a h;@sddlmZmZddlZddlmZddlmZddl m Z m Z m Z m Z mZmZddlmZmZddlmZdd lmZeeZGd d d ed d ZGddde ZGddde d d ZGddde ZdgZdS))OptionalUnionN) BatchFeature) ImageInput) ImagesKwargsMultiModalDataProcessingKwargsProcessorMixinUnpack VideosKwargs)PreTokenizedInput TextInput)logging) VideoInputc@s"eZdZUeeeefed<dS)Glm4vVideosProcessorKwargsfpsN)__name__ __module__ __qualname__rlistfloat__annotations__rrf/var/www/html/assistant/venv/lib/python3.9/site-packages/transformers/models/glm4v/processing_glm4v.pyr$s rF)totalc@s2eZdZUeeed<eeed<eeed<dS)Glm4vImagesKwargsZ patch_sizeZtemporal_patch_size merge_sizeN)rrrrintrrrrrr(s   rc@s2eZdZUeed<dddddidZeed<dS) Glm4vProcessorKwargs images_kwargsF)paddingreturn_mm_token_type_idsreturn_metadataT) text_kwargs videos_kwargsr%N)rrrrr _defaultsrrrrrr.s rcszeZdZdZgdZdZdZdZdfdd Zde e e e e e e e feeeed d d Zdd d ZdddZZS)Glm4vProcessora Constructs a GLM-4V processor which wraps a GLM-4V image processor and a GLM-4 tokenizer into a single processor. [`~Glm4vProcessor.__call__`] and [`~Glm4vProcessor.decode`] for more information. Args: image_processor ([`Glm4vProcessor`], *optional*): The image processor is a required input. tokenizer ([`PreTrainedTokenizerFast`], *optional*): The tokenizer is a required input. video_processor ([`Glm4vVideoProcessor`], *optional*): The video processor is a required input. chat_template (`str`, *optional*): A Jinja template which will be used to convert lists of messages in a chat into a tokenizable string. )image_processor tokenizervideo_processorZAutoImageProcessorZAutoVideoProcessor)ZPreTrainedTokenizerZPreTrainedTokenizerFastNc stj||||dt|ds"dn|j|_t|ds8dn|j|_t|ddrR|jn ||j|_t|ddrr|jn ||j|_dS)N) chat_template image_tokenz <|image|> video_tokenz <|video|>image_token_idvideo_token_id) super__init__hasattrr,r-getattrr.Zconvert_tokens_to_idsr/)selfr(r)r*r+kwargs __class__rrr1Os    zGlm4vProcessor.__init__)imagestextvideosr5returncKs|jtfd|jji|}|durD|jfd|i|d}|d}ni}d}|dur|jfd|i|d}d|vr|d } n|d } |d } ni}d} t|ts|g}| }|dur:|jj d } d } t t |D]`} |j || vr"|| | }|| |j d |d|| <| d7} q|| d |j || <q| dur|jj d } d }t t |D]} |j|| vr| |d }d}| | }|jdurtd|jdurdn|j|_|jddd }g}t d t |D]}|||q|d|}t ||kr*||r |dnd qt |D],}||}d|j dt|}||7}q2|| |j|d|| <| || | |d }t |D]2}|j || vr|| |j d |d|| <q|d7}qf|| d |j || <q`|ddd}|ddd}|j|fi|d}|j||ddgd|r|t|d}t|d}d|||jk<||d<ti||||dS)a^ Main method to prepare for the model one or several sequences(s) and image(s). This method forwards the `text` and `kwargs` arguments to PreTrainedTokenizerFast's [`~PreTrainedTokenizerFast.__call__`] if `text` is not `None` to encode the text. 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). videos (`np.ndarray`, `torch.Tensor`, `List[np.ndarray]`, `List[torch.Tensor]`): The image or batch of videos to be prepared. Each video can be a 4D NumPy array or PyTorch tensor, or a nested list of 3D frames. Both channels-first and channels-last formats are supported. 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`. - **pixel_values_videos** -- Pixel values of videos to be fed to a model. Returned when `videos` is not `None`. - **image_grid_thw** -- List of image 3D grid in LLM. Returned when `images` is not `None`. - **video_grid_thw** -- List of video 3D grid in LLM. Returned when `videos` is not `None`. Ztokenizer_init_kwargsNr8r image_grid_thwr:r%r#video_metadatavideo_grid_thwrz<|placeholder|>aSmolVLM requires frame timestamps to construct prompts, but the `fps` of the input video could not be inferred. Probably `video_metadata` was missing from inputs and you passed pre-sampled frames. Defaulting to `fps=24`. Please provide `video_metadata` for more accurate results.z<|begin_of_image|>z<|end_of_image|>r$return_tensorsr"FimageZvideo)Z modalitiesZ input_idsmm_token_type_ids)dataZ tensor_type)Z _merge_kwargsrr)Z init_kwargsr(r*pop isinstancercopyrrangelenr,prodreplacer-rloggerZ warning_once timestampsappendrZ_check_special_mm_tokensnparrayZ zeros_liker.tolistr)r4r8r9r:r5Z output_kwargsZ image_inputsr<Z videos_inputsr=r>Z merge_lengthindexinum_image_tokensZ video_indexZ num_framesZvideo_structuremetadatarPZunique_timestampsidxZselected_timestampsZ frame_idxZ timestamp_secZframe_structurerDr"Z text_inputsZ array_idsrFrrr__call__^s*                  zGlm4vProcessor.__call__c  si}|durntjdi|ddp6jjfdd|D}fdd|D}|||d|durtjdi|fd d|D}fd d|D}||d <tfi|S) aK Computes the number of placeholder tokens needed for multimodal inputs with the given sizes. Args: image_sizes (`list[list[int]]`, *optional*): The input sizes formatted as (height, width) per each image. video_sizes (`list[list[int]]`, *optional*): The input sizes formatted as (num_frames, height, width) per each video. Returns: `MultiModalData`: A `MultiModalData` object holding number of tokens per each of the provided input modalities, along with other useful data. Nr rcs"g|]}jjg|RqSr)r(Zget_number_of_image_patches).0Z image_size)r r4rr sz=Glm4vProcessor._get_num_multimodal_tokens..csg|]}|dqSr?rr[Z num_patchesrrrr\)rWnum_image_patchesr%cs"g|]}jjg|RqSr)r*Zget_number_of_video_patches)r[Z video_size)r4r%rrr\scsg|]}|dqSr]rr^r_rrr\r`num_video_tokens)rr&getupdater(rr) r4Z image_sizesZ video_sizesr5Z vision_datararWZnum_video_patchesrbr)r rr4r%r_get_num_multimodal_tokenss&     z)Glm4vProcessor._get_num_multimodal_tokensTFcKs|jj|f||d|S)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,)`. skip_special_tokens (`bool`, *optional*, defaults to `True`): Whether or not to remove special tokens in the output. Argument passed to the tokenizer's `batch_decode` method. clean_up_tokenization_spaces (`bool`, *optional*, defaults to `False`): Whether or not to clean up the tokenization spaces. Argument passed to the tokenizer's `batch_decode` method. **kwargs: Additional arguments to be passed to the tokenizer's `batch_decode method`. Returns: `list[str]`: The decoded text. )skip_special_tokensclean_up_tokenization_spaces)r)Z batch_decode)r4Zgenerated_outputsrfrgr5rrrpost_process_image_text_to_text sz.Glm4vProcessor.post_process_image_text_to_text)NNNN)NNN)NN)TF)rrr__doc__ attributesZimage_processor_classZvideo_processor_classZtokenizer_classr1rrrr rrr rrrZrerh __classcell__rrr6rr':s(  'r')typingrrnumpyrRZfeature_extraction_utilsrZ image_utilsrZprocessing_utilsrrr r r r Ztokenization_utils_baser rutilsrZ video_utilsrZ get_loggerrrOrrrr'__all__rrrrs       m