g" ddlmZmZddlmZmZmZmZddlm Z m Z er ddl m Z ddl mZer ddlmZdd lmZejeZee d d Gd d e ZdS))ListUnion)add_end_docstringsis_torch_availableis_vision_availablelogging)Pipelinebuild_pipeline_init_args)Image) load_image)1MODEL_FOR_VISUAL_QUESTION_ANSWERING_MAPPING_NAMES) KeyDatasetT) has_tokenizerhas_image_processorceZdZdZfdZddZ ddedeedeedfdeeeefffd Z dd Z d Z ddZ xZ S)VisualQuestionAnsweringPipelinea Visual Question Answering pipeline using a `AutoModelForVisualQuestionAnswering`. This pipeline is currently only available in PyTorch. Example: ```python >>> from transformers import pipeline >>> oracle = pipeline(model="dandelin/vilt-b32-finetuned-vqa") >>> image_url = "https://huggingface.co/datasets/Narsil/image_dummy/raw/main/lena.png" >>> oracle(question="What is she wearing ?", image=image_url) [{'score': 0.948, 'answer': 'hat'}, {'score': 0.009, 'answer': 'fedora'}, {'score': 0.003, 'answer': 'clothes'}, {'score': 0.003, 'answer': 'sun hat'}, {'score': 0.002, 'answer': 'nothing'}] >>> oracle(question="What is she wearing ?", image=image_url, top_k=1) [{'score': 0.948, 'answer': 'hat'}] >>> oracle(question="Is this a person ?", image=image_url, top_k=1) [{'score': 0.993, 'answer': 'yes'}] >>> oracle(question="Is this a man ?", image=image_url, top_k=1) [{'score': 0.996, 'answer': 'no'}] ``` Learn more about the basics of using a pipeline in the [pipeline tutorial](../pipeline_tutorial) This visual question answering pipeline can currently be loaded from [`pipeline`] using the following task identifiers: `"visual-question-answering", "vqa"`. The models that this pipeline can use are models that have been fine-tuned on a visual question answering task. See the up-to-date list of available models on [huggingface.co/models](https://huggingface.co/models?filter=visual-question-answering). cntj|i||tdSN)super__init__check_model_typer)selfargskwargs __class__s l/var/www/html/ai-engine/env/lib/python3.11/site-packages/transformers/pipelines/visual_question_answering.pyrz(VisualQuestionAnsweringPipeline.__init__7s:$)&))) OPPPPPNc Lii}}|||d<|||d<|||d<|||d<|i|fS)Npadding truncationtimeouttop_k)rr$r!r"r#rpreprocess_paramspostprocess_paramss r_sanitize_parametersz4VisualQuestionAnsweringPipeline._sanitize_parameters;s_02B-  +2 i (  !.8 l +  +2 i (  */ w ' "&888rimagez Image.Imagerquestionc tt}ttotdD}ttotdD}ttjt frtt rd}n|s|r$tt rfdD}nbttjt fr|rfdD}n0|s|r)|r'g}D]} D]} || | d |}n }tj|fi|} | S)a Answers open-ended questions about images. The pipeline accepts several types of inputs which are detailed below: - `pipeline(image=image, question=question)` - `pipeline({"image": image, "question": question})` - `pipeline([{"image": image, "question": question}])` - `pipeline([{"image": image, "question": question}, {"image": image, "question": question}])` Args: image (`str`, `List[str]`, `PIL.Image`, `List[PIL.Image]` or `KeyDataset`): The pipeline handles three types of images: - A string containing a http 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. If given a single image, it can be broadcasted to multiple