^j8'ddlZddlZddlmcmZddlmZddl m Z m Z ddl m Z mZmZmZddlmZmZmZddlmZmZddlmZej4eZGd d ed Zd ZedDcgc] }d|dd c}edDcgc] }d|dd c}zZ eedGddeZ!dgZ"ycc}wcc}w)N) BatchFeature) ImageInputmake_nested_list_of_images)MultiModalDataProcessingKwargsProcessorMixinUnpack) AddedTokenPreTokenizedInput TextInput)auto_docstringlogging)requiresc"eZdZddddddidZy) PI0ProcessorKwargs max_length0right)paddingr padding_sidereturn_tensorspt) text_kwargs common_kwargsN)__name__ __module__ __qualname__ _defaultsq/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/pi0/processing_pi0.pyrr$s#$# +D1 Ir!rF)totalziz4>z3)visiontorch)backendsc0eZdZd fd Ze d deeezeeezdzdeezeezeezdzdee jze jzdzdee jze jzdzde edef d Zd d Zefd ZxZS) PI0ProcessorNc |jd|jdc|_|_|jdgd}|jdgd}|jdgd}|jd gd }t j ||_t j ||_t j ||_t j ||_ |jd d |_ |jd d|_ t|ds td|j|_t|dsNtt dd} d| gi} |j#| |j%t |_t |_n"|j&|_|j(|_|j+t,d|_d|_t2 |i|||y)Nheightwidth state_mean)ggsgr?g h"l?gW2D@g\ AcgZd;OſgB>٬?g QI state_std)gt$~?gL F%u?g.!u?g/n?g6?gx?g]K=?gF%u? actions_mean)g&S?gX?gW[재gHPsr?gg?g gHPs actions_std)gGz?g`"?g9#J{?gvOjM?g>yX5ͫ?g46|d|j:|j8dz f}|jAd|jB|j:|d<|t%j2||jDz |jFdzz }|j8d|j:kr2t=j>|d|j:|j8dz f}|jAd|j:|d<tI||S)aC actions (`list | np.ndarray | torch.Tensor`, *optional*): Actions to be predicted by the model. If provided, padding, mean and std normalization will be applied. state (`list | np.ndarray | torch.Tensor`, *optional*): Robotic states to be predicted by the model. If provided, padding, mean and std normalization will be applied. 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`. - **pixel_attention_mask** -- Pixel values padding mask to be fed to a model. Returned when `images` is not `None`. - **state** -- Robot state compatible with model if `state` is not None - **actions** -- Label-actions compatible with training if `actions` is not None tokenizer_init_kwargsNzPYou are using PI0 without a text prefix. The processor will use an empty prompt.z Received z image samples for z\ prompts. Each prompt should be associated with one sample (with one or more camera images).rr images_kwargs c32K|]}t|ywN)len).0 sample_imagess r" z(PI0Processor.__call__..sU]c-0Us)dtyperrT pixel_values)rbpixel_attention_maskg:0yE>rrSrT)data tensor_type)% _merge_kwargsrrM init_kwargslogger warning_once isinstancestrrr]r@popzipr7r6 bos_tokenappendmaxr'zerosboolr,r- enumeraterLr>r0r1shaper2Fpadviewr4r.r/r)rKrQrRrSrTrN output_kwargsbatched_imagesrprompt_stringssample image_list text_inputsmax_num_camerasrcpadded_pixel_valuesbatchr_ processed num_cameras return_datas r"__call__zPI0Processor.__call__Xs|8+**  6:nn6P6P TZ  <    r sD dC 6D3F; ~ #d) +C/00CCI;Oee  '}599:JDQo&**+;TB"%dN"; * FJ##d&;&;;c*oMNt~~OgOgNhiohpprs   ! !& )  * %dnn^T}]7ST UnUU${{C,?+QY^YcYcd#kk#n*=PQSWS^S^`d`j`jk$-n$= Q E=,,,]r4rS`apSqrIm,K8<   !4 57@7P | | 3 4  Q  /$8  ||G,t/@/@@TEUEUX]E]^G}}R 4#5#55%%!T-?-?'--PRBS-S)TU%,\\"dootGYGY%ZK "  \\%(4??:t~~PU?UVE{{2!3!33ea););ekk"o)M%NO#(::b$2D2D#EK .IIr!c i}|<|jgt|z}dgt|z}|j||dtdi|S)a Computes the number of placeholder tokens needed for multimodal inputs with the given sizes. Args: image_sizes (list[list[str]], *optional*): The input sizes formatted as (height, width) per each image. Returns: `MultiModalData`: A `MultiModalData` object holding number of tokens per each of the provided input modalities, along with other useful data. )num_image_tokensnum_image_patchesr )r6r]updater)rK image_sizesrN vision_datarrs r"_get_num_multimodal_tokensz'PI0Processor._get_num_multimodal_tokenss]  " $ 5 56[9II !"c+&6 6    4D[lm n, ,,r!c t|dgzS)Nrc)rImodel_input_names)rKrPs r"rzPI0Processor.model_input_namessw(,B+CCCr!)NNNr\)rrrrJrrlistr r npndarrayr'Tensorr rrrrpropertyr __classcell__)rPs@r"r*r*3s RDbf;?9= WJT*--T*5E0FFMWJ++d9o=EV@WWZ^^WJ "U\\1D8 WJ bjj 5<</$6 WJ +, WJ WJWJr-$DDr!r*)#numpyrr'torch.nn.functionalnn functionalrvfeature_extraction_utilsr image_utilsrrprocessing_utilsrrr r tokenization_utils_baser r r utilsrrutils.import_utilsr get_loggerrrirrArangerFr*__all__)is0r"rs* 4AXXOO,*   H %) ).t5A$qgQ5RWX[R\8]Q4#wa8]]  &'QD>QD(QDh  s68]s /B< C