+h SSKrSSKJr SSKJr SS\RS\S\ S\ S\ S\ S\ S \R4S jjr "S S \R5r "S S\R5rg)N timesteps embedding_dim freq_shift min_timescale max_timescaleflip_sin_to_cosscalereturncURS:XdS5eUS-S:Xd SUS35e[US-5n[R"XC- 5Xr- - nU[R "[R "U[RS9U*-5-n [R"US5[R"U S5-n Xj-n U(a@[R"[R"U 5[R"U 5/SS9n O?[R"[R"U 5[R"U 5/SS9n [R"U [R"U5SU/5n U $) aReturns the positional encoding (same as Tensor2Tensor). Args: timesteps (`jnp.ndarray` of shape `(N,)`): A 1-D array of N indices, one per batch element. These may be fractional. embedding_dim (`int`): The number of output channels. freq_shift (`float`, *optional*, defaults to `1`): Shift applied to the frequency scaling of the embeddings. min_timescale (`float`, *optional*, defaults to `1`): The smallest time unit used in the sinusoidal calculation (should probably be 0.0). max_timescale (`float`, *optional*, defaults to `1.0e4`): The largest time unit used in the sinusoidal calculation. flip_sin_to_cos (`bool`, *optional*, defaults to `False`): Whether to flip the order of sinusoidal components to cosine first. scale (`float`, *optional*, defaults to `1.0`): A scaling factor applied to the positional embeddings. Returns: a Tensor of timing signals [N, num_channels] zTimesteps should be a 1d-arrayrzEmbedding dimension z should be even)dtype)axis)ndimfloatmathlogjnpexparangefloat32 expand_dims concatenatecossinreshapeshape) rrrrrrr num_timescaleslog_timescale_incrementinv_timescalesemb scaled_timesignals Z/home/james-whalen/.local/lib/python3.13/site-packages/diffusers/models/embeddings_flax.pyget_sinusoidal_embeddingsr%s6< >>Q @ @@  1  !X%9-#XX !=A-.N"hh}'DEIde"SWWSZZckk-Z^u]u-u%vvN //)Q '#//.!*L LC+K#''+"6 8L!MTUV#''+"6 8L!MTUV [[#))I"6q"9=!I JF Mc\rSrSr%SrSr\\S'\Rr \R\S'\ RS5r Srg) FlaxTimestepEmbeddingDa! Time step Embedding Module. Learns embeddings for input time steps. Args: time_embed_dim (`int`, *optional*, defaults to `32`): Time step embedding dimension. dtype (`jnp.dtype`, *optional*, defaults to `jnp.float32`): The data type for the embedding parameters. time_embed_dimrc[R"URURSS9"U5n[R"U5n[R"URURSS9"U5nU$)Nlinear_1)rnamelinear_2)nnDenser+rsilu)selftembs r$__call__FlaxTimestepEmbedding.__call__RsXxx++4::JOPTUwwt}xx++4::JOPTU r&N)__name__ __module__ __qualname____firstlineno____doc__r+int__annotations__rrrr0compactr5__static_attributes__r7r&r$r(r(Ds;NC{{E399"ZZr&r(ch\rSrSr%SrSr\\S'Sr\ \S'Sr \ \S'\ RS 5rS rg ) FlaxTimestepsZa Wrapper Module for sinusoidal Time step Embeddings as described in https://huggingface.co/papers/2006.11239 Args: dim (`int`, *optional*, defaults to `32`): Time step embedding dimension. flip_sin_to_cos (`bool`, *optional*, defaults to `False`): Whether to flip the sinusoidal function from sine to cosine. freq_shift (`float`, *optional*, defaults to `1`): Frequency shift applied to the sinusoidal embeddings. r*dimFrr rcT[XRURURS9$)N)rrr)r%rDrr)r3rs r$r5FlaxTimesteps.__call__ks'( XXt?S?S`d`o`o  r&r7N)r8r9r:r;r<rDr=r>rboolrrr0r?r5r@r7r&r$rBrBZs< CM!OT!JZZ  r&rB)r r g@Fg?)r flax.linenlinenr0 jax.numpynumpyrndarrayr=rrGr%Moduler(rBr7r&r$rNs   !-{{--- -  -  - - [[-`BII, BII r&