U 5Af @sddlmZmZddlmZmZmZmZmZm Z ddl m Z m Z er^ddl mZddlmZerpddlmZerdd lZdd lmZeeZee d d d Gd dde Zd S))ListUnion)add_end_docstringsis_tf_availableis_torch_availableis_vision_availableloggingrequires_backends)Pipelinebuild_pipeline_init_args)Image) load_image)'TF_MODEL_FOR_VISION_2_SEQ_MAPPING_NAMESN)$MODEL_FOR_VISION_2_SEQ_MAPPING_NAMEST)Z has_tokenizerZhas_image_processorcsjeZdZdZfddZdddZeeeededfdfd d Z dd d Z d dZ ddZ Z S)ImageToTextPipelinea Image To Text pipeline using a `AutoModelForVision2Seq`. This pipeline predicts a caption for a given image. Example: ```python >>> from transformers import pipeline >>> captioner = pipeline(model="ydshieh/vit-gpt2-coco-en") >>> captioner("https://huggingface.co/datasets/Narsil/image_dummy/raw/main/parrots.png") [{'generated_text': 'two birds are standing next to each other '}] ``` Learn more about the basics of using a pipeline in the [pipeline tutorial](../pipeline_tutorial) This image to text pipeline can currently be loaded from pipeline() using the following task identifier: "image-to-text". See the list of available models on [huggingface.co/models](https://huggingface.co/models?pipeline_tag=image-to-text). cs4tj||t|d||jdkr*tntdS)NZvisiontf)super__init__r Zcheck_model_type frameworkrr)selfargskwargs __class__H/tmp/pip-unpacked-wheel-zw5xktn0/transformers/pipelines/image_to_text.pyrEs  zImageToTextPipeline.__init__NcCsli}i}|dk r||d<|dk r(||d<|dk r8||d<|dk rb|dk rXd|krXtd||||ifS)Nprompttimeoutmax_new_tokenszp`max_new_tokens` is defined both as an argument and inside `generate_kwargs` argument, please use only 1 version) ValueErrorupdate)rr generate_kwargsrrZforward_paramsZpreprocess_paramsrrr_sanitize_parametersLs z(ImageToTextPipeline._sanitize_parametersz Image.Image)imagesc stj|f|S)a Assign labels to the image(s) passed as inputs. Args: images (`str`, `List[str]`, `PIL.Image` or `List[PIL.Image]`): The pipeline handles three types of images: - A string containing a HTTP(s) link pointing to an image - A string containing a local path to an image - An image loaded in PIL directly The pipeline accepts either a single image or a batch of images. max_new_tokens (`int`, *optional*): The amount of maximum tokens to generate. By default it will use `generate` default. generate_kwargs (`Dict`, *optional*): Pass it to send all of these arguments directly to `generate` allowing full control of this function. timeout (`float`, *optional*, defaults to None): The maximum time in seconds to wait for fetching images from the web. If None, no timeout is set and the call may block forever. Return: A list or a list of list of `dict`: Each result comes as a dictionary with the following key: - **generated_text** (`str`) -- The generated text. )r__call__)rr%rrrrr&aszImageToTextPipeline.__call__cCst||d}|dk rsz/ImageToTextPipeline._forward..)r,listallpopr/Zmain_input_namegenerate)rr6r#inputs model_outputsrrr_forwards zImageToTextPipeline._forwardcCs0g}|D]"}d|jj|ddi}||q|S)NZgenerated_textT)Zskip_special_tokens)r3decodeappend)rr@recordsZ output_idsrecordrrr postprocesss zImageToTextPipeline.postprocess)NNNN)NN)__name__ __module__ __qualname____doc__rr$rr-rr&r7rArF __classcell__rrrrr-s  & /r)typingrrutilsrrrrr r baser r ZPILrZ image_utilsrZmodels.auto.modeling_tf_autorr4Zmodels.auto.modeling_autorZ get_loggerrGloggerrrrrrs