S^(hS dZddlZddlZddlZddlZddlmZddlmZddl m Z m Z m Z m Z mZddlZddlmcmZddlmZmZddlmZddlmZd d lmZd d lmZmZmZmZm Z m!Z!m"Z"d d l#m$Z$d d l%m&Z&d dl'm(Z(d dl)m*Z*d dl+m,Z,m-Z-m.Z.m/Z/m0Z0d dl1m2Z2ddl3m4Z4e/jje6Z7da8dZ9GddeZ:e,r ddl;mZ>e!rddl?m@Z@e.rddlAmBZBerddlCmDZDe/jje6Z7dZEdZFeGddeZGeGd d!eZHeGd"d#eZId$ZJdid%ZKGd&d'ejZMd(ZNGd)d*ejZOGd+d,ejZPGd-d.ejZQd/ZRd0ed1ed2ed3ed4ef d5ZSGd6d7ejZTGd8d9ejZUGd:d;ejZVGd<d=ejZWGd>d?e(ZXd@ZYdAZZGdBdCeXZ[GdDdEeXZ\edFeYGdGdHeXZ]edIeYGdJdKeXZ^dLZ_djdMe`dNe`fdOZaGdPdQejZbGdRdSejZcGdTdUejZddVed4efdWZedXed4efdYZfdZZgd[Zhd\ZiGd]d^Zjd_ej(d`ekdae`dbekfdcZlddZmGdedfejZnGdgdhejZoy)kzPyTorch DETA model.N) dataclass)Path)DictListOptionalTupleUnion)Tensornn)Function)once_differentiable)ACT2FN) ModelOutputadd_start_docstrings%add_start_docstrings_to_model_forwardis_scipy_availableis_torch_cuda_availableis_vision_availablereplace_return_docstrings)_prepare_4d_attention_mask)BaseModelOutput)PreTrainedModel)meshgrid)is_accelerate_availableis_ninja_availableis_torchvision_availableloggingrequires_backends) load_backbone) DetaConfigc tddlm}ttj j j j dz dz }dt jjddt jjdd fDcgc]}||z  }}|d |d t|gd ggd ycc}w)Nr)loadkernelsdetaz vision.cppcpuzms_deform_attn_cpu.cppcudazms_deform_attn_cuda.cuMultiScaleDeformableAttentionTz -DWITH_CUDA=1)z-DCUDA_HAS_FP16=1z-D__CUDA_NO_HALF_OPERATORS__z-D__CUDA_NO_HALF_CONVERSIONS__z-D__CUDA_NO_HALF2_OPERATORS__) with_cudaextra_include_paths extra_cflagsextra_cuda_cflags) torch.utils.cpp_extensionr$r__file__resolveparentospathjoinstr)r$rootfilename src_filess o/var/www/skyplay_api_hub/venv/lib/python3.12/site-packages/transformers/models/deprecated/deta/modeling_deta.pyload_cuda_kernelsr:7s. > ! ! # * * 1 1 8 89 Dv MD  GGLL 8 9 GGLL!9 :   xI ' YK%& s B5c6eZdZedZeedZy)%MultiScaleDeformableAttentionFunctionc||_tj||||||j}|j||||||SN) im2col_stepr)ms_deform_attn_forwardsave_for_backward)contextvaluevalue_spatial_shapesvalue_level_start_indexsampling_locationsattention_weightsr?outputs r9forwardz-MultiScaleDeformableAttentionFunction.forwardVsZ*.EE  #        !! ')@BTVg  c |j\}}}}}tj|||||||j\}}} |dd|| dfSr>) saved_tensorsr)ms_deform_attn_backwardr?) rB grad_outputrCrDrErFrG grad_valuegrad_sampling_locgrad_attn_weights r9backwardz.MultiScaleDeformableAttentionFunction.backwardnsl  ! !   #  :W:o:o  #      ; 7 %'74'8:JDPPrJN)__name__ __module__ __qualname__ staticmethodrIr rRrJr9r<r<Us2.QQrJr<) PartialState)reduce)center_to_corners_format) batched_nms)linear_sum_assignmentr"zjozhang97/deta-swin-large-o365ceZdZUdZdZeejed<dZ eejed<dZ eejed<dZ ee ejed<dZ ee ejed<dZee ejed<y) DetaDecoderOutputa Base class for outputs of the DetaDecoder. This class adds two attributes to BaseModelOutputWithCrossAttentions, namely: - a stacked tensor of intermediate decoder hidden states (i.e. the output of each decoder layer) - a stacked tensor of intermediate reference points. Args: last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`): Sequence of hidden-states at the output of the last layer of the model. intermediate_hidden_states (`torch.FloatTensor` of shape `(batch_size, config.decoder_layers, num_queries, hidden_size)`): Stacked intermediate hidden states (output of each layer of the decoder). intermediate_reference_points (`torch.FloatTensor` of shape `(batch_size, config.decoder_layers, sequence_length, hidden_size)`): Stacked intermediate reference points (reference points of each layer of the decoder). hidden_states (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`): Tuple of `torch.FloatTensor` (one for the output of the embeddings + one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`. Hidden-states of the model at the output of each layer plus the initial embedding outputs. attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`): Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads. cross_attentions (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` and `config.add_cross_attention=True` is passed or when `config.output_attentions=True`): Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length, sequence_length)`. Attentions weights of the decoder's cross-attention layer, after the attention softmax, used to compute the weighted average in the cross-attention heads. Nlast_hidden_stateintermediate_hidden_statesintermediate_reference_points hidden_states attentionscross_attentions)rSrTrU__doc__r_rtorch FloatTensor__annotations__r`rarbrrcrdrWrJr9r^r^s66:x 1 129>B):): ;BAE!8E,=,=#>E8