| This operator fills the elements of the output
| tensor with a constant value specified by the
| value argument.
|
| - The data type is specified by the dtype
| argument
|
| - Currently, the data types supported are f32,
| int32, int64, and bool
|
| - If the dtype argument is not provided, the
| data type of value is used
|
| - The output tensor shape is either specified by
| the shape argument or will match the shape of
| the input tensor if one is provided (if an input
| tensor is provided, a shape argument should not be
| set)
|
| - Optional additional dimensions can be appended
| at the end as specified by extra_shape argument
|
| - If input_as_shape is set to True, the input
| should be a 1D tensor containing the desired
| output shape (the dimensions specified in
| extra_shape will also be appended)
|
| - If a second input V is passed, fill the output
| with the first element of V
|
| When specifying dtype argument, use the integer
| keys from the DataType enum in TensorProto:
|
| | message TensorProto { | ... | enum DataType { | UNDEFINED = 0; | FLOAT = 1; // float | INT32 = 2; // int | BYTE = 3; // BYTE, when deserialized, is going to be restored as uint8. | STRING = 4; // string | BOOL = 5; // bool | UINT8 = 6; // uint8_t | INT8 = 7; // int8_t | UINT16 = 8; // uint16_t | INT16 = 9; // int16_t | INT64 = 10; // int64_t | FLOAT16 = 12; // at::Half | DOUBLE = 13; // double | } |
|
| Github Links:
|
| - https://github.com/pytorch/pytorch/blob/master/caffe2/operators/filler_op.cc
| The operator fills the diagonal elements
| of the output tensor (>= 2D) with a constant
| value specified by the ‘value’ argument,
| and others 0. If number of dimensions
| of the output tensor is greater than
| 2, all dimensions must be equal.
|
| The data type is specified by the ‘dtype’
| argument. The ‘dtype’ argument must
| be one of the data types specified in
| the ‘DataType’ enum field in the TensorProto
| message. If the ‘dtype’ argument is
| not provided, the data type of ‘value’
| is used.
|
| The output tensor shape is specified
| by the ‘shape’ argument. If the number
| of input is 1, the shape will be identical
| to that of the input at run time with optional
| additional dimensions appended at
| the end as specified by ‘extra_shape’
| argument. In that case the ‘shape’ argument
| should not be set.
|
| If input_as_shape is set to true, then
| the input should be a 1D tensor containing
| the desired output shape (the dimensions
| specified in extra_shape will also
| be appended)
|
| ———–
| @note
|
| Currently, it supports data type of
| float, int32, int64, and bool.
|
| FillerOp takes in either zero or one
| input.
|
| If the number of input is 1, the shape
| will be identical to that of the input
| at run time with optional additional
| dimensions appended at the end as specified
| by “extra_shape” argument. In that
| case the “shape” parameter should not
| be set.
|
| If the number of inputs is 0, the full
| shape must be provided via “shape” argument
|
| This op fills an output tensor with samples drawn
| from a normal distribution specified by the mean
| and standard deviation arguments. The output
| tensor shape is specified by the shape argument.
|
| However, if input_as_shape is set to true,
| then the input should be a 1D tensor containing
| the desired output shape (the dimensions specified
| in extra_shape will also be appended). In this
| case, the shape argument should not be set.
|
| Note: cannot set the shape argument and pass in
| an input at the same time.
|
| Github Links:
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/filler_op.h
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/filler_op.cc
| This op fills a uint8 output tensor with
| the data specified by the value argument.
| The data must previously be serialized
| as a byte string. The output tensor shape
| is specified by the shape argument.
| Beware, when using this argument value
| should have a value for every element
| of the output*, as missing values will
| not be initialized automatically.
| If input_as_shape is set to true,
| then the input should be a 1D tensor
| containing the desired output shape
| (the dimensions specified in extra_shape
| will also be appended). In this case,
| the shape argument should not
| be set.
|
| This op allows us to write uint8 tensors
| to
|
| Protobuf as byte strings and read them
| back as uint8 tensors in order to avoid
| the Protobuf uint32_t varint encoding
| size penalty.
|
| This op fills an output tensor with the
| data specified by the value and dtype
| arguments.
|
| The output tensor shape is specified
| by the shape* argument. Beware, when
| using this argument value* should have
| a value for every element of the output,
| as missing values will not be initialized
| automatically. If input_as_shape
| is set to true, then the input should
| be a 1D tensor containing the desired
| output shape (the dimensions specified
| in extra_shape will also be appended).
| In this case, the shape argument should
| not be set.
|
| ———–
| @note
|
| Do not set the shape argument and pass
| in an input at the same time.*
|
| Github Links:
|
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/given_tensor_fill_op.h
|
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/given_tensor_fill_op.cc
|
| The LengthsRangeFill op takes a single
| input lengths* and outputs a single
| tensor range_sequence*. For each element
| of lengths, the op appends the range(0,lengths)
| vector to the end of range_sequence.
| For example, if input=[2,4,1], the
| output would be [0,1,0,1,2,3,0].
|
| Github Links:
|
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/filler_op.h
|
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/filler_op.cc
|
| This is mostly used just as a debugging
| purpose stuff: it fills a tensor sequentially
| with values 0, 1, 2…, which can then be used
| to check e.g. reshape operations by allowing
| one to read the indices more easily.
| Fill the output tensor with float samples from
| uniform distribution [min, max].
|
| - The range can be defined either by arguments or
| input blobs. min and max are inclusive.
|
| - If the range is given by input blobs, you
| also need to give the shape as input.
|
| - When the range is given as arguments, this
| operator enforces min <= max. When the range is
| given as inputs, the constraint is not enforced.
|
| - When the range is given as inputs and max
| < min, the first dimension of the output is set to
| 0. This behavior is allowed so that dynamically
| sampling indices into a dynamically sized tensor
| is possible.
|
| - The shape of the output can be given as argument
| or input.
|
| Github Links:
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/filler_op.h
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/filler_op.cc
|
| Fill the output tensor with uniform
| samples between min and max (inclusive).
|
| If the second input is given, its elements
| will be excluded from uniform sampling.
| Using the second input will require
| you to provide shape via the first input.
|
| This op fills an output tensor with values
| sampled from a uniform distribution
| with the range determined by the desired
| shape of the output.
|
| Rather, than specifying the range of
| values manually, the novelty of Xavier
| Fill is that it automatically scales
| the range of the distribution it draws
| from based on the size of the desired
| output tensor.
|
| For more information check out the paper
|
Understanding the difficulty of training
| deep feedforward neural networks.
| The output tensor shape is specified
| by the
shape argument.
|
| However, if
input_as_shape is set
| to
true, then the
input should be
| a 1D tensor containing the desired output
| shape (the dimensions specified in
|
extra_shape will also be appended).
| In this case, the
shape argument should
|
not be set.
|
| ———–
| @note
|
| Do not set the shape argument and pass
| in an input at the same time.*
|
| Github Links:
|
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/filler_op.h
|
| - https://github.com/caffe2/caffe2/blob/master/caffe2/operators/filler_op.cc
|