qiixddlZddlmZddlmZddlZddlmcmZ ddlmZddl m Z ddl mZddlmZmZdd lmZdd lmZmZdd lmZdd lmZdd lmZddlmZmZddl m!Z!m"Z"ddl#m$Z$m%Z%ddl&m'Z'ddl(m)Z)m*Z*m+Z+ddl,m-Z-m.Z.m/Z/ddl0m1Z1ddl2m3Z3edGddejhZ5GddejhZ6GddejhZ7dZ8edd;d Z9d!ejtd"e;d#ejtfd$Z< dd=d.Z?d>d/Z@Gd0d1ejhZAGd2d3eZBe*Gd4d5e%ZCe*Gd6d7eCZDe*Gd8d9eCeZEgd:ZFy)?N)Callable)Optional)nn)initialization)ACT2FN)Cache DynamicCache)GenerationMixin)use_kernel_forward_from_hubuse_kernel_func_from_hub)create_causal_mask)FlashAttentionKwargs)GradientCheckpointingLayer)BaseModelOutputWithPastCausalLMOutputWithPast)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)TransformersKwargsauto_docstringcan_return_tuple)is_flash_attention_requestedmaybe_autocastmerge_with_config_defaults)capture_outputs) YoutuConfigRMSNormcheZdZddeddffd Zdej dej fdZdZxZ S) YoutuRMSNormepsreturnNct|tjt j ||_||_y)z; YoutuRMSNorm is equivalent to T5LayerNorm N)super__init__r Parametertorchonesweightvariance_epsilon)self hidden_sizer$ __class__s Z/opt/pipecat/venv/lib/python3.12/site-packages/transformers/models/youtu/modeling_youtu.pyr(zYoutuRMSNorm.__init__6s1 ll5::k#:; # hidden_statesc"|j}|jtj}|j dj dd}|tj ||jzz}|j|j|zS)NT)keepdim) dtypetor*float32powmeanrsqrtr-r,)r.r3 input_dtypevariances r1forwardzYoutuRMSNorm.forward>sy#)) %((7  $$Q',,R,>% Ht?T?T4T(UU {{]--k:::r2c^t|jjd|jS)Nz, eps=)tupler,shaper-)r.s r1 extra_reprzYoutuRMSNorm.extra_reprEs* ))*+6$2G2G1HIIr2)gư>) __name__ __module__ __qualname__floatr(r*Tensorr@rD __classcell__r0s@r1r#r#4s7$$$$;U\\;ell;Jr2r#c eZdZUejed<d deffd Ze d dedzde dde dzde d e ffd Z ejed ZxZS)YoutuRotaryEmbeddinginv_freqNconfigct||j|_|j|_||_|j j d|_|j}|jdk7rt|j}||j |\}|_ |jd|d|jd|jdy)N rope_typedefaultrNF) persistentoriginal_inv_freq) r'r(max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenrOrope_parametersrQcompute_default_rope_parametersrattention_scalingregister_bufferclone)r.rOdevice rope_init_fnrNr0s r1r(zYoutuRotaryEmbedding.__init__Ls "("@"@$*$B$B! 44[A!%!E!E >>Y &.t~~>L+7 V+L($( ZeD 0(..2BuUr2r]z torch.deviceseq_lenr%z torch.Tensorc|jd}t|ddxs|j|jz}d}d|t j d|dtj j|tj|z zz }||fS) a Computes the inverse frequencies according to the original RoPE implementation Args: config ([`~transformers.PreTrainedConfig`]): The model configuration. device (`torch.device`): The device to use for initialization of the inverse frequencies. seq_len (`int`, *optional*): The current sequence length. Unused for this type of RoPE. Returns: Tuple of (`torch.Tensor`, `float`), containing the inverse frequencies for the RoPE embeddings and the post-processing scaling factor applied to the computed cos/sin (unused in this type of RoPE). rope_thetahead_dimN?rr5r8)r]r8) rXgetattrr/num_attention_headsr*arangeint64r9rH)rOr]r_basedimattention_factorrNs r1rYz4YoutuRotaryEmbedding.compute_default_rope_parameters\s&%%l3fj$/c63E3EIcIc3c U\\!S!5;;?BB&X]XcXcBdgjj k )))r2cN|jddddfjj|jdddj |j }|dddddfj}t |j jtr/|j jdk7r|j jnd}t|d5|j|jzjdd}tj||fd }|j|jz}|j|jz} dddj |j  j |j fS#1swYAxYw) Nrr6rmpscpuF) device_typeenabledr5rjrd)rNrHexpandrCr9r] isinstancetypestrr transposer*catcosrZsinr8) r.x position_idsinv_freq_expandedposition_ids_expandedrofreqsembrxrys r1r@zYoutuRotaryEmbedding.forwardzsR!MM$4-8>>@GG HZHZ[\H]_acdehhijiqiqr ,QaZ 8 > > @'1!((--'E!((--[`J`ahhmmfk  U C 5&,,.1F1L1L1NNYYZ[]^_E))UEN3C'')d444C'')d444C  5 vvAGGv$cff177f&;;;  5 5s BFF$N)NNN)rErFrGr*rI__annotations__r r( staticmethodrintrBrHrYno_gradrr@rJrKs@r1rMrMIsllV{V %)+/"*d"*(*t* ~u$ % **:U]]_ <