: [PATCH 16/35] make fix-copies --- .../florence/configuration_florence2.py | 63 ++++++++++--------- src/transformers/utils/dummy_pt_objects.py | 28 +++++++++ .../utils/dummy_vision_objects.py | 14 +++++ 3 files changed, 76 insertions(+), 29 deletions(-) diff --git a/src/transformers/models/florence/configuration_florence2.py b/src/transformers/models/florence/configuration_florence2.py index 3b8467187571..5206ddb460e0 100644 --- a/src/transformers/models/florence/configuration_florence2.py +++ b/src/transformers/models/florence/configuration_florence2.py @@ -35,23 +35,23 @@ class Florence2VisionConfig(PretrainedConfig): Args: drop_path_rate (`float`, *optional*, defaults to 0.1): The dropout rate of the drop path layer. - patch_size (`List[int]`, *optional*, defaults to [7, 3, 3, 3]): + patch_size (`List[int]`, *optional*, defaults to `[7, 3, 3, 3]`): The patch size of the image. - patch_stride (`List[int]`, *optional*, defaults to [4, 2, 2, 2]): + patch_stride (`List[int]`, *optional*, defaults to `[4, 2, 2, 2]`): The patch stride of the image. - patch_padding (`List[int]`, *optional*, defaults to [3, 1, 1, 1]): + patch_padding (`List[int]`, *optional*, defaults to `[3, 1, 1, 1]`): The patch padding of the image. - patch_prenorm (`List[bool]`, *optional*, defaults to [false, true, true, true]): + patch_prenorm (`List[bool]`, *optional*, defaults to `[False, True, True, True]`): Whether to apply layer normalization before the patch embedding layer. - enable_checkpoint (`bool`, *optional*, defaults to False): + enable_checkpoint (`bool`, *optional*, defaults to `False`): Whether to enable checkpointing. - dim_embed (`List[int]`, *optional*, defaults to [256, 512, 1024, 2048]): + dim_embed (`List[int]`, *optional*, defaults to `[256, 512, 1024, 2048]`): The dimension of the embedding layer. - num_heads (`List[int]`, *optional*, defaults to [8, 16, 32, 64]): + num_heads (`List[int]`, *optional*, defaults to `[8, 16, 32, 64]`): The number of attention heads. - num_groups (`List[int]`, *optional*, defaults to [8, 16, 32, 64]): + num_groups (`List[int]`, *optional*, defaults to `[8, 16, 32, 64]`): The number of groups. - depths (`List[int]`, *optional*, defaults to [1, 1, 9, 1]): + depths (`List[int]`, *optional*, defaults to `[1, 1, 9, 1]`): The depth of the model. window_size (`int`, *optional*, defaults to 12): The window size of the model. @@ -61,7 +61,7 @@ class Florence2VisionConfig(PretrainedConfig): The configuration of the visual temporal embedding. image_pos_embed (`dict`, *optional*): The configuration of the image position embedding. - image_feature_source (`List[str]`, *optional*, defaults to ["spatial_avg_pool", "temporal_avg_pool"]): + image_feature_source (`List[str]`, *optional*, defaults to `['spatial_avg_pool', 'temporal_avg_pool']`): The source of the image feature. Example: @@ -134,48 +134,53 @@ class Florence2LanguageConfig(PretrainedConfig): vocab_size (`int`, *optional*, defaults to 51289): Vocabulary size of the Florence2Language model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [`Florence2LanguageModel`]. - d_model (`int`, *optional*, defaults to 1024): - Dimensionality of the layers and the pooler layer. + max_position_embeddings (`int`, *optional*, defaults to 1024): + The maximum sequence length that this model might ever be used with. Typically set this to something large + just in case (e.g., 512 or 1024 or 2048). encoder_layers (`int`, *optional*, defaults to 12): Number of encoder layers. - decoder_layers (`int`, *optional*, defaults to 12): - Number of decoder layers. + encoder_ffn_dim (`int`, *optional*, defaults to 4096): + Dimensionality of the "intermediate" (often named feed-forward) layer in decoder. encoder_attention_heads (`int`, *optional*, defaults to 16): Number of attention heads for each attention layer in the Transformer encoder. - decoder_attention_heads (`int`, *optional*, defaults to 16): - Number of attention heads for each attention layer in the Transformer decoder. + decoder_layers (`int`, *optional*, defaults to 12): + Number of decoder layers. decoder_ffn_dim (`int`, *optional*, defaults to 4096): Dimensionality of the "intermediate" (often named feed-forward) layer in decoder. - encoder_ffn_dim (`int`, *optional*, defaults to 4096): - Dimensionality of the "intermediate" (often named feed-forward) layer in decoder. + decoder_attention_heads (`int`, *optional*, defaults to 16): + Number of attention heads for each attention layer in the Transformer decoder. + encoder_layerdrop (`float`, *optional*, defaults to 0.0): + The LayerDrop probability for the encoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556) + for more details. + decoder_layerdrop (`float`, *optional*, defaults to 0.0): + The LayerDrop probability for the decoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556) + for more details. activation_function (`str` or `function`, *optional*, defaults to `"gelu"`): The non-linear activation function (function or string) in the encoder and pooler. If string, `"gelu"`, `"relu"`, `"silu"` and `"gelu_new"` are supported. + d_model (`int`, *optional*, defaults to 1024): + Dimensionality of the layers and the pooler layer. dropout (`float`, *optional*, defaults to 0.1): The dropout probability for all fully connected layers in the embeddings, encoder, and pooler. attention_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for the attention probabilities. activation_dropout (`float`, *optional*, defaults to 0.0): The dropout ratio for activations inside the fully connected layer. - classifier_dropout (`float`, *optional*, defaults to 0.0): - The dropout ratio for classifier. - max_position_embeddings (`int`, *optional*, defaults to 1024): - The maximum sequence length that this model might ever be used with. Typically set this to something large - just in case (e.g., 512 or 1024 or 2048). init_std (`float`, *optional*, defaults to 0.02): The standard deviation of the truncated_normal_initializer for initializing all weight matrices. - encoder_layerdrop (`float`, *optional*, defaults to 0.0): - The LayerDrop probability for the encoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556) - for more details. - decoder_layerdrop (`float`, *optional*, defaults to 0.0): - The LayerDrop probability for the decoder. See the [LayerDrop paper](see https://arxiv.org/abs/1909.11556) - for more details. + classifier_dropout (`float`, *optional*, defaults to 0.0): + The dropout ratio for classifier. scale_embedding (`bool`, *optional*, defaults to `False`): Scale embeddings by diving by sqrt(d_model). use_cache (`bool`, *optional*, defaults to `True`): Whether or not the model should return the last key/values attentions (not used by all models). num_labels (`int`, *optional*, defaults to 3): The number of labels to use in [`Florence2LanguageForSequenceClassification`]. + pad_token_id (``, *optional*, defaults to 1): + bos_token_id (``, *optional*, defaults to 0): + eos_token_id (``, *optional*, defaults to 2): + is_encoder_decoder (``, *optional*, defaults to `True`): + decoder_start_token_id (``, *optional*, defaults to 2): forced_eos_token_id (`int`, *optional*, defaults to 2): The id of the token to force as the last generated token when `max_length` is reached. Usually set to `eos_token_id`. diff --git a/src/transformers/utils/dummy_pt_objects.py b/src/transformers/utils/dummy_pt_objects.py index 0cda4ed7b963..5660963802fc 100644 --- a/src/transformers/utils/dummy_pt_objects.py +++ b/src/transformers/utils/dummy_pt_objects.py @@ -3955,6 +3955,34 @@ def __init__(self, *args, **kwargs): requires_backends(self, ["torch"]) +class Florence2ForConditionalGeneration(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + +class Florence2PreTrainedModel(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + +class Florence2VisionModel(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + +class Florence2VisionModelWithProjection(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + class FNetForMaskedLM(metaclass=DummyObject): _backends = ["torch"] diff --git a/src/transformers/utils/dummy_vision_objects.py b/src/transformers/utils/dummy_vision_objects.py index a27dc024447f..591c46cb1b5b 100644 --- a/src/transformers/utils/dummy_vision_objects.py +++ b/src/transformers/utils/dummy_vision_objects.py @@ -240,6 +240,20 @@ def __init__(self, *args, **kwargs): requires_backends(self, ["vision"]) +class Florence2Processor(metaclass=DummyObject): + _backends = ["vision"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["vision"]) + + +class Florence2PostProcesser(metaclass=DummyObject): + _backends = ["vision"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["vision"]) + + class FuyuImageProcessor(metaclass=DummyObject): _backends = ["vision"] From 2d81135826977a0f790eca95c2dff2a0bcead0bd Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 18:34:58 +0200 Subject: [PATCH 17/35] fix formatting & typo --- src/transformers/__init__.py | 6 +++--- src/transformers/models/florence/__init__.py | 6 ++---- src/transformers/models/florence/processing_florence2.py | 4 ++-- src/transformers/utils/dummy_vision_objects.py | 4 ++-- 4 files changed, 9 insertions(+), 11 deletions(-) diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py index df678da14048..5733bcda94bc 100755 --- a/src/transformers/__init__.py +++ b/src/transformers/__init__.py @@ -1135,7 +1135,7 @@ _import_structure["models.dpt"].extend(["DPTFeatureExtractor", "DPTImageProcessor"]) _import_structure["models.efficientnet"].append("EfficientNetImageProcessor") _import_structure["models.flava"].extend(["FlavaFeatureExtractor", "FlavaImageProcessor", "FlavaProcessor"]) - _import_structure["models.florence"].extend(["Florence2PostProcesser", "Florence2Processor"]) + _import_structure["models.florence"].extend(["Florence2PostProcessor", "Florence2Processor"]) _import_structure["models.fuyu"].extend(["FuyuImageProcessor", "FuyuProcessor"]) _import_structure["models.glpn"].extend(["GLPNFeatureExtractor", "GLPNImageProcessor"]) _import_structure["models.grounding_dino"].extend(["GroundingDinoImageProcessor"]) @@ -2127,10 +2127,10 @@ ) _import_structure["models.florence"].extend( [ + "Florence2ForConditionalGeneration", "Florence2PreTrainedModel", "Florence2VisionModel", "Florence2VisionModelWithProjection", - "Florence2ForConditionalGeneration", ] ) _import_structure["models.fnet"].extend( @@ -5770,7 +5770,7 @@ FlavaImageProcessor, FlavaProcessor, ) - from .models.florence import Florence2Processor, Florence2PostProcesser + from .models.florence import Florence2PostProcessor, Florence2Processor from .models.fuyu import FuyuImageProcessor, FuyuProcessor from .models.glpn import GLPNFeatureExtractor, GLPNImageProcessor from .models.grounding_dino import GroundingDinoImageProcessor diff --git a/src/transformers/models/florence/__init__.py b/src/transformers/models/florence/__init__.py index 693c2700e0e5..f3dfbf59eba6 100644 --- a/src/transformers/models/florence/__init__.py +++ b/src/transformers/models/florence/__init__.py @@ -27,9 +27,7 @@ "Florence2LanguageConfig", "Florence2VisionConfig", ], - "processing_florence2": [ - "Florence2PostProcesser" - ], + "processing_florence2": ["Florence2PostProcessor"], } try: @@ -62,7 +60,7 @@ Florence2VisionConfig, ) from .processing_florence2 import ( - Florence2PostProcesser, + Florence2PostProcessor, ) try: diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py index 1d04201b324a..3c4568609960 100644 --- a/src/transformers/models/florence/processing_florence2.py +++ b/src/transformers/models/florence/processing_florence2.py @@ -163,7 +163,7 @@ def __init__( "": "What text is in the region {input}?", } - self.post_processor = Florence2PostProcesser(tokenizer=tokenizer) + self.post_processor = Florence2PostProcessor(tokenizer=tokenizer) super().__init__(image_processor, tokenizer) @@ -515,7 +515,7 @@ def dequantize(self, coordinates: torch.Tensor, size): return dequantized_coordinates -class Florence2PostProcesser(object): +class Florence2PostProcessor(object): r""" Florence-2 post process for converting text prediction to various tasks results. diff --git a/src/transformers/utils/dummy_vision_objects.py b/src/transformers/utils/dummy_vision_objects.py index 591c46cb1b5b..7355881d81a0 100644 --- a/src/transformers/utils/dummy_vision_objects.py +++ b/src/transformers/utils/dummy_vision_objects.py @@ -240,14 +240,14 @@ def __init__(self, *args, **kwargs): requires_backends(self, ["vision"]) -class Florence2Processor(metaclass=DummyObject): +class Florence2PostProcessor(metaclass=DummyObject): _backends = ["vision"] def __init__(self, *args, **kwargs): requires_backends(self, ["vision"]) -class Florence2PostProcesser(metaclass=DummyObject): +class Florence2Processor(metaclass=DummyObject): _backends = ["vision"] def __init__(self, *args, **kwargs): From e9f86d077591aca2c3d19784d30018f08c33ceaf Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 18:51:33 +0200 Subject: [PATCH 18/35] fix __init__s & dummies --- src/transformers/__init__.py | 6 ++++++ src/transformers/models/florence/__init__.py | 6 ++++++ src/transformers/utils/dummy_pt_objects.py | 21 ++++++++++++++++++++ 3 files changed, 33 insertions(+) diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py index 5733bcda94bc..f08fd93b1916 100755 --- a/src/transformers/__init__.py +++ b/src/transformers/__init__.py @@ -2128,6 +2128,9 @@ _import_structure["models.florence"].extend( [ "Florence2ForConditionalGeneration", + "Florence2LanguageForConditionalGeneration", + "Florence2LanguageModel", + "Florence2LanguagePreTrainedModel", "Florence2PreTrainedModel", "Florence2VisionModel", "Florence2VisionModelWithProjection", @@ -6615,6 +6618,9 @@ ) from .models.florence import ( Florence2ForConditionalGeneration, + Florence2LanguageForConditionalGeneration, + Florence2LanguageModel, + Florence2LanguagePreTrainedModel, Florence2PreTrainedModel, Florence2VisionModel, Florence2VisionModelWithProjection, diff --git a/src/transformers/models/florence/__init__.py b/src/transformers/models/florence/__init__.py index f3dfbf59eba6..bcde843c77c4 100644 --- a/src/transformers/models/florence/__init__.py +++ b/src/transformers/models/florence/__init__.py @@ -50,6 +50,9 @@ "Florence2VisionModel", "Florence2VisionModelWithProjection", "Florence2ForConditionalGeneration", + "Florence2LanguageModel", + "Florence2LanguagePreTrainedModel", + "Florence2LanguageForConditionalGeneration", ] @@ -79,6 +82,9 @@ else: from .modeling_florence2 import ( Florence2ForConditionalGeneration, + Florence2LanguageForConditionalGeneration, + Florence2LanguageModel, + Florence2LanguagePreTrainedModel, Florence2PreTrainedModel, Florence2VisionModel, Florence2VisionModelWithProjection, diff --git a/src/transformers/utils/dummy_pt_objects.py b/src/transformers/utils/dummy_pt_objects.py index 5660963802fc..f15adc92395c 100644 --- a/src/transformers/utils/dummy_pt_objects.py +++ b/src/transformers/utils/dummy_pt_objects.py @@ -3962,6 +3962,27 @@ def __init__(self, *args, **kwargs): requires_backends(self, ["torch"]) +class Florence2LanguageForConditionalGeneration(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + +class Florence2LanguageModel(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + +class Florence2LanguagePreTrainedModel(metaclass=DummyObject): + _backends = ["torch"] + + def __init__(self, *args, **kwargs): + requires_backends(self, ["torch"]) + + class Florence2PreTrainedModel(metaclass=DummyObject): _backends = ["torch"] From 172a1dd0e26fb9115a8b5aed3d29b6ae3bb1d911 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 19:47:40 +0200 Subject: [PATCH 19/35] add tests florence2 processing --- .../florence/test_processing_florence2.py | 153 ++++++++++++++++++ 1 file changed, 153 insertions(+) create mode 100644 tests/models/florence/test_processing_florence2.py diff --git a/tests/models/florence/test_processing_florence2.py b/tests/models/florence/test_processing_florence2.py new file mode 100644 index 000000000000..d114b4ef5f17 --- /dev/null +++ b/tests/models/florence/test_processing_florence2.py @@ -0,0 +1,153 @@ +# Copyright 2024 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. + +import json +import os +import random +import shutil +import tempfile +import unittest + +import numpy as np +import pytest + +from transformers import BartTokenizer, BartTokenizerFast +from transformers.models.roberta.tokenization_roberta import VOCAB_FILES_NAMES # TODO can I use this? +from transformers.testing_utils import require_vision +from transformers.utils import IMAGE_PROCESSOR_NAME, is_vision_available + +if is_vision_available(): + from PIL import Image + + from transformers import CLIPImageProcessor, Florence2Processor + +@require_vision +class Florence2ProcessorTest(unittest.TestCase): + def setUp(self): + self.tmpdirname = tempfile.mkdtemp() + + vocab_tokens = [] # TODO: add vocab tokens + self.vocab_file = os.path.join(self.tmpdirname, VOCAB_FILES_NAMES["vocab_file"]) + + with open(self.vocab_file, "w", encoding="utf-8") as fp: + fp.write("".join([x + "\n" for x in vocab_tokens])) + + self.processor_file = os.path.join(self.tmpdirname, IMAGE_PROCESSOR_NAME) + + def get_tokenizer(self, **kwargs): + return BartTokenizer.from_pretrained(self.tmpdirname, **kwargs) + + def get_rust_tokenizer(self, **kwargs): + return BartTokenizerFast.from_pretrained(self.tmpdirname, **kwargs) + + def get_image_processor(self, **kwargs): + return CLIPImageProcessor.from_pretrained(self.tmpdirname, **kwargs) + + def tearDown(self): + shutil.rmtree(self.tmpdirname) + + def prepare_image_inputs(self): + """This function prepares a list of PIL images, or a list of numpy arrays if one specifies numpify=True, + or a list of PyTorch tensors if one specifies torchify=True. + """ + image_inputs = [np.random.randint(255, size=(3, 30, 400), dtype=np.uint8)] + + image_inputs = [Image.fromarray(np.moveaxis(x, 0, -1)) for x in image_inputs] + + return image_inputs + + def test_save_load_pretrained_default(self): + tokenizer_slow = self.get_tokenizer() + tokenizer_fast = self.get_rust_tokenizer() + image_processor = self.get_image_processor() + + processor_slow = Florence2Processor(tokenizer=tokenizer_slow, image_processor=image_processor) + processor_slow.save_pretrained(self.tmpdirname) + processor_slow = Florence2Processor.from_pretrained(self.tmpdirname) + + processor_fast = Florence2Processor(tokenizer=tokenizer_fast, image_processor=image_processor) + processor_fast.save_pretrained(self.tmpdirname) + processor_fast = Florence2Processor.from_pretrained(self.tmpdirname) + + self.assertEqual(processor_slow.tokenizer.get_vocab(), tokenizer_slow.get_vocab()) + self.assertEqual(processor_fast.tokenizer.get_vocab(), tokenizer_fast.get_vocab()) + self.assertEqual(tokenizer_slow.get_vocab(), tokenizer_fast.get_vocab()) + self.assertIsInstance(processor_slow.tokenizer, BartTokenizer) + self.assertIsInstance(processor_fast.tokenizer, BartTokenizerFast) + + self.assertEqual(processor_slow.image_processor.to_json_string(), image_processor.to_json_string()) + self.assertEqual(processor_fast.image_processor.to_json_string(), image_processor.to_json_string()) + self.assertIsInstance(processor_slow.image_processor, CLIPImageProcessor) + self.assertIsInstance(processor_fast.image_processor, CLIPImageProcessor) + + def test_save_load_pretrained_additional_features(self): + processor = Florence2Processor(tokenizer=self.get_tokenizer(), image_processor=self.get_image_processor()) + processor.save_pretrained(self.tmpdirname) + + tokenizer_add_kwargs = self.get_tokenizer(bos_token="(BOS)", eos_token="(EOS)") + image_processor_add_kwargs = self.get_image_processor(do_normalize=False, padding_value=1.0) + + processor = Florence2Processor.from_pretrained( + self.tmpdirname, bos_token="(BOS)", eos_token="(EOS)", do_normalize=False, padding_value=1.0 + ) + + self.assertEqual(processor.tokenizer.get_vocab(), tokenizer_add_kwargs.get_vocab()) + self.assertIsInstance(processor.tokenizer, BartTokenizer) + + self.assertEqual(processor.image_processor.to_json_string(), image_processor_add_kwargs.to_json_string()) + self.assertIsInstance(processor.image_processor, CLIPImageProcessor) + + def test_image_processor(self): + image_processor = self.get_image_processor() + tokenizer = self.get_tokenizer() + processor = Florence2Processor(tokenizer=tokenizer, image_processor=image_processor) + + image_inputs = self.prepare_image_inputs() + + input_feat_extract = image_processor(image_inputs, return_tensors="np") + input_processor = processor(images=image_inputs, return_tensors="np") + + for key in input_feat_extract.keys(): + self.assertAlmostEqual(input_feat_extract[key].sum(), input_processor[key].sum(), delta=1e-2) + + def test_tokenizer(self): + image_processor = self.get_image_processor() + tokenizer = self.get_tokenizer() + processor = Florence2Processor(tokenizer=tokenizer, image_processor=image_processor) + + input_str = "huggingface is cool" + + encoded_processor = processor(text=input_str) + encoded_tok = tokenizer(input_str) + for key in encoded_tok.keys(): + self.assertListEqual(encoded_processor[key], encoded_tok[key]) + + + def test_processor(self): + image_processor = self.get_image_processor() + tokenizer = self.get_tokenizer() + + processor = Florence2Processor(tokenizer=tokenizer, image_processor=image_processor) + + input_str = "huggingface is cool" + image_inputs = self.prepare_image_inputs() + + encoded_processor = processor(text=input_str, images=image_inputs) + + self.assertListEqual(list(encoded_processor.keys()), ["input_ids", "token_type_ids", "attention_mask", "pixel_values"]) + + # test if it raises when no input is passed + with pytest.raises(ValueError): + processor() + From 087ef0e9e05a49d284ae5d3ed3aba104e601989d Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 23:53:50 +0200 Subject: [PATCH 20/35] add tests florence2 modeling --- .../florence/test_modeling_florence2.py | 234 ++++++++++++++++++ 1 file changed, 234 insertions(+) create mode 100644 tests/models/florence/test_modeling_florence2.py diff --git a/tests/models/florence/test_modeling_florence2.py b/tests/models/florence/test_modeling_florence2.py new file mode 100644 index 000000000000..6823b4918120 --- /dev/null +++ b/tests/models/florence/test_modeling_florence2.py @@ -0,0 +1,234 @@ +# Copyright 2024 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. + +import json +import os +import random +import requests +import shutil +import tempfile +import unittest + +import numpy as np +import pytest + +from transformers import ( + Florence2VisionConfig, + Florence2LanguageConfig, + Florence2Config, +) + +from transformers.testing_utils import require_torch, require_vision, slow, torch_device +from transformers.utils import is_torch_available, is_vision_available + +from ...test_configuration_common import ConfigTester +from ...test_modeling_common import ( + ModelTesterMixin, +) + + +if is_torch_available(): + import torch + from torch import nn + + from transformers import ( + Florence2LanguageModel, + Florence2VisionModelWithProjection, + Florence2LanguageForConditionalGeneration, + Florence2VisionModel, + Florence2ForConditionalGeneration, + Florence2LanguagePreTrainedModel, + Florence2PreTrainedModel, + ) +else: + torch = {} + +if is_vision_available(): + from PIL import Image + + from transformers import Florence2Processor + +MODEL_ID = "microsoft/Florence-2-base-ft" + +class Florence2LanguageModelTester: + def __init__(self): + pass + +@require_torch +class Florence2LanguageModelTest(ModelTesterMixin, unittest.TestCase): + def setUp(self): + self.model_tester = Florence2LanguageModelTester(self) + self.config_tester = ConfigTester(self, config_class=Florence2LanguageConfig) + + @slow + def test_model_from_pretrained(self): + model = Florence2LanguageModel.from_pretrained(MODEL_ID) + self.assertIsNotNone(model) + + +class Florence2VisionModelTester: + def __init__(self): + pass + +@require_torch +class Florence2VisionModelTest(ModelTesterMixin, unittest.TestCase): + def setUp(self): + self.model_tester = Florence2VisionModelTester(self) + self.config_tester = ConfigTester(self, config_class=Florence2VisionConfig) + + @slow + def test_model_from_pretrained(self): + model = Florence2VisionModel.from_pretrained(MODEL_ID) + self.assertIsNotNone(model) + + +class Florence2VisionModelWithProjectionTester: + def __init__(self): + pass + +@require_torch +class Florence2VisionModelWithProjectionTest(ModelTesterMixin, unittest.TestCase): + def setUp(self): + self.model_tester = Florence2VisionModelWithProjectionTester(self) + self.config_tester = ConfigTester(self, config_class=Florence2VisionConfig) + + @slow + def test_model_from_pretrained(self): + model = Florence2VisionModelWithProjection.from_pretrained(MODEL_ID) + self.assertIsNotNone(model) + + +class Florence2ForConditionalGenerationTester: + def __init__(self): + pass + +@require_torch +class Florence2ForConditionalGenerationTest(ModelTesterMixin, unittest.TestCase): + def setUp(self): + self.model_tester = Florence2ForConditionalGenerationTester(self) + self.config_tester = ConfigTester(self, config_class=Florence2Config) + + @slow + def test_model_from_pretrained(self): + model = Florence2ForConditionalGeneration.from_pretrained(MODEL_ID) + self.assertIsNotNone(model) + + +class Florence2LanguageForConditionalGenerationTester: + def __init__(self): + pass + +@require_torch +class Florence2LanguageForConditionalGenerationTest(ModelTesterMixin, unittest.TestCase): + def setUp(self): + self.model_tester = Florence2LanguageForConditionalGenerationTester(self) + self.config_tester = ConfigTester(self, config_class=Florence2LanguageConfig) + + @slow + def test_model_from_pretrained(self): + model = Florence2LanguageForConditionalGeneration.from_pretrained(MODEL_ID) + self.assertIsNotNone(model) + + +def prepare_img(): + url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.jpg?download=true" + im = Image.open(requests.get(url, stream=True).raw) + return im + + +@require_vision +@require_torch +class Florence2LanguageModelIntegrationTest(unittest.TestCase): + @slow + def test_inference(self): + model = Florence2LanguageModel.from_pretrained(MODEL_ID).to(torch_device) + processor = Florence2Processor.from_pretrained(MODEL_ID) + + img = prepare_img() + inputs = processor(img, return_tensors="pt").to(torch_device) + + with torch.no_grad(): + outputs = model(**inputs) + + # TODO: write test test condition + #self.assertEqual(outputs.last_hidden_state.shape, (1, 1, 768)) + + +@require_vision +@require_torch +class Florence2VisionModelIntegrationTest(unittest.TestCase): + @slow + def test_inference(self): + model = Florence2VisionModel.from_pretrained(MODEL_ID).to(torch_device) + processor = Florence2Processor.from_pretrained(MODEL_ID) + + img = prepare_img() + inputs = processor(img, return_tensors="pt").to(torch_device) + + with torch.no_grad(): + outputs = model(**inputs) + + # TODO: write test condition + + +@require_vision +@require_torch +class Florence2VisionModelWithProjectionIntegrationTest(unittest.TestCase): + @slow + def test_inference(self): + model = Florence2VisionModelWithProjection.from_pretrained(MODEL_ID).to(torch_device) + processor = Florence2Processor.from_pretrained(MODEL_ID) + + img = prepare_img() + inputs = processor(img, return_tensors="pt").to(torch_device) + + with torch.no_grad(): + outputs = model(**inputs) + + # TODO: write test condition + + +@require_vision +@require_torch +class Florence2ForConditionalGenerationIntegrationTest(unittest.TestCase): + @slow + def test_inference(self): + model = Florence2ForConditionalGeneration.from_pretrained(MODEL_ID).to(torch_device) + processor = Florence2Processor.from_pretrained(MODEL_ID) + + img = prepare_img() + inputs = processor(img, return_tensors="pt").to(torch_device) + + with torch.no_grad(): + outputs = model(**inputs) + + # TODO: write test condition + + +@require_vision +@require_torch +class Florence2LanguageForConditionalGenerationIntegrationTest(unittest.TestCase): + @slow + def test_inference(self): + model = Florence2LanguageForConditionalGeneration.from_pretrained(MODEL_ID).to(torch_device) + processor = Florence2Processor.from_pretrained(MODEL_ID) + + img = prepare_img() + inputs = processor(img, return_tensors="pt").to(torch_device) + + with torch.no_grad(): + outputs = model(**inputs) + + # TODO: write test condition + From fb234ef24d8b20f2b6b9bba92a89fd095e6511bd Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Thu, 20 Jun 2024 23:58:16 +0200 Subject: [PATCH 21/35] fix tests florence2 processign vocabfile --- tests/models/florence/test_processing_florence2.