sg? FdZddlmZmZmZddlZddlmZddl m Z ddl m Z m Z mZmZddlmZmZd d lmZGd d e d ZGdde d ZdeededeeefdZdeeeedeeedededej2f dZdedededefdZGddeZy) zProcessor class for Mllama.)ListOptionalUnionN) BatchFeature) ImageInput) ImagesKwargsProcessingKwargsProcessorMixinUnpack)PreTokenizedInput TextInput)make_list_of_imagesceZdZUeeed<y)MllamaImagesKwargsmax_image_tilesN)__name__ __module__ __qualname__rint__annotations___/var/www/html/venv/lib/python3.12/site-packages/transformers/models/mllama/processing_mllama.pyrr"s c]"rrF)totalc$eZdZUeed<dddiiZy)MllamaProcessorKwargs images_kwargs image_kwargsrN)rrrrr _defaultsrrrrr&s%%  q Irr input_idsimage_token_idreturnct|Dcgc] \}}||k(s |}}}t|dk(rgSt|dk(r|ddggSt|dd|ddDcgc] \}}||g }}}|j|dt|g|dd}|dddD]} | d| ddz k(r|| d<| d}|Scc}}wcc}}w)a Generate a cross-attention token mask for image tokens in the input sequence. This function identifies the positions of image tokens in the input sequence and creates a mask that defines which subsequent tokens each image token should attend to. Args: input_ids (List[int]): A list of token ids representing the input sequence. image_token_id (int): The id of the token used to represent images in the sequence. Returns: List[List[int]]: A list of [start, end] pairs, where each pair represents the range of tokens an image token should attend to. Notes: - If no image tokens are present, an empty list is returned. - For a single image token, it attends to all subsequent tokens until the end of the sequence. - For multiple image tokens, each attends to tokens up to the next image token or the end of the sequence. - Consecutive image tokens are treated as a group and attend to all subsequent tokens together. rrN) enumeratelenzipappend) r#r$itokenimage_token_locationsloc1loc2 vision_masks last_mask_end vision_masks rget_cross_attention_token_maskr40s,09/C_81euP^G^Q__ !Q&  !Q&&q)2.//367LSb7QShijikSl3mnZT4T4LnLn.r2C NCD !$Q'M#DbD)' q>[^a/ /*KN#A ' /`os B>B>Ccross_attention_token_mask num_tiles max_num_tileslengthc t|}t|Dcgc] }t|c}}tj||||ftj}t t ||D]\\}\} } t t | | D]<\} \} } t| dk(s| \}}t||}|dk(r|}d||||| d| f<>^|Scc}w)a Convert the cross attention mask indices to a cross attention mask 4D array. This function takes a sparse representation of cross attention masks and converts it to a dense 4D numpy array. The sparse representation is a nested list structure that defines attention ranges for each image in each batch item. Args: cross_attention_token_mask (List[List[List[int]]]): A nested list structure where: - The outer list represents the batch dimension. - The middle list represents different images within each batch item. - The inner list contains pairs of integers [start, end] representing token ranges for each image. num_tiles (List[List[int]]): A nested list structure specifying the number of tiles for each image in each batch item. max_num_tiles (int): The maximum possible number of tiles. length (int): The total sequence length of the input. Returns: np.ndarray: A 4D numpy array of shape (batch_size, length, max_num_images, max_num_tiles) The array contains `1` where attention is allowed and `0` where it is not. Note: - Special handling is done for cases where the end token is -1, which is interpreted as attending to the end of the sequence. )shapedtyper'rN)r)maxnpzerosint64r(r*min)r5r6r7r8 batch_sizemasksmax_num_imagescross_attention_mask sample_idx sample_maskssample_num_tilesmask_idx locationsmask_num_tilesstartends r,convert_sparse_cross_attention_mask_to_denserN`s:/0J2LM#e*MNN886>=Ahh 9B#F`bkBl8m[4 4\#35>s>> build_string_from_input("Hello world", "", "<|image|>") 'Hello world' >>> build_string_from_input("<|image|>Hello world", "", "<|image|>") '<|image|>Hello world' >>> build_string_from_input("Hello world", "", "<|image|>") 'Hello world' rNr) startswithr))rOrPrQnum_image_tokens_on_starts rbuild_string_from_inputrUsn4F !   K (K(*+!Q&!   K (55 6yk& JJrc eZdZdZddgZdZdZfdZ ddee dee e e e e e e fd eed efd Zd Zd ZdZedZxZS)MllamaProcessoraw Constructs a Mllama processor which wraps [`MllamaImageProcessor`] and [`PretrainedTokenizerFast`] into a single processor that inherits both the image processor and tokenizer functionalities. See the [`~MllamaProcessor.