Vpf(ddlmZddlmZddlZddlmZddlZddlm Z ddl m Z ddl m Z eZ dddZ d d!dZdS)") annotations)SequenceN)Any) lax_numpy)lax) convolutionlhsjax.typing.ArrayLike filter_shape Sequence[int]window_stridespaddingstr | Sequence[tuple[int, int]] lhs_dilationSequence[int] | None rhs_dilationdimension_numbers5convolution.ConvGeneralDilatedDimensionNumbers | None precisionlax.Precision | Nonepreferred_element_type DType | Nonereturn jax.Arrayc ntj|} t|}tj| jd|z|}|\} } } t j|} | j| d}tj| | j  |dz}| | df|z}tj ||fd|j dz zz}tj |d| d| df}tj| |||||||dn|tjjf|| }|S) a Extract patches subject to the receptive field of `conv_general_dilated`. Runs the input through a convolution with given parameters. The kernel of the convolution is constructed such that the output channel dimension `"C"` contains flattened image patches, so instead a single `"C"` dimension represents, for example, three dimensions `"chw"` collapsed. The order of these dimensions is `"c" + ''.join(c for c in rhs_spec if c not in 'OI')`, where `rhs_spec == dimension_numbers[1]`, and the size of this `"C"` dimension is therefore the size of each patch, i.e. `np.prod(filter_shape) * lhs.shape[lhs_spec.index('C')]`, where `lhs_spec == dimension_numbers[0]`. Docstring below adapted from `jax.lax.conv_general_dilated`. See Also: https://www.tensorflow.org/xla/operation_semantics#conv_convolution Args: lhs: a rank `n+2` dimensional input array. filter_shape: a sequence of `n` integers, representing the receptive window spatial shape in the order as specified in `rhs_spec = dimension_numbers[1]`. window_strides: a sequence of `n` integers, representing the inter-window strides. padding: either the string `'SAME'`, the string `'VALID'`, or a sequence of `n` `(low, high)` integer pairs that give the padding to apply before and after each spatial dimension. lhs_dilation: `None`, or a sequence of `n` integers, giving the dilation factor to apply in each spatial dimension of `lhs`. LHS dilation is also known as transposed convolution. rhs_dilation: `None`, or a sequence of `n` integers, giving the dilation factor to apply in each spatial dimension of `rhs`. RHS dilation is also known as atrous convolution. dimension_numbers: either `None`, or a 3-tuple `(lhs_spec, rhs_spec, out_spec)`, where each element is a string of length `n+2`. `None` defaults to `("NCHWD..., OIHWD..., NCHWD...")`. precision: Optional. Either ``None``, which means the default precision for the backend, or a :class:`~jax.lax.Precision` enum value (``Precision.DEFAULT``, ``Precision.HIGH`` or ``Precision.HIGHEST``). preferred_element_type: Optional. Either ``None``, which means the default accumulation type for the input types, or a datatype, indicating to accumulate results to and return a result with that datatype. Returns: A rank `n+2` array containing the flattened image patches in the output channel (`"C"`) dimension. For example if `dimension_numbers = ("NcHW", "OIwh", "CNHW")`, the output has dimension numbers `"CNHW" = "{cwh}NHW"`, with the size of dimension `"C"` equal to the size of each patch (`np.prod(filter_shape) * lhs.shape[lhs_spec.index('C')]`). rr)dtype)r)rrrN) r rhsr rrrrrfeature_group_countr)jnpasarraytuplerconv_dimension_numbersshapemathprodeyerreshapetilendimmoveaxisconv_general_dilatedr PrecisionDEFAULT)r r r rrrrrr lhs_arraylhs_specrhs_specout_spec spatial_size n_channelsr outs R/var/www/html/nettyfy-visnx/env/lib/python3.11/site-packages/jax/_src/lax/other.pyconv_general_dilated_patchesr9sN~k#)|$$,!8ov ,.?AA"3(Hh<((,x{+*  IO444<<\A=MNN# \1% 455# zmdchl&;;<<# S&8A; "<==#(  #)!) 