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/models/florence/test_processing_florence2.py b/tests/models/florence/test_processing_florence2.py index d114b4ef5f17..a9dba89aed5a 100644 --- a/tests/models/florence/test_processing_florence2.py +++ b/tests/models/florence/test_processing_florence2.py @@ -23,7 +23,7 @@ import pytest from transformers import BartTokenizer, BartTokenizerFast -from transformers.models.roberta.tokenization_roberta import VOCAB_FILES_NAMES # TODO can I use this? +from transformers.models.bart.tokenization_bart import VOCAB_FILES_NAMES from transformers.testing_utils import require_vision from transformers.utils import IMAGE_PROCESSOR_NAME, is_vision_available From 0a615baf4870450829d93f4a794b88169b20510f Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Fri, 21 Jun 2024 03:33:10 +0200 Subject: [PATCH 22/35] fix rm import Florence2PostProcessor --- src/transformers/__init__.py | 4 ++-- src/transformers/models/florence/__init__.py | 6 +----- src/transformers/utils/dummy_vision_objects.py | 7 ------- 3 files changed, 3 insertions(+), 14 deletions(-) diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py index f08fd93b1916..77bf3bdb211a 100755 --- a/src/transformers/__init__.py +++ b/src/transformers/__init__.py @@ -1135,7 +1135,7 @@ _import_structure["models.dpt"].extend(["DPTFeatureExtractor", "DPTImageProcessor"]) _import_structure["models.efficientnet"].append("EfficientNetImageProcessor") _import_structure["models.flava"].extend(["FlavaFeatureExtractor", "FlavaImageProcessor", "FlavaProcessor"]) - _import_structure["models.florence"].extend(["Florence2PostProcessor", "Florence2Processor"]) + _import_structure["models.florence"].extend(["Florence2Processor"]) _import_structure["models.fuyu"].extend(["FuyuImageProcessor", "FuyuProcessor"]) _import_structure["models.glpn"].extend(["GLPNFeatureExtractor", "GLPNImageProcessor"]) _import_structure["models.grounding_dino"].extend(["GroundingDinoImageProcessor"]) @@ -5773,7 +5773,7 @@ FlavaImageProcessor, FlavaProcessor, ) - from .models.florence import Florence2PostProcessor, Florence2Processor + from .models.florence import Florence2Processor from .models.fuyu import FuyuImageProcessor, FuyuProcessor from .models.glpn import GLPNFeatureExtractor, GLPNImageProcessor from .models.grounding_dino import GroundingDinoImageProcessor diff --git a/src/transformers/models/florence/__init__.py b/src/transformers/models/florence/__init__.py index bcde843c77c4..0ce98bda9450 100644 --- a/src/transformers/models/florence/__init__.py +++ b/src/transformers/models/florence/__init__.py @@ -27,7 +27,6 @@ "Florence2LanguageConfig", "Florence2VisionConfig", ], - "processing_florence2": ["Florence2PostProcessor"], } try: @@ -36,7 +35,7 @@ except OptionalDependencyNotAvailable: pass else: - _import_structure["processing_florence2"].append("Florence2Processor") + _import_structure["processing_florence2"] = ["Florence2Processor"] try: @@ -62,9 +61,6 @@ Florence2LanguageConfig, Florence2VisionConfig, ) - from .processing_florence2 import ( - Florence2PostProcessor, - ) try: if not is_vision_available(): diff --git a/src/transformers/utils/dummy_vision_objects.py b/src/transformers/utils/dummy_vision_objects.py index 7355881d81a0..04002cae9b0b 100644 --- a/src/transformers/utils/dummy_vision_objects.py +++ b/src/transformers/utils/dummy_vision_objects.py @@ -240,13 +240,6 @@ def __init__(self, *args, **kwargs): requires_backends(self, ["vision"]) -class Florence2PostProcessor(metaclass=DummyObject): - _backends = ["vision"] - - def __init__(self, *args, **kwargs): - requires_backends(self, ["vision"]) - - class Florence2Processor(metaclass=DummyObject): _backends = ["vision"] From 26e1c23eff0129bea76d394095e401b3d0e63ef0 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Fri, 21 Jun 2024 03:51:55 +0200 Subject: [PATCH 23/35] fix rm unnecessary imports --- src/transformers/__init__.py | 8 -- src/transformers/models/florence/__init__.py | 10 +- src/transformers/utils/dummy_pt_objects.py | 21 ----- .../florence/test_modeling_florence2.py | 94 +++---------------- 4 files changed, 12 insertions(+), 121 deletions(-) diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py index 77bf3bdb211a..8b31ac0260a3 100755 --- a/src/transformers/__init__.py +++ b/src/transformers/__init__.py @@ -425,7 +425,6 @@ ], "models.florence": [ "Florence2Config", - "Florence2LanguageConfig", "Florence2VisionConfig", ], "models.fnet": ["FNetConfig"], @@ -2128,9 +2127,6 @@ _import_structure["models.florence"].extend( [ "Florence2ForConditionalGeneration", - "Florence2LanguageForConditionalGeneration", - "Florence2LanguageModel", - "Florence2LanguagePreTrainedModel", "Florence2PreTrainedModel", "Florence2VisionModel", "Florence2VisionModelWithProjection", @@ -5032,7 +5028,6 @@ ) from .models.florence import ( Florence2Config, - Florence2LanguageConfig, Florence2VisionConfig, ) from .models.fnet import FNetConfig @@ -6618,9 +6613,6 @@ ) from .models.florence import ( Florence2ForConditionalGeneration, - Florence2LanguageForConditionalGeneration, - Florence2LanguageModel, - Florence2LanguagePreTrainedModel, Florence2PreTrainedModel, Florence2VisionModel, Florence2VisionModelWithProjection, diff --git a/src/transformers/models/florence/__init__.py b/src/transformers/models/florence/__init__.py index 0ce98bda9450..c2a08b783125 100644 --- a/src/transformers/models/florence/__init__.py +++ b/src/transformers/models/florence/__init__.py @@ -24,7 +24,6 @@ _import_structure = { "configuration_florence2": [ "Florence2Config", - "Florence2LanguageConfig", "Florence2VisionConfig", ], } @@ -45,20 +44,16 @@ pass else: _import_structure["modeling_florence2"] = [ + "Florence2ForConditionalGeneration", "Florence2PreTrainedModel", "Florence2VisionModel", "Florence2VisionModelWithProjection", - "Florence2ForConditionalGeneration", - "Florence2LanguageModel", - "Florence2LanguagePreTrainedModel", - "Florence2LanguageForConditionalGeneration", ] if TYPE_CHECKING: from .configuration_florence2 import ( Florence2Config, - Florence2LanguageConfig, Florence2VisionConfig, ) @@ -78,9 +73,6 @@ else: from .modeling_florence2 import ( Florence2ForConditionalGeneration, - Florence2LanguageForConditionalGeneration, - Florence2LanguageModel, - Florence2LanguagePreTrainedModel, Florence2PreTrainedModel, Florence2VisionModel, Florence2VisionModelWithProjection, diff --git a/src/transformers/utils/dummy_pt_objects.py b/src/transformers/utils/dummy_pt_objects.py index f15adc92395c..5660963802fc 100644 --- a/src/transformers/utils/dummy_pt_objects.py +++ b/src/transformers/utils/dummy_pt_objects.py @@ -3962,27 +3962,6 @@ def __init__(self, *args, **kwargs): requires_backends(self, ["torch"]) -class Florence2LanguageForConditionalGeneration(metaclass=DummyObject): - _backends = ["torch"] - - def __init__(self, *args, **kwargs): - requires_backends(self, ["torch"]) - - -class Florence2LanguageModel(metaclass=DummyObject): - _backends = ["torch"] - - def __init__(self, *args, **kwargs): - requires_backends(self, ["torch"]) - - -class Florence2LanguagePreTrainedModel(metaclass=DummyObject): - _backends = ["torch"] - - def __init__(self, *args, **kwargs): - requires_backends(self, ["torch"]) - - class Florence2PreTrainedModel(metaclass=DummyObject): _backends = ["torch"] diff --git a/tests/models/florence/test_modeling_florence2.py b/tests/models/florence/test_modeling_florence2.py index 6823b4918120..1d2929ca6f95 100644 --- a/tests/models/florence/test_modeling_florence2.py +++ b/tests/models/florence/test_modeling_florence2.py @@ -25,7 +25,6 @@ from transformers import ( Florence2VisionConfig, - Florence2LanguageConfig, Florence2Config, ) @@ -43,13 +42,9 @@ from torch import nn from transformers import ( - Florence2LanguageModel, - Florence2VisionModelWithProjection, - Florence2LanguageForConditionalGeneration, - Florence2VisionModel, Florence2ForConditionalGeneration, - Florence2LanguagePreTrainedModel, - Florence2PreTrainedModel, + Florence2VisionModel, + Florence2VisionModelWithProjection, ) else: torch = {} @@ -61,19 +56,20 @@ MODEL_ID = "microsoft/Florence-2-base-ft" -class Florence2LanguageModelTester: + +class Florence2ForConditionalGenerationTester: def __init__(self): pass @require_torch -class Florence2LanguageModelTest(ModelTesterMixin, unittest.TestCase): +class Florence2ForConditionalGenerationTest(ModelTesterMixin, unittest.TestCase): def setUp(self): - self.model_tester = Florence2LanguageModelTester(self) - self.config_tester = ConfigTester(self, config_class=Florence2LanguageConfig) + self.model_tester = Florence2ForConditionalGenerationTester(self) + self.config_tester = ConfigTester(self, config_class=Florence2Config) @slow def test_model_from_pretrained(self): - model = Florence2LanguageModel.from_pretrained(MODEL_ID) + model = Florence2ForConditionalGeneration.from_pretrained(MODEL_ID) self.assertIsNotNone(model) @@ -108,39 +104,6 @@ def test_model_from_pretrained(self): model = Florence2VisionModelWithProjection.from_pretrained(MODEL_ID) self.assertIsNotNone(model) - -class Florence2ForConditionalGenerationTester: - def __init__(self): - pass - -@require_torch -class Florence2ForConditionalGenerationTest(ModelTesterMixin, unittest.TestCase): - def setUp(self): - self.model_tester = Florence2ForConditionalGenerationTester(self) - self.config_tester = ConfigTester(self, config_class=Florence2Config) - - @slow - def test_model_from_pretrained(self): - model = Florence2ForConditionalGeneration.from_pretrained(MODEL_ID) - self.assertIsNotNone(model) - - -class Florence2LanguageForConditionalGenerationTester: - def __init__(self): - pass - -@require_torch -class Florence2LanguageForConditionalGenerationTest(ModelTesterMixin, unittest.TestCase): - def setUp(self): - self.model_tester = Florence2LanguageForConditionalGenerationTester(self) - self.config_tester = ConfigTester(self, config_class=Florence2LanguageConfig) - - @slow - def test_model_from_pretrained(self): - model = Florence2LanguageForConditionalGeneration.from_pretrained(MODEL_ID) - self.assertIsNotNone(model) - - def prepare_img(): url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.jpg?download=true" im = Image.open(requests.get(url, stream=True).raw) @@ -149,10 +112,10 @@ def prepare_img(): @require_vision @require_torch -class Florence2LanguageModelIntegrationTest(unittest.TestCase): +class Florence2ForConditionalGenerationIntegrationTest(unittest.TestCase): @slow def test_inference(self): - model = Florence2LanguageModel.from_pretrained(MODEL_ID).to(torch_device) + model = Florence2ForConditionalGeneration.from_pretrained(MODEL_ID).to(torch_device) processor = Florence2Processor.from_pretrained(MODEL_ID) img = prepare_img() @@ -161,8 +124,7 @@ def test_inference(self): with torch.no_grad(): outputs = model(**inputs) - # TODO: write test test condition - #self.assertEqual(outputs.last_hidden_state.shape, (1, 1, 768)) + # TODO: write