questions. For dataset: the passed in dataset must be of type `transformers.pipelines.pt_utils.KeyDataset` Example: ```python >>> from transformers.pipelines.pt_utils import KeyDataset >>> from datasets import load_dataset >>> dataset = load_dataset("detection-datasets/coco") >>> oracle(image=KeyDataset(dataset, "image"), question="What's in this image?") ``` question (`str`, `List[str]`): The question(s) asked. If given a single question, it can be broadcasted to multiple images. If multiple images and questions are given, each and every question will be broadcasted to all images (same effect as a Cartesian product) top_k (`int`, *optional*, defaults to 5): The number of top labels that will be returned by the pipeline. If the provided number is higher than the number of labels available in the model configuration, it will default to the number of labels. 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 dictionary or a list of dictionaries containing the result. The dictionaries contain the following keys: - **label** (`str`) -- The label identified by the model. - **score** (`int`) -- The score attributed by the model for that label. c3XK|]%}t|tjtfV&dSr) isinstancer str.0items r z;VisualQuestionAnsweringPipeline.__call__..{s58p8pbfD5;X[J\9]9]8p8p8p8p8p8prc3@K|]}t|tVdSr)r-r.r/s rr2z;VisualQuestionAnsweringPipeline.__call__..|s.>j>jY]z$PS?T?T>j>j>j>j>j>jrr)r*cg|]}|dSr4r%)r0imr*s r z.s!JJJb99JJJrcg|]}|dSr6r%)r0qr)s rr8z.s!HHH!155HHHr) r-rlistallr r.appendr__call__) rr)r*r is_datasetis_image_batchis_question_batchinputsquestion_image_pairsr:r7resultsrs `` rr>z(VisualQuestionAnsweringPipeline.__call__Gsf z22 #E400pS8p8pjo8p8p8p5p5p&x66j3>j>jai>j>j>j;j;j eek3/ 0 0 Z#5N5N $(;;FF    8S0I0I JJJJEJJJFF  S1 2 2 7H HHHHxHHHFF  0A #%  N NNNB(//"!0L0LMMMMN)FF  F"%''"644V44rFc(t|d|}||d|j||}|||j}|jdkr||j}|||S)Nr))r#r*)return_tensorsr!r")imagesrFpt)r tokenizer frameworkimage_processorto torch_dtypeupdate)rrBr!r"r#r) model_inputsimage_featuress r preprocessz*VisualQuestionAnsweringPipeline.preprocesss6'?G<<<~~ : >! &  --U4>-ZZ >T ! !+..t/?@@NN+++rc |jr#d|vr |j|d<|jjdi||}n |jdi|}|S)Ngeneration_configr%)model can_generaterSgenerate)rrOgenerate_kwargs model_outputss r_forwardz(VisualQuestionAnsweringPipeline._forwardso : " " $ $ 7"/997;7M 34/DJ/RR,R/RRMM&DJ6666Mrcjrfd|DS|jjjkrjjj}jdkr8|jd}||\}}ntdj| }| }fdt||DS)Ncpg|]2}dj|di3S)answerT)skip_special_tokens)rIdecodestrip)r0 output_idsrs rr8z?VisualQuestionAnsweringPipeline.postprocess..sP4>00QU0VV\\^^_rrHrzUnsupported framework: cJg|]\}}|jjj|d S))scorer])rTconfigid2label)r0rc_idrs rr8z?VisualQuestionAnsweringPipeline.postprocess..s5sssT^TY[^etz/@/I#/NOOsssr) rTrUrd num_labelsrJlogitssigmoidtopk ValueErrortolistzip)rrXr$probsscoresidss` r postprocessz+VisualQuestionAnsweringPipeline.postprocesss : " " $ $ t"/  tz(333 )4~%%%,4466q9#jj//  !K4>!K!KLLL]]__F**,,Cssssbeflnqbrbrsss sr)NNNNr)FFN)rZ)__name__ __module__ __qualname____doc__rr(rr.rr>rQrYrq __classcell__)rs@rrrs  DQQQQQ 9 9 9 9+/LL]Cm)rsJXXXXXXXXXXXX44444444)((((((%]]]]]]$$$$$$  H % %,,4UYZZZ[[ltltltltlthltlt\[ltltltr