+ jg!DRt^RIt^RIt^RIHt^RIHt^RIHt^RI t ^RI H uH t ^RI H t ^RIHt^RIHt^RIHt^R IHt^R IHtHt^R IHtHt^R IHt^R IH t ^RI!H"t"H#t#H$t$H%t%H&t&H't'^RI(H)t)^RI*H+t+^RI,H-t-H.t.H/t/]&P`!]14t2Rt3Rt4Rt5RRlt6]$!RR7]!RR]"444t7]$!RR7]!RR]"444t8]$]!RR]"444t9!R R!] Pt4t;!R"R#] Pt4t<!R$R%] Pt4t=!R&R'] Pt4t>!R(R)] Pt4t?!R*R+] Pt4t@!R,R-] Pt4tA!R.R/] Pt4tB!R0R1] Pt4tC!R2R3]4tD!R4R5] Pt4tE!R6R7] Pt4tF!R8R9] Pt4tG!R:R;] Pt4tHR[R<R=lltI!R>R?] Pt4tJ!R@RA] Pt4tK!RBRC] Pt4tL!RDRE] Pt4tM!RFRG] Pt4tN!RHRI]4tO!RJRK] Pt4tP!RLRM] Pt4tQ]$!RNRO]44tR!RPRQ]R4tS]$!RRR7!RSRT]R44tT]$!RURV]R44tU]$!RWRX]R44tV]$!RYRZ]R44tW.R\OtXR#)]zPyTorch CLAP model.N)Callable) dataclass)Any)nn)initialization)ACT2FN)create_bidirectional_mask)GradientCheckpointingLayer)BaseModelOutputBaseModelOutputWithPooling)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)apply_chunking_to_forward) ModelOutputTransformersKwargsauto_docstringcan_return_tuplelogging torch_int)merge_with_config_defaults)capture_outputs)ClapAudioConfig ClapConfigClapTextConfigcVPwr#pVR,P^^V^4pVPW#V,V4pV#)aI Interpolate data in time domain. This is used to compensate the resolution reduction in downsampling of a CNN. Args: hidden_states (`torch.FloatTensor` of shape (batch_size, time_length, classes_num)): Input hidden states ratio (`int`): The ratio of the length of the output to the length of the input. )NNNrNr)shaperepeatreshape) hidden_statesratio batch_size time_length classes_num upsampleds&& g/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/models/clap/modeling_clap.py interpolater'/sL.;-@-@*Zkm,33Aq%CI!!*E.A;OI cVPwr#rEVPW#V,WV,W4pVP^^^^^^4P4PRWV4pV#)a2 Returns the resized hidden states. The output shape should be `(batch_size * num_windows, window_size, window_size, num_channels)` Args: hidden_states (`torch.FloatTensor` of shape `(batch_size, height, width, num_channels)`): Input hidden states window_size (`int`): Window size rviewpermute contiguous)r window_sizer"heightwidth num_channelswindowss&& r&window_partitionr4@so/<.A.A+J!&&k);8LkM##Aq!Q15@@BGGKfrsG Nr(cVPR,pVPRW!,W1,WV4pVP^^^^^^4P4PRW#V4pV#)a_ Merges windows to produce higher resolution features. Args: windows (`torch.FloatTensor` of shape `(num_windows * batch_size, window_size, window_size, num_channels)`): Input windows window_size (`int`): Window size height (`int`): Height of the resized audio width (`int`): Width of the resized audio r*r+)r3r/r0r1r2s&&&& r&window_reverser6Use==$Lll2v4e6JKfrsGooaAq!Q/::<AA"fUabG Nr(cXV^8dQhR\PR\P/#)logitsreturntorchTensor)formats"r& __annotate__r?js"77U\\7ell7r(c\P!\V4VPR7p\P P W4#)device)r<arangelenrBr functional cross_entropy)r9labelss& r&contrastive_lossrHjs/ \\#f+fmm