test condition @require_vision @@ -198,37 +160,3 @@ def test_inference(self): # TODO: write test condition - -@require_vision -@require_torch -class Florence2ForConditionalGenerationIntegrationTest(unittest.TestCase): - @slow - def test_inference(self): - model = Florence2ForConditionalGeneration.from_pretrained(MODEL_ID).to(torch_device) - processor = Florence2Processor.from_pretrained(MODEL_ID) - - img = prepare_img() - inputs = processor(img, return_tensors="pt").to(torch_device) - - with torch.no_grad(): - outputs = model(**inputs) - - # TODO: write test condition - - -@require_vision -@require_torch -class Florence2LanguageForConditionalGenerationIntegrationTest(unittest.TestCase): - @slow - def test_inference(self): - model = Florence2LanguageForConditionalGeneration.from_pretrained(MODEL_ID).to(torch_device) - processor = Florence2Processor.from_pretrained(MODEL_ID) - - img = prepare_img() - inputs = processor(img, return_tensors="pt").to(torch_device) - - with torch.no_grad(): - outputs = model(**inputs) - - # TODO: write test condition - From f11c993ed5e80bc2c4c427a9049757b8933de2fd Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Fri, 21 Jun 2024 14:18:53 +0200 Subject: [PATCH 24/35] fix processing PILImageResampling --- src/transformers/models/florence/processing_florence2.py | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence/processing_florence2.py index 3c4568609960..abedf0a6d06d 100644 --- a/src/transformers/models/florence/processing_florence2.py +++ b/src/transformers/models/florence/processing_florence2.py @@ -43,7 +43,8 @@ if is_vision_available(): - from ...image_utils import PILImageResampling + import PIL.Image + from PIL.Image import Image as PILImage logger = logging.getLogger(__name__) @@ -200,7 +201,7 @@ def __call__( image_std: Optional[Union[float, List[float]]] = None, data_format: Optional[ChannelDimension] = "channels_first", input_data_format: Optional[Union[str, ChannelDimension]] = None, - resample: PILImageResampling = None, # noqa: F821 + resample: Optional[Union[PILImage, PIL.Image.Resampling]] = None, # noqa: F821 do_convert_rgb: bool = None, do_thumbnail: bool = None, do_align_long_axis: bool = None, From 311da9c247bb45bd0081d0c865b50f9218153b18 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Fri, 21 Jun 2024 14:44:16 +0200 Subject: [PATCH 25/35] fix make repo-consistency error --- utils/check_repo.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/utils/check_repo.py b/utils/check_repo.py index dcb1374d8e0b..4232b5b0ff09 100644 --- a/utils/check_repo.py +++ b/utils/check_repo.py @@ -81,6 +81,9 @@ "SeamlessM4Tv2TextToUnitModel", "SeamlessM4Tv2CodeHifiGan", "SeamlessM4Tv2TextToUnitForConditionalGeneration", + "Florence2LanguageForConditionalGeneration", + "Florence2LanguageModel", + "Florence2LanguagePreTrainedModel", ] # Update this list for models that are not tested with a comment explaining the reason it should not be. From b4298146e6e492b9a7f67a307f8806e2376a3656 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Sat, 22 Jun 2024 14:38:04 +0200 Subject: [PATCH 26/35] update tests --- tests/models/florence/__init__.py | 0 .../florence/test_modeling_florence2.py | 42 ++++++++++++------- .../florence/test_processing_florence2.py | 25 +++++------ 3 files changed, 38 insertions(+), 29 deletions(-) create mode 100644 tests/models/florence/__init__.py diff --git a/tests/models/florence/__init__.py b/tests/models/florence/__init__.py new file mode 100644 index 000000000000..e69de29bb2d1 diff --git a/tests/models/florence/test_modeling_florence2.py b/tests/models/florence/test_modeling_florence2.py index 1d2929ca6f95..b0bd628aa972 100644 --- a/tests/models/florence/test_modeling_florence2.py +++ b/tests/models/florence/test_modeling_florence2.py @@ -12,22 +12,13 @@ # See the License for the specific language governing permissions and # limitations under the License. -import json -import os -import random -import requests -import shutil -import tempfile import unittest - -import numpy as np -import pytest +import requests from transformers import ( - Florence2VisionConfig, Florence2Config, + Florence2VisionConfig, ) - from transformers.testing_utils import require_torch, require_vision, slow, torch_device from transformers.utils import is_torch_available, is_vision_available @@ -39,7 +30,6 @@ if is_torch_available(): import torch - from torch import nn from transformers import ( Florence2ForConditionalGeneration, @@ -58,11 +48,17 @@ class Florence2ForConditionalGenerationTester: - def __init__(self): + def __init__(self, parent): + pass + + def prepare_config_and_inputs_for_common(self): pass + @require_torch class Florence2ForConditionalGenerationTest(ModelTesterMixin, unittest.TestCase): + all_model_classes = (Florence2ForConditionalGeneration,) if is_torch_available() else () + def setUp(self): self.model_tester = Florence2ForConditionalGenerationTester(self) self.config_tester = ConfigTester(self, config_class=Florence2Config) @@ -74,11 +70,17 @@ def test_model_from_pretrained(self): class Florence2VisionModelTester: - def __init__(self): + def __init__(self, parent): pass + def prepare_config_and_inputs_for_common(self): + pass + + @require_torch class Florence2VisionModelTest(ModelTesterMixin, unittest.TestCase): + all_model_classes = (Florence2VisionModel,) if is_torch_available() else () + def setUp(self): self.model_tester = Florence2VisionModelTester(self) self.config_tester = ConfigTester(self, config_class=Florence2VisionConfig) @@ -90,11 +92,17 @@ def test_model_from_pretrained(self): class Florence2VisionModelWithProjectionTester: - def __init__(self): + def __init__(self, parent): pass + def prepare_config_and_inputs_for_common(self): + pass + + @require_torch class Florence2VisionModelWithProjectionTest(ModelTesterMixin, unittest.TestCase): + all_model_classes = (Florence2VisionModelWithProjection,) if is_torch_available() else () + def setUp(self): self.model_tester = Florence2VisionModelWithProjectionTester(self) self.config_tester = ConfigTester(self, config_class=Florence2VisionConfig) @@ -104,6 +112,7 @@ def test_model_from_pretrained(self): model = Florence2VisionModelWithProjection.from_pretrained(MODEL_ID) self.assertIsNotNone(model) + def prepare_img(): url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/tasks/car.jpg?download=true" im = Image.open(requests.get(url, stream=True).raw) @@ -125,6 +134,7 @@ def test_inference(self): outputs = model(**inputs) # TODO: write test condition + self.assertIsNotNone(outputs) @require_vision @@ -142,6 +152,7 @@ def test_inference(self): outputs = model(**inputs) # TODO: write test condition + self.assertIsNotNone(outputs) @require_vision @@ -159,4 +170,5 @@ def test_inference(self): outputs = model(**inputs) # TODO: write test condition + self.assertIsNotNone(outputs) diff --git a/tests/models/florence/test_processing_florence2.py b/tests/models/florence/test_processing_florence2.py index a9dba89aed5a..ba5de9f65e38 100644 --- a/tests/models/florence/test_processing_florence2.py +++ b/tests/models/florence/test_processing_florence2.py @@ -12,9 +12,7 @@ # See the License for the specific language governing permissions and # limitations under the License. -import json import os -import random import shutil import tempfile import unittest @@ -27,42 +25,41 @@ from transformers.testing_utils import require_vision from transformers.utils import IMAGE_PROCESSOR_NAME, is_vision_available + if is_vision_available(): from PIL import Image from transformers import CLIPImageProcessor, Florence2Processor + @require_vision class Florence2ProcessorTest(unittest.TestCase): def setUp(self): self.tmpdirname = tempfile.mkdtemp() - vocab_tokens = [] # TODO: add vocab tokens + vocab_tokens = [] # TODO: add vocab tokens self.vocab_file = os.path.join(self.tmpdirname, VOCAB_FILES_NAMES["vocab_file"]) with open(self.vocab_file, "w", encoding="utf-8") as fp: fp.write("".join([x + "\n" for x in vocab_tokens])) - + self.processor_file = os.path.join(self.tmpdirname, IMAGE_PROCESSOR_NAME) def get_tokenizer(self, **kwargs): return BartTokenizer.from_pretrained(self.tmpdirname, **kwargs) - + def get_rust_tokenizer(self, **kwargs): return BartTokenizerFast.from_pretrained(self.tmpdirname, **kwargs) - + def get_image_processor(self, **kwargs): return CLIPImageProcessor.from_pretrained(self.tmpdirname, **kwargs) - + def tearDown(self): shutil.rmtree(self.tmpdirname) def prepare_image_inputs(self): - """This function prepares a list of PIL images, or a list of numpy arrays if one specifies numpify=True, - or a list of PyTorch tensors if one specifies torchify=True. - """ + """This function prepares a list of PIL images""" image_inputs = [np.random.randint(255, size=(3, 30, 400), dtype=np.uint8)] - image_inputs = [Image.fromarray(np.moveaxis(x, 0, -1)) for x in image_inputs] return image_inputs @@ -133,7 +130,6 @@ def test_tokenizer(self): for key in encoded_tok.keys(): self.assertListEqual(encoded_processor[key], encoded_tok[key]) - def test_processor(self): image_processor = self.get_image_processor() tokenizer = self.get_tokenizer() @@ -145,9 +141,10 @@ def test_processor(self): encoded_processor = processor(text=input_str, images=image_inputs) - self.assertListEqual(list(encoded_processor.keys()), ["input_ids", "token_type_ids", "attention_mask", "pixel_values"]) + self.assertListEqual( + list(encoded_processor.keys()), ["input_ids", "token_type_ids", "attention_mask", "pixel_values"] + ) # test if it raises when no input is passed with pytest.raises(ValueError): processor() - From 3d67bebc5c7bbac47445f619cb0d1edd453abb93 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Sat, 22 Jun 2024 14:39:18 +0200 Subject: [PATCH 27/35] rename folders --- src/transformers/__init__.py | 6 +++--- src/transformers/models/__init__.py | 2 +- src/transformers/models/{florence => florence2}/__init__.py | 0 .../{florence => florence2}/configuration_florence2.py | 0 .../models/{florence => florence2}/modeling_florence2.py | 0 .../models/{florence => florence2}/processing_florence2.py | 0 tests/models/{florence => florence2}/__init__.py | 0 .../{florence => florence2}/test_modeling_florence2.py | 0 .../{florence => florence2}/test_processing_florence2.py | 0 9 files changed, 4 insertions(+), 4 deletions(-) rename src/transformers/models/{florence => florence2}/__init__.py (100%) rename src/transformers/models/{florence => florence2}/configuration_florence2.py (100%) rename src/transformers/models/{florence => florence2}/modeling_florence2.py (100%) rename src/transformers/models/{florence => florence2}/processing_florence2.py (100%) rename tests/models/{florence => florence2}/__init__.py (100%) rename tests/models/{florence => florence2}/test_modeling_florence2.py (100%) rename tests/models/{florence => florence2}/test_processing_florence2.py (100%) diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py index 8b31ac0260a3..fea2b07e7d74 100755 --- a/src/transformers/__init__.py +++ b/src/transformers/__init__.py @@ -5026,7 +5026,7 @@ FlavaMultimodalConfig, FlavaTextConfig, ) - from .models.florence import ( + from .models.florence2 import ( Florence2Config, Florence2VisionConfig, ) @@ -5768,7 +5768,7 @@ FlavaImageProcessor, FlavaProcessor, ) - from .models.florence import Florence2Processor + from .models.florence2 import Florence2Processor from .models.fuyu import FuyuImageProcessor, FuyuProcessor from .models.glpn import GLPNFeatureExtractor, GLPNImageProcessor from .models.grounding_dino import GroundingDinoImageProcessor @@ -6611,7 +6611,7 @@ FlavaPreTrainedModel, FlavaTextModel, ) - from .models.florence import ( + from .models.florence2 import ( Florence2ForConditionalGeneration, Florence2PreTrainedModel, Florence2VisionModel, diff --git a/src/transformers/models/__init__.py b/src/transformers/models/__init__.py index 83f6427d8710..ee6abde95e9c 100644 --- a/src/transformers/models/__init__.py +++ b/src/transformers/models/__init__.py @@ -86,7 +86,7 @@ fastspeech2_conformer, flaubert, flava, - florence, + florence2, fnet, focalnet, fsmt, diff --git a/src/transformers/models/florence/__init__.py b/src/transformers/models/florence2/__init__.py similarity index 100% rename from src/transformers/models/florence/__init__.py rename to src/transformers/models/florence2/__init__.py diff --git a/src/transformers/models/florence/configuration_florence2.py b/src/transformers/models/florence2/configuration_florence2.py similarity index 100% rename from src/transformers/models/florence/configuration_florence2.py rename to src/transformers/models/florence2/configuration_florence2.py diff --git a/src/transformers/models/florence/modeling_florence2.py b/src/transformers/models/florence2/modeling_florence2.py similarity index 100% rename from src/transformers/models/florence/modeling_florence2.py rename to src/transformers/models/florence2/modeling_florence2.py diff --git a/src/transformers/models/florence/processing_florence2.py b/src/transformers/models/florence2/processing_florence2.py similarity index 100% rename from src/transformers/models/florence/processing_florence2.py rename to src/transformers/models/florence2/processing_florence2.py diff --git a/tests/models/florence/__init__.py b/tests/models/florence2/__init__.py similarity index 100% rename from tests/models/florence/__init__.py rename to tests/models/florence2/__init__.py diff --git a/tests/models/florence/test_modeling_florence2.py b/tests/models/florence2/test_modeling_florence2.py similarity index 100% rename from tests/models/florence/test_modeling_florence2.py rename to tests/models/florence2/test_modeling_florence2.py diff --git a/tests/models/florence/test_processing_florence2.py b/tests/models/florence2/test_processing_florence2.py similarity index 100% rename from tests/models/florence/test_processing_florence2.py rename to tests/models/florence2/test_processing_florence2.py From 8b88ca743da68aefa9cb28a9202d3334a40c5c8f Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Sat, 22 Jun 2024 14:49:56 +0200 Subject: [PATCH 28/35] fix folder renaming --- src/transformers/__init__.py | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/transformers/__init__.py b/src/transformers/__init__.py index fea2b07e7d74..0f436e11f2dd 100755 --- a/src/transformers/__init__.py +++ b/src/transformers/__init__.py @@ -423,7 +423,7 @@ "FlavaMultimodalConfig", "FlavaTextConfig", ], - "models.florence": [ + "models.florence2": [ "Florence2Config", "Florence2VisionConfig", ], @@ -1134,7 +1134,7 @@ _import_structure["models.dpt"].extend(["DPTFeatureExtractor", "DPTImageProcessor"]) _import_structure["models.efficientnet"].append("EfficientNetImageProcessor") _import_structure["models.flava"].extend(["FlavaFeatureExtractor", "FlavaImageProcessor", "FlavaProcessor"]) - _import_structure["models.florence"].extend(["Florence2Processor"]) + _import_structure["models.florence2"].extend(["Florence2Processor"]) _import_structure["models.fuyu"].extend(["FuyuImageProcessor", "FuyuProcessor"]) _import_structure["models.glpn"].extend(["GLPNFeatureExtractor", "GLPNImageProcessor"]) _import_structure["models.grounding_dino"].extend(["GroundingDinoImageProcessor"]) @@ -2124,7 +2124,7 @@ "FlavaTextModel", ] ) - _import_structure["models.florence"].extend( + _import_structure["models.florence2"].extend( [ "Florence2ForConditionalGeneration", "Florence2PreTrainedModel", From 299670b512dedbd0c9823ee8789c2ea511e12996 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Mon, 24 Jun 2024 01:19:55 +0200 Subject: [PATCH 29/35] fix florence2 modeling, copy DropPath from beit --- .../models/florence2/modeling_florence2.py | 49 ++++++++++++------- 1 file changed, 31 insertions(+), 18 deletions(-) diff --git a/src/transformers/models/florence2/modeling_florence2.py b/src/transformers/models/florence2/modeling_florence2.py index a4a7a5d0aafb..e70cee3ea3cb 100644 --- a/src/transformers/models/florence2/modeling_florence2.py +++ b/src/transformers/models/florence2/modeling_florence2.py @@ -328,28 +328,41 @@ def forward(self, x, size): x = self.proj(x) return x, size +# TODO verify behavior is same as https://github.com/huggingface/pytorch-image-models/blob/main/timm/layers/drop.py#L150 +# Copied from transformers.models.beit.modeling_beit.drop_path +def drop_path(input: torch.Tensor, drop_prob: float = 0.0, training: bool = False) -> torch.Tensor: + """ + Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). + + Comment by Ross Wightman: This is the same as the DropConnect impl I created for EfficientNet, etc networks, + however, the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper... + See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for changing the + layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use 'survival rate' as the + argument. + """ + if drop_prob == 0.0 or not training: + return input + keep_prob = 1 - drop_prob + shape = (input.shape[0],) + (1,) * (input.ndim - 1) # work with diff dim tensors, not just 2D ConvNets + random_tensor = keep_prob + torch.rand(shape, dtype=input.dtype, device=input.device) + random_tensor.floor_() # binarize + output = input.div(keep_prob) * random_tensor + return output + -# https://github.com/huggingface/pytorch-image-models/blob/main/timm/layers/drop.py -class DropPath(nn.Module): - """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).""" +# Copied from transformers.models.beit.modeling_beit.BeitDropPath with Beit->Florence2 +class Florence2DropPath(nn.Module): + """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).""" - def __init__(self, drop_prob: float = 0.0, scale_by_keep: bool = True): - super(DropPath, self).__init__() + def __init__(self, drop_prob: Optional[float] = None) -> None: + super().__init__() self.drop_prob = drop_prob - self.scale_by_keep = scale_by_keep - def forward(self, x): - if self.drop_prob == 0.0 or not self.training: - return x - keep_prob = 1 - self.drop_prob - shape = (x.shape[0],) + (1,) * (x.ndim - 1) # work with diff dim tensors, not just 2D ConvNets - random_tensor = x.new_empty(shape).bernoulli_(keep_prob) - if keep_prob > 0.0 and self.scale_by_keep: - random_tensor.div_(keep_prob) - return x * random_tensor + def forward(self, hidden_states: torch.Tensor) -> torch.Tensor: + return drop_path(hidden_states, self.drop_prob, self.training) - def extra_repr(self): - return f"drop_prob={round(self.drop_prob,3):0.3f}" + def extra_repr(self) -> str: + return "p={}".format(self.drop_prob) # TODO: can this be replaced with torch.nn.init.trunc_normal_? @@ -426,7 +439,7 @@ def __init__( ): super().