fThSrSSKrSSKrSSKJr SSKJrJrJrJ r J r SSK r SSK J s Jr SSK J r SSKJr SSKJrJrJr SS KJr SS KJrJrJrJr SS KJrJrJ r J!r! S S K"J#r#J$r$J%r% \ RL"\'5r(Sr)Sr*Sr+SWSjr,S\ RZS\ RZ4Sjr.\"SS\55r/\"SS\55r0\"SS\55r1"SS\ Rd5r3"SS\ Rd5r4"SS \ Rd5r5"S!S"\ Rd5r6"S#S$\ Rd5r7"S%S&\ Rd5r8"S'S(\ Rd5r9"S)S*\ Rd5r:"S+S,\ Rd5r;"S-S.\ Rd5r<"S/S0\ Rd5r="S1S2\ Rd5r>"S3S4\ Rd5r?"S5S6\ Rd5r@"S7S8\ Rd5rA"S9S:\ Rd5rBS;\A0rC"S<S=\ Rd5rD"S>S?\ Rd5rE"S@SA\ Rd5rF"SBSC\ Rd5rG"SDSE\ Rd5rH"SFSG\ Rd5rI\"SHSI\55rJ"SJSK\J5rK\"SLSM9"SNSO\J55rL\"SPSQ\J55rM\"SRSS\J55rN\"STSU\J55rO/SVQrPg)XzPyTorch CLAP model.N) dataclass)AnyListOptionalTupleUnion)nn)ACT2FN))BaseModelOutputWithPastAndCrossAttentionsBaseModelOutputWithPooling,BaseModelOutputWithPoolingAndCrossAttentions)PreTrainedModel)apply_chunking_to_forward find_pruneable_heads_and_indicesmeshgridprune_linear_layer) ModelOutputauto_docstringlogging torch_int)ClapAudioConfig ClapConfigClapTextConfigcURup#nUSS2SS2SSS24RSSUS5nURX#U-U5nU$)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. Nr)shaperepeatreshape) hidden_statesratio batch_size time_length classes_num upsampleds ^/var/www/auris/envauris/lib/python3.13/site-packages/transformers/models/clap/modeling_clap.py interpolater'*sT.;-@-@*ZkaD!m,33Aq%CI!!*E.A;OI cURup#pEURX#U-XU-X5nURSSSSSS5R5RSXU5nU$)a2 Returns the resized hidden states. The output shape should be `(batch_size * num_windows, window_size, window_size, num_channels)` Args: hidden_states (`torch.FloatTensor` of shape `(batch_size, height, width, num_channels)`): Input hidden states window_size (`int`): Window size rrr rviewpermute contiguous)r window_sizer"heightwidth num_channelswindowss r&window_partitionr7;so/<.A.A+J!&&k);8LkM##Aq!Q15@@BGGKfrsG Nr(cURSnURSX!-X1-XU5nURSSSSSS5R5RSX#U5nU$)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-rrr r*r+r,r.)r6r2r3r4r5s r&window_reverser9Pse==$Lll2v4e6JKfrsGooaAq!Q/::<AA"fUabG Nr(cURU5R5n[R"USS9R U5U-U-nUR 5U-$)z Replace non-padding symbols with their position numbers. Position numbers begin at padding_idx+1. Padding symbols are ignored. This is modified from fairseq's `utils.make_positions`. Args: x: torch.Tensor x: Returns: torch.Tensor rdim)neinttorchcumsumtype_aslong) input_ids padding_idxpast_key_values_lengthmaskincremental_indicess r&"create_position_ids_from_input_idsrHdsW << $ ( ( *D <<!4<