^j)YdZddlmZddlmZddlZddlmZddlmZddl m Z m Z dd l m Z dd lmZdd lmZdd lmZmZmZdd lmZmZddlmZmZddlmZmZddlm Z ddl!m"Z"m#Z#m$Z$m%Z%ddl&m'Z'm(Z(ddl)m*Z*ddl+m,Z,e%jZe.Z/Gddej`Z1dZ2d3dZ3Gddej`Z4 d4dej`dejjdejjdejjd ejjdzd!e6d"e6fd#Z7Gd$d%ej`Z8Gd&d'eZ9e#Gd(d)eZ:e#Gd*d+e:Z;Gd,d-e:e Z<Gd.d/ee:Z=Gd0d1ee:Z>gd2Z?y)5zPyTorch Persimmon model.)Callable)OptionalN)nn)ACT2FN)Cache DynamicCache)GenerationMixin)create_causal_mask)FlashAttentionKwargs) GenericForSequenceClassificationGenericForTokenClassificationGradientCheckpointingLayer)BaseModelOutputWithPastCausalLMOutputWithPast)ROPE_INIT_FUNCTIONSdynamic_rope_update)ALL_ATTENTION_FUNCTIONSPreTrainedModel)Unpack)TransformersKwargsauto_docstringcan_return_tuplelogging)maybe_autocastmerge_with_config_defaults)capture_outputs)PersimmonConfigc eZdZUejed<d deffd Ze d dedzde dde dzde d e ffd Z ejed ZxZS)PersimmonRotaryEmbeddinginv_freqNconfigct||j|_|j|_||_|j j d|_|j}|jdk7rt|j}||j |\}|_ |jd|d|jd|jdy)N rope_typedefaultr"F) persistentoriginal_inv_freq) super__init__max_position_embeddingsmax_seq_len_cachedoriginal_max_seq_lenr#rope_parametersr%compute_default_rope_parametersrattention_scalingregister_bufferclone)selfr#device rope_init_fnr" __class__s {/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/persimmon/modeling_persimmon.pyr*z!PersimmonRotaryEmbedding.__init__<s "("@"@$*$B$B! 44[A!%!E!E >>Y &.t~~>L+7 V+L($( ZeD 0(..2BuUr4z torch.deviceseq_lenreturnz torch.Tensorcn|jd}|jjdd}t|ddxs|j|jz}t ||z}d}d|t jd|dt jj|t j |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_thetapartial_rotary_factorg?head_dimNrdtype)r4rA) r.getgetattr hidden_sizenum_attention_headsinttorcharangeint64tofloat) r#r4r9baser=r>dimattention_factorr"s r7r/z8PersimmonRotaryEmbedding.compute_default_rope_parametersLs(%%l3 & 6 6 : :;RTW X6:t4h8J8JfNhNh8h(223 U\\!S!5;;?BB&X]XcXcBdgjj k )))r8cN|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) NrrmpscpuF) device_typeenabledr?rMr@)r"rKexpandshaperJr4 isinstancetypestrr transposerGcatcosr0sinrA) r3x position_idsinv_freq_expandedposition_ids_expandedrSfreqsembr]r^s r7forwardz PersimmonRotaryEmbedding.forwardmsR!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$N)NNN)__name__ __module__ __qualname__rGTensor__annotations__rr* staticmethodrrFtuplerKr/no_gradrre __classcell__r6s@r7r!r!9sllVV *.+/"*$&*(*t* ~u$ % **>U]]_ <