__call__`] and [`~OwlViTProcessor.decode`] for more information. The preferred way of passing kwargs is as a dictionary per modality, see usage example below. ```python from transformers import MllamaProcessor from PIL import Image processor = MllamaProcessor.from_pretrained("meta-llama/Llama-3.2-11B-Vision") processor( images=your_pil_image, text=["<|image|>If I had to write a haiku for this one"], images_kwargs = {"size": {"height": 448, "width": 448}}, text_kwargs = {"padding": "right"}, common_kwargs = {"return_tensors": "pt"}, ) ``` Args: image_processor ([`MllamaImageProcessor`]): The image processor is a required input. tokenizer ([`PreTrainedTokenizer`, `PreTrainedTokenizerFast`]): The tokenizer is a required input. image_processor tokenizerMllamaImageProcessorPreTrainedTokenizerFastcdt|ds(d|_|j|j|_n"|j|_|j|_d|_|j|j|_|j |_|j|_t|%||y)NrQz <|image|>z<|python_tag|>) hasattrrQconvert_tokens_to_idsr$ python_tokenpython_token_idrP chat_templatesuper__init__)selfrXrY __class__s rrczMllamaProcessor.__init__sy-0*D "+"A"A$BRBR"SD (44D "+":":D ,(>>t?P?PQ",,&44 )4rimagestextkwargsr%c *| | td|jtfd|jji|}|d}|d}|d} i} |t |t r|g}n3t |ttfrtd|Ds td|D cgc]} | j|j} } |D cgc]#} t| |j|j%}} |jd d}|j|fi|}| j|d g}|#t!|}|Dcgc] }t#|}}|wt%d  Drtd | Ds td t'|t'| k7r1| tdtdt'| dt'|d|5|j(|fi|}|jd}| j||c|adDcgc]}t+||j,}}t/||j(j0t3d|dD}|| d<| jdd}t5| |}|Scc} wcc} wcc}wcc}w)a& Main method to prepare text(s) and image(s) to be fed as input to the model. This method forwards the `text` arguments to PreTrainedTokenizerFast's [`~PreTrainedTokenizerFast.__call__`] if `text` is not `None` to encode the text. To prepare the image(s), this method forwards the `images` arguments to MllamaImageProcessor's [`~MllamaImageProcessor.__call__`] if `images` is not `None`. Please refer to the docstring of the above two methods for more information. Args: 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. Both channels-first and channels-last formats are supported. 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). 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`. - **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`. TODO: add aspect_ratio_ids and aspect_ratio_mask and cross_attention_mask Nz'You must specify either text or images.tokenizer_init_kwargs text_kwargsr common_kwargsc3<K|]}t|tywN) isinstancestr).0ts r z+MllamaProcessor.__call__..s=_UVjC>P=_szAInvalid input text. Please provide a string, or a list of strings padding_siderc3&K|] }|dk( ywrNrrq batch_imgs rrsz+MllamaProcessor.__call__..+sDi9>Dc3&K|] }|dk( ywrvrrws rrsz+MllamaProcessor.__call__..+sQ#, QQryzaIf a batch of text is provided, there should be either no images or at least one image per samplez@No image were provided, but there are image tokens in the promptzThe number of image token (z:) should be the same as in the number of provided images ()r6r#c32K|]}t|ywrn)r))rqr#s rrsz+MllamaProcessor.__call__..GsQi3y>Qs)r6r7r8rEreturn_tensors)data tensor_type) ValueError _merge_kwargsrrY init_kwargsrorplisttupleallcountrQrUrPpopupdaterr)anysumrXr4r$rNrr=r)rdrfrgaudiovideosrh output_kwargsrkrrlr~rrn_images_in_text text_item_encodingn_images_in_imagessampleimage_featuresr6 token_idsr5rEr} batch_features r__call__zMllamaProcessor.__call__sN *??>$%ghh$5c:J6K5LMGHKL^H_G``ab  1T11&JMJN&**;7I KK '  $"2`hit`u*S\.y$:M:MN* &*$P*#"22BBQ8K;PQQ $ ,@D' (&**+;TB$$NK _ Io"D0*s"J(J7J Jc:|jj|i|S)z This method forwards all its arguments to PreTrainedTokenizerFast's [`~PreTrainedTokenizer.batch_decode`]. Please refer to the docstring of this method for more information. rY batch_decoderdargsrhs rrzMllamaProcessor.batch_decodePs +t~~**D;F;;rc:|jj|i|S)z This method forwards all its arguments to PreTrainedTokenizerFast's [`~PreTrainedTokenizer.decode`]. Please refer to the docstring of this method for more information. )rYdecoders rrzMllamaProcessor.decodeWs %t~~$$d5f55rc>|jj|ddS)a Post-process the output of the model to decode the text. Args: generated_outputs (`torch.Tensor` or `np.ndarray`): The output of the model `generate` function. The output is expected to be a tensor of shape `(batch_size, sequence_length)` or `(sequence_length,)`. Returns: `List[str]`: The decoded text. TF)skip_special_tokensclean_up_tokenization_spacesr)rdgenerated_outputss rpost_process_image_text_to_textz/MllamaProcessor.post_process_image_text_to_text^s(~~** 4V[+  rc~|jj}|jj}t||zdgzS)NrE)rYmodel_input_namesrXr)rdtokenizer_input_namesimage_processor_input_namess rrz!MllamaProcessor.model_input_namesns? $ @ @&*&:&:&L&L#),GGKaJbbccr)NNNN)rrr__doc__ attributesimage_processor_classtokenizer_classrcrrrrr rr rrrrrrpropertyr __classcell__)res@rrWrWs:$[1J2/O 5 (,hl i$iuY(94 ?DQbLccdei ./ i iV<6 ddrrW)rtypingrrrnumpyr>feature_extraction_utilsr image_utilsrprocessing_utilsr r r r tokenization_utils_baser rimage_processing_mllamarrrrr4ndarrayrNrprUrWrrrrs "((4%VV 9#U#,E-d3i--QUVZ[^V_Q`-`- $T$s)_ 5- DI- -   - ZZ - `"KC"KC"Kc"Kc"KJ}dn}dr