__init__() - drop_path = DropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() + drop_path = Florence2DropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() self.conv1 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None self.channel_attn = PreNorm( From 689ad98c1ea55cbe48e9d0c2caa551628c10fb30 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Tue, 25 Jun 2024 22:42:08 +0200 Subject: [PATCH 30/35] update Florence2DropPath --- .../models/florence2/modeling_florence2.py | 37 +++++++++---------- 1 file changed, 18 insertions(+), 19 deletions(-) diff --git a/src/transformers/models/florence2/modeling_florence2.py b/src/transformers/models/florence2/modeling_florence2.py index e70cee3ea3cb..f19b21fe248a 100644 --- a/src/transformers/models/florence2/modeling_florence2.py +++ b/src/transformers/models/florence2/modeling_florence2.py @@ -328,33 +328,32 @@ def forward(self, x, size): x = self.proj(x) return x, size -# TODO verify behavior is same as https://github.com/huggingface/pytorch-image-models/blob/main/timm/layers/drop.py#L150 -# Copied from transformers.models.beit.modeling_beit.drop_path -def drop_path(input: torch.Tensor, drop_prob: float = 0.0, training: bool = False) -> torch.Tensor: - """ - Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). +# https://github.com/huggingface/pytorch-image-models/blob/main/timm/layers/drop.py#L150 +def drop_path(x, drop_prob: float = 0., training: bool = False, scale_by_keep: bool = True): + """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). + + This is the same as the DropConnect impl I created for EfficientNet, etc networks, however, + the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper... + See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for + changing the layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use + 'survival rate' as the argument. - Comment by Ross Wightman: This is the same as the DropConnect impl I created for EfficientNet, etc networks, - however, the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper... - See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for changing the - layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use 'survival rate' as the - argument. """ - if drop_prob == 0.0 or not training: - return input + if drop_prob == 0. or not training: + return x keep_prob = 1 - drop_prob - shape = (input.shape[0],) + (1,) * (input.ndim - 1) # work with diff dim tensors, not just 2D ConvNets - random_tensor = keep_prob + torch.rand(shape, dtype=input.dtype, device=input.device) - random_tensor.floor_() # binarize - output = input.div(keep_prob) * random_tensor - return output + shape = (x.shape[0],) + (1,) * (x.ndim - 1) # work with diff dim tensors, not just 2D ConvNets + random_tensor = x.new_empty(shape).bernoulli_(keep_prob) + if keep_prob > 0.0 and scale_by_keep: + random_tensor.div_(keep_prob) + return x * random_tensor # Copied from transformers.models.beit.modeling_beit.BeitDropPath with Beit->Florence2 class Florence2DropPath(nn.Module): """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).""" - def __init__(self, drop_prob: Optional[float] = None) -> None: + def __init__(self, drop_prob: float = 0.) -> None: super().__init__() self.drop_prob = drop_prob @@ -550,7 +549,7 @@ def __init__( ): super().__init__() - drop_path = DropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() + drop_path = Florence2DropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() self.conv1 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None self.window_attn = PreNorm( From 6b7feb6ac27d7f2bdfb625d15334995a6669e5cb Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Tue, 25 Jun 2024 22:48:58 +0200 Subject: [PATCH 31/35] update Florence2Sequential --- src/transformers/models/florence2/modeling_florence2.py | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/src/transformers/models/florence2/modeling_florence2.py b/src/transformers/models/florence2/modeling_florence2.py index f19b21fe248a..d724a592430b 100644 --- a/src/transformers/models/florence2/modeling_florence2.py +++ b/src/transformers/models/florence2/modeling_florence2.py @@ -190,7 +190,7 @@ def forward(self, seq_embeds: torch.Tensor) -> torch.Tensor: return pos_embeds -class MySequential(nn.Sequential): +class Florence2Sequential(nn.Sequential): def forward(self, *inputs): for module in self._modules.values(): if type(inputs) == tuple: @@ -364,7 +364,6 @@ def extra_repr(self) -> str: return "p={}".format(self.drop_prob) -# TODO: can this be replaced with torch.nn.init.trunc_normal_? # https://github.com/huggingface/pytorch-image-models/blob/b28945ff056d454b174f0fb8682e362b87150141/timm/layers/weight_init.py def trunc_normal_(tensor: torch.Tensor, mean=0.0, std=1.0, a=-2.0, b=2.0) -> torch.Tensor: r"""Fills the input Tensor with values drawn from a truncated @@ -643,9 +642,9 @@ def __init__( ) convs.append(conv_embed) - block = MySequential( + block = Florence2Sequential( *[ - MySequential( + Florence2Sequential( OrderedDict( [ ( From 3addc028a2e97bc6d5616119ba82cddd34efdbd2 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Tue, 25 Jun 2024 23:07:28 +0200 Subject: [PATCH 32/35] update modeling_florence2, add model-prefix --- .../models/florence2/modeling_florence2.py | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/src/transformers/models/florence2/modeling_florence2.py b/src/transformers/models/florence2/modeling_florence2.py index d724a592430b..09d2a21111f8 100644 --- a/src/transformers/models/florence2/modeling_florence2.py +++ b/src/transformers/models/florence2/modeling_florence2.py @@ -66,7 +66,7 @@ _CONFIG_FOR_DOC = "Florence2Config" -class LearnedAbsolutePositionEmbedding2D(nn.Module): +class Florence2LearnedAbsolutePositionEmbedding2D(nn.Module): """ This module learns positional embeddings up to a fixed maximum size. """ @@ -100,7 +100,7 @@ def forward(self, pixel_values): return pos -class PositionalEmbeddingCosine1D(nn.Module): +class Florence2PositionalEmbeddingCosine1D(nn.Module): """ This class implements a very simple positional encoding. It follows closely the encoder from the link below: @@ -113,7 +113,7 @@ class PositionalEmbeddingCosine1D(nn.Module): """ def __init__(self, embed_dim: int = 512, max_seq_len: int = 1024) -> None: - super(PositionalEmbeddingCosine1D, self).__init__() + super(Florence2PositionalEmbeddingCosine1D, self).__init__() self.embed_dim = embed_dim self.max_seq_len = max_seq_len # Generate the sinusoidal arrays. @@ -152,7 +152,7 @@ def forward(self, seq_embeds: torch.Tensor) -> torch.Tensor: return pos_embeds -class LearnedAbsolutePositionEmbedding1D(nn.Module): +class Florence2LearnedAbsolutePositionEmbedding1D(nn.Module): """ Learnable absolute positional embeddings for 1D sequences. @@ -162,7 +162,7 @@ class LearnedAbsolutePositionEmbedding1D(nn.Module): """ def __init__(self, embedding_dim: int = 512, num_pos: int = 1024) -> None: - super(LearnedAbsolutePositionEmbedding1D, self).__init__() + super(Florence2LearnedAbsolutePositionEmbedding1D, self).__init__() self.embeddings = nn.Embedding(num_pos, embedding_dim) self.num_pos = num_pos @@ -2528,7 +2528,7 @@ def _build_image_projection_layers(self, config): self.image_proj_norm = nn.LayerNorm(dim_projection) image_pos_embed_config = config.image_pos_embed if image_pos_embed_config["type"] == "learned_abs_2d": - self.image_pos_embed = LearnedAbsolutePositionEmbedding2D( + self.image_pos_embed = Florence2LearnedAbsolutePositionEmbedding2D( embedding_dim=image_dim_out, num_pos=image_pos_embed_config["max_pos_embeddings"] ) else: @@ -2539,7 +2539,7 @@ def _build_image_projection_layers(self, config): # temporal embedding visual_temporal_embedding_config = config.visual_temporal_embedding if visual_temporal_embedding_config["type"] == "COSINE": - self.visual_temporal_embed = PositionalEmbeddingCosine1D( + self.visual_temporal_embed = Florence2PositionalEmbeddingCosine1D( embed_dim=image_dim_out, max_seq_len=visual_temporal_embedding_config["max_temporal_embeddings"] ) else: @@ -2626,7 +2626,7 @@ def _build_image_projection_layers(self, config): self.image_proj_norm = nn.LayerNorm(dim_projection) image_pos_embed_config = config.vision_config.image_pos_embed if image_pos_embed_config["type"] == "learned_abs_2d": - self.image_pos_embed = LearnedAbsolutePositionEmbedding2D( + self.image_pos_embed = Florence2LearnedAbsolutePositionEmbedding2D( embedding_dim=image_dim_out, num_pos=image_pos_embed_config["max_pos_embeddings"] ) else: @@ -2637,7 +2637,7 @@ def _build_image_projection_layers(self, config): # temporal embedding visual_temporal_embedding_config = config.vision_config.visual_temporal_embedding if visual_temporal_embedding_config["type"] == "COSINE": - self.visual_temporal_embed = PositionalEmbeddingCosine1D( + self.visual_temporal_embed = Florence2PositionalEmbeddingCosine1D( embed_dim=image_dim_out, max_seq_len=visual_temporal_embedding_config["max_temporal_embeddings"] ) else: From 0f624acb955a5641776d0e23dd12f93f5d21bfce Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Wed, 26 Jun 2024 00:05:17 +0200 Subject: [PATCH 33/35] update modeling_florence2, add DaViT prefixes --- .../models/florence2/modeling_florence2.py | 73 +++++++++---------- 1 file changed, 34 insertions(+), 39 deletions(-) diff --git a/src/transformers/models/florence2/modeling_florence2.py b/src/transformers/models/florence2/modeling_florence2.py index 09d2a21111f8..700d1d5ac2de 100644 --- a/src/transformers/models/florence2/modeling_florence2.py +++ b/src/transformers/models/florence2/modeling_florence2.py @@ -59,6 +59,7 @@ if is_flash_attn_2_available(): + from flash_attn import flash_attn_func, flash_attn_varlen_func from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input # noqa logger = logging.get_logger(__name__) @@ -190,7 +191,7 @@ def forward(self, seq_embeds: torch.Tensor) -> torch.Tensor: return pos_embeds -class Florence2Sequential(nn.Sequential): +class Florence2DaViTSequential(nn.Sequential): def forward(self, *inputs): for module in self._modules.values(): if type(inputs) == tuple: @@ -200,7 +201,7 @@ def forward(self, *inputs): return inputs -class PreNorm(nn.Module): +class Florence2DaViTPreNorm(nn.Module): def __init__(self, norm, fn, drop_path=None): super().__init__() self.norm = norm @@ -222,7 +223,7 @@ def forward(self, x, *args, **kwargs): return x, size -class Mlp(nn.Module): +class Florence2DaViTMlp(nn.Module): def __init__( self, in_features, @@ -247,7 +248,7 @@ def forward(self, x, size): return self.net(x), size -class DepthWiseConv2d(nn.Module): +class Florence2DaViTDepthWiseConv2d(nn.Module): def __init__( self, dim_in, @@ -272,7 +273,7 @@ def forward(self, x, size): return x, size -class ConvEmbed(nn.Module): +class Florence2DaViTConvEmbed(nn.Module): """Image to Patch Embedding""" def __init__(self, patch_size=7, in_chans=3, embed_dim=64, stride=4, padding=2, norm_layer=None, pre_norm=True): @@ -306,7 +307,7 @@ def forward(self, x, size): return x, (H, W) -class ChannelAttention(nn.Module): +class Florence2DaViTChannelAttention(nn.Module): def __init__(self, dim, groups=8, qkv_bias=True): super().__init__() @@ -349,8 +350,7 @@ def drop_path(x, drop_prob: float = 0., training: bool = False, scale_by_keep: b return x * random_tensor -# Copied from transformers.models.beit.modeling_beit.BeitDropPath with Beit->Florence2 -class Florence2DropPath(nn.Module): +class Florence2DaViTDropPath(nn.Module): """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).""" def __init__(self, drop_prob: float = 0.) -> None: @@ -422,7 +422,7 @@ def norm_cdf(x): return tensor -class ChannelBlock(nn.Module): +class Florence2DaViTChannelBlock(nn.Module): def __init__( self, dim, @@ -437,15 +437,15 @@ def __init__( ): super().__init__() - drop_path = Florence2DropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() + drop_path = Florence2DaViTDropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() - self.conv1 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None - self.channel_attn = PreNorm( - norm_layer(dim), ChannelAttention(dim, groups=groups, qkv_bias=qkv_bias), drop_path + self.conv1 = Florence2DaViTPreNorm(None, Florence2DaViTDepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None + self.channel_attn = Florence2DaViTPreNorm( + norm_layer(dim), Florence2DaViTChannelAttention(dim, groups=groups, qkv_bias=qkv_bias), drop_path ) - self.conv2 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_ffn else None - self.ffn = PreNorm( - norm_layer(dim), Mlp(in_features=dim, hidden_features=int(dim * mlp_ratio), act_layer=act_layer), drop_path + self.conv2 = Florence2DaViTPreNorm(None, Florence2DaViTDepthWiseConv2d(dim, 3, 1, 1)) if conv_at_ffn else None + self.ffn = Florence2DaViTPreNorm( + norm_layer(dim), Florence2DaViTMlp(in_features=dim, hidden_features=int(dim * mlp_ratio), act_layer=act_layer), drop_path ) def forward(self, x, size): @@ -476,7 +476,7 @@ def window_reverse(windows, batch_size: int, window_size: int, H: int, W: int): return x -class WindowAttention(nn.Module): +class Florence2DaViTWindowAttention(nn.Module): def __init__(self, dim, num_heads, window_size, qkv_bias=True): super().__init__() self.dim = dim @@ -532,7 +532,7 @@ def forward(self, x, size): return x, size -class SpatialBlock(nn.Module): +class Florence2DaViTSpatialBlock(nn.Module): def __init__( self, dim, @@ -548,15 +548,15 @@ def __init__( ): super().__init__() - drop_path = Florence2DropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() + drop_path = Florence2DaViTDropPath(drop_path_rate) if drop_path_rate > 0.0 else nn.Identity() - self.conv1 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None - self.window_attn = PreNorm( - norm_layer(dim), WindowAttention(dim, num_heads, window_size, qkv_bias=qkv_bias), drop_path + self.conv1 = Florence2DaViTPreNorm(None, Florence2DaViTDepthWiseConv2d(dim, 3, 1, 1)) if conv_at_attn else None + self.window_attn = Florence2DaViTPreNorm( + norm_layer(dim), Florence2DaViTWindowAttention(dim, num_heads, window_size, qkv_bias=qkv_bias), drop_path ) - self.conv2 = PreNorm(None, DepthWiseConv2d(dim, 3, 1, 1)) if conv_at_ffn else None - self.ffn = PreNorm( - norm_layer(dim), Mlp(in_features=dim, hidden_features=int(dim * mlp_ratio), act_layer=act_layer), drop_path + self.conv2 = Florence2DaViTPreNorm(None, Florence2DaViTDepthWiseConv2d(dim, 3, 1, 1)) if conv_at_ffn else None + self.ffn = Florence2DaViTPreNorm( + norm_layer(dim), Florence2DaViTMlp(in_features=dim, hidden_features=int(dim * mlp_ratio), act_layer=act_layer), drop_path ) def forward(self, x, size): @@ -570,7 +570,7 @@ def forward(self, x, size): return x, size -class DaViT(nn.Module): +class Florence2DaViT(nn.Module): """DaViT: Dual-Attention Transformer Args: @@ -631,7 +631,7 @@ def __init__( convs = [] blocks = [] for i in range(num_stages): - conv_embed = ConvEmbed( + conv_embed = Florence2DaViTConvEmbed( patch_size=patch_size[i], stride=patch_stride[i], padding=patch_padding[i], @@ -642,14 +642,14 @@ def __init__( ) convs.append(conv_embed) - block = Florence2Sequential( + block = Florence2DaViTSequential( *[ - Florence2Sequential( + Florence2DaViTSequential( OrderedDict( [ ( "spatial_block", - SpatialBlock( + Florence2DaViTSpatialBlock( embed_dims[i], num_heads[i], window_size, @@ -662,7 +662,7 @@ def __init__( ), ( "channel_block", - ChannelBlock( + Florence2DaViTChannelBlock( embed_dims[i], num_groups[i], drop_path_rate=dpr[depth_offset + j * 2 + 1], @@ -758,11 +758,6 @@ def from_config(cls, config): ) -if is_flash_attn_2_available(): - from flash_attn import flash_attn_func, flash_attn_varlen_func - from flash_attn.bert_padding import index_first_axis, pad_input, unpad_input # noqa - - # Copied from transformers.models.llama.modeling_llama._get_unpad_data def _get_unpad_data(attention_mask): seqlens_in_batch = attention_mask.sum(dim=-1, dtype=torch.int32) @@ -2495,7 +2490,7 @@ class Florence2VisionModel(Florence2PreTrainedModel): def __init__(self, config: Florence2VisionConfig): super().__init__(config) assert config.model_type == "davit", "only DaViT is supported for now" - self.vision_tower = DaViT.from_config(config=config) + self.vision_tower = Florence2DaViT.from_config(config=config) self.post_init() @@ -2515,7 +2510,7 @@ class Florence2VisionModelWithProjection(Florence2PreTrainedModel): def __init__(self, config: Florence2VisionConfig): super().__init__(config) assert config.model_type == "davit", "only DaViT is supported for now" - self.vision_tower = DaViT.from_config(config=config) + self.vision_tower = Florence2DaViT.from_config(config=config) self._build_image_projection_layers(config) @@ -2601,7 +2596,7 @@ def __init__(self, config: Florence2Config): super().__init__(config) assert config.vision_config.model_type == "davit", "only DaViT is supported for now" del config.vision_config.model_type - self.vision_tower = DaViT.from_config(config=config.vision_config) + self.vision_tower = Florence2DaViT.from_config(config=config.vision_config) # remove unused layers del self.vision_tower.head del self.vision_tower.norms From 880811f2f50de704508f8c9451c43c795d40f0a3 Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Wed, 26 Jun 2024 01:27:00 +0200 Subject: [PATCH 34/35] fix base_model_prefix --- src/transformers/models/florence2/modeling_florence2.py | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/src/transformers/models/florence2/modeling_florence2.py b/src/transformers/models/florence2/modeling_florence2.py index 700d1d5ac2de..58035118ba71 100644 --- a/src/transformers/models/florence2/modeling_florence2.py +++ b/src/transformers/models/florence2/modeling_florence2.py @@ -1504,7 +1504,7 @@ def forward( class Florence2LanguagePreTrainedModel(PreTrainedModel): config_class = Florence2LanguageConfig - base_model_prefix = "model" + base_model_prefix = "florence2" supports_gradient_checkpointing = True _keys_to_ignore_on_load_unexpected = ["encoder.version", "decoder.version"] _no_split_modules = [r"Florence2EncoderLayer", r"Florence2DecoderLayer"] @@ -2132,7 +2132,6 @@ def forward( class Florence2LanguageForConditionalGeneration(Florence2LanguagePreTrainedModel): - base_model_prefix = "model" _tied_weights_keys = ["encoder.embed_tokens.weight", "decoder.embed_tokens.weight", "lm_head.weight"] _keys_to_ignore_on_load_missing = ["final_logits_bias"] @@ -2395,7 +2394,7 @@ class Florence2Seq2SeqLMOutput(ModelOutput): ) class Florence2PreTrainedModel(PreTrainedModel): config_class = Florence2Config - base_model_prefix = "model" + base_model_prefix = "florence2" supports_gradient_checkpointing = True _skip_keys_device_placement = "past_key_values" From 4069145e4fb79764618ec89d5043bf282a168e9b Mon Sep 17 00:00:00 2001 From: Dave <69651599+D4ve-R@users.noreply.github.com> Date: Wed, 26 Jun 2024 16:26:36 +0200 Subject: [PATCH 35/35] fix missing variables --- .../models/florence2/modeling_florence2.py | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/src/transformers/models/florence2/modeling_florence2.py b/src/transformers/models/florence2/modeling_florence2.py index 58035118ba71..ce91d2050384 100644 --- a/src/transformers/models/florence2/modeling_florence2.py +++ b/src/transformers/models/florence2/modeling_florence2.py @@ -2312,6 +2312,10 @@ class Florence2Seq2SeqLMOutput(ModelOutput): decoding. Args: + loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided): + Language modeling loss. + 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). last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`): Sequence of hidden-states at the output of the last layer of the decoder of the model. @@ -2360,7 +2364,8 @@ class Florence2Seq2SeqLMOutput(ModelOutput): image_hidden_states of the model produced by the vision encoder """ - + loss: Optional[torch.FloatTensor] = None + logits: torch.FloatTensor = None last_hidden_state: torch.FloatTensor = None past_key_values: Optional[Tuple[Tuple[torch.FloatTensor]]] = None decoder_hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None @@ -2369,6 +2374,7 @@ class Florence2Seq2SeqLMOutput(ModelOutput): encoder_last_hidden_state: Optional[torch.FloatTensor] = None encoder_hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None encoder_attentions: Optional[Tuple[torch.FloatTensor, ...]] = None + image_hidden_states: Optional[Tuple[torch.FloatTensor, ...]] = None FLORENCE2_START_DOCSTRING = r""" @@ -2594,7 +2600,6 @@ class Florence2ForConditionalGeneration(Florence2PreTrainedModel): def __init__(self, config: Florence2Config): super().__init__(config) assert config.vision_config.model_type == "davit", "only DaViT is supported for now" - del config.vision_config.model_type self.vision_tower = Florence2DaViT.from_config(config=config.vision_config) # remove unused layers del self.vision_tower.head @@ -2793,7 +2798,8 @@ def forward( image_features, inputs_embeds ) - attention_mask = attention_mask.to(inputs_embeds.dtype) + if inputs_embeds is not None: + attention_mask = attention_mask.to(inputs_embeds.dtype) outputs = self.language_model( attention_mask=attention_mask, labels=labels,