03 0E0G$3   # *r lax.PrecisionLikec tj|} tj|} t | t rt | dkr| dn| } t| ||||||| } tj | j dt |z|\} }}|dg|dg}}|dd|ddfdttt fd Dt||fff}tj | ||| }tj|d |d|df}|S) a3General n-dimensional unshared convolution operator with optional dilation. Also known as locally connected layer, the operation is equivalent to convolution with a separate (unshared) `rhs` kernel used at each output spatial location. Docstring below adapted from `jax.lax.conv_general_dilated`. See Also: https://www.tensorflow.org/xla/operation_semantics#conv_convolution Args: lhs: a rank `n+2` dimensional input array. rhs: a rank `n+2` dimensional array of kernel weights. Unlike in regular CNNs, its spatial coordinates (`H`, `W`, ...) correspond to output spatial locations, while input spatial locations are fused with the input channel locations in the single `I` dimension, in the order of `"C" + ''.join(c for c in rhs_spec if c not in 'OI')`, where `rhs_spec = dimension_numbers[1]`. For example, if `rhs_spec == "WHIO", the unfolded kernel shape is `"[output W][output H]{I[receptive window W][receptive window H]}O"`. window_strides: a sequence of `n` integers, representing the inter-window strides. padding: either the string `'SAME'`, the string `'VALID'`, or a sequence of `n` `(low, high)` integer pairs that give the padding to apply before and after each spatial dimension. filter_shape: a sequence of `n` integers, representing the receptive window spatial shape in the order as specified in `rhs_spec = dimension_numbers[1]`. lhs_dilation: `None`, or a sequence of `n` integers, giving the dilation factor to apply in each spatial dimension of `lhs`. LHS dilation is also known as transposed convolution. rhs_dilation: `None`, or a sequence of `n` integers, giving the dilation factor to apply in each input spatial dimension of `rhs`. RHS dilation is also known as atrous convolution. dimension_numbers: either `None`, a `ConvDimensionNumbers` object, or a 3-tuple `(lhs_spec, rhs_spec, out_spec)`, where each element is a string of length `n+2`. precision: Optional. Either ``None``, which means the default precision for the backend, a ``lax.Precision`` enum value (``Precision.DEFAULT``, ``Precision.HIGH`` or ``Precision.HIGHEST``) or a tuple of two ``lax.Precision`` enums indicating precision of ``lhs``` and ``rhs``. Returns: An array containing the unshared convolution result. In the string case of `dimension_numbers`, each character identifies by position: - the batch dimensions in `lhs`, `rhs`, and the output with the character 'N', - the feature dimensions in `lhs` and the output with the character 'C', - the input and output feature dimensions in rhs with the characters 'I' and 'O' respectively, and - spatial dimension correspondences between `lhs`, `rhs`, and the output using any distinct characters. For example, to indicate dimension numbers consistent with the `conv` function with two spatial dimensions, one could use `('NCHW', 'OIHW', 'NCHW')`. As another example, to indicate dimension numbers consistent with the TensorFlow Conv2D operation, one could use `('NHWC', 'HWIO', 'NHWC')`. When using the latter form of convolution dimension specification, window strides are associated with spatial dimension character labels according to the order in which the labels appear in the `rhs_spec` string, so that `window_strides[0]` is matched with the dimension corresponding to the first character appearing in rhs_spec that is not `'I'` or `'O'`. If `dimension_numbers` is `None`, the default is `('NCHW', 'OIHW', 'NCHW')` (for a 2D convolution). rr)r r r rrrrrrrNc g|] }| Sr>).0i rhs_b_dimss r8 z.conv_general_dilated_local..s/LLL! 1 LLLr:c|S)Nr>)k lhs_b_dimss r8z,conv_general_dilated_local..s JqMr:)key)rr))r"r#rcanonicalize_precision isinstancer$lenr9rr%r&sortedrange dot_generalr-)r r r rr rrrrr1 c_precision lhs_precisionpatchesr2r3r4 lhs_c_dims rhs_c_dimsdnr7rErAs @@r8conv_general_dilated_localrV|s^k#)*955+ [% ( (-0-=-=-B-B!nn  ) #)   '"-!Covl 3 335F"H"H(Hh%QK=8A;-j*|*|*LLLLveC OO.D.D2I2I2I2I(K(K(KLLL*j!!* Z :z":;" iPPP# S(Xa[(1+$>??# *r:)NNNNN)r r r r r r rrrrrrrrrrrrrr)NNNN)r r r r r r rrr r rrrrrrrr;rr) __future__rcollections.abcrr'typingrjaxjax._src.numpyrr" jax._src.laxrrDTyper9rVr>r:r8r^s#"""""$$$$$$  ++++++$$$$$$ *.)-OS&*+/] ] ] ] ] L*.)-OS#'r r r r r r r r: