JTh] S SK Jr S SKrS SKJr S SKJrJr S SKJ r SSK
J r SS /r " S
S\ 5 r
" S S \ 5 rg) )OptionalN)Tensor)
functionalinit) Parameter )Module Embedding EmbeddingBagc D ^ \ rS rSr% Sr/ SQr\\S' \\S' \\ \S' \\ \S' \ \S' \
\S ' \ \S
' \
\S ' \
\S ' SS\S\S\\ S\\ S\ S \
S \
S\\ S\
SS
4U 4S jjjr SS jr
SS jrS\ S\ 4S jrS\4S jr\ SS j5 rSrU =r$ ) r
a] A simple lookup table that stores embeddings of a fixed dictionary and size.
This module is often used to store word embeddings and retrieve them using indices.
The input to the module is a list of indices, and the output is the corresponding
word embeddings.
Args:
num_embeddings (int): size of the dictionary of embeddings
embedding_dim (int): the size of each embedding vector
padding_idx (int, optional): If specified, the entries at :attr:`padding_idx` do not contribute to the gradient;
therefore, the embedding vector at :attr:`padding_idx` is not updated during training,
i.e. it remains as a fixed "pad". For a newly constructed Embedding,
the embedding vector at :attr:`padding_idx` will default to all zeros,
but can be updated to another value to be used as the padding vector.
max_norm (float, optional): If given, each embedding vector with norm larger than :attr:`max_norm`
is renormalized to have norm :attr:`max_norm`.
norm_type (float, optional): The p of the p-norm to compute for the :attr:`max_norm` option. Default ``2``.
scale_grad_by_freq (bool, optional): If given, this will scale gradients by the inverse of frequency of
the words in the mini-batch. Default ``False``.
sparse (bool, optional): If ``True``, gradient w.r.t. :attr:`weight` matrix will be a sparse tensor.
See Notes for more details regarding sparse gradients.
Attributes:
weight (Tensor): the learnable weights of the module of shape (num_embeddings, embedding_dim)
initialized from :math:`\mathcal{N}(0, 1)`
Shape:
- Input: :math:`(*)`, IntTensor or LongTensor of arbitrary shape containing the indices to extract
- Output: :math:`(*, H)`, where `*` is the input shape and :math:`H=\text{embedding\_dim}`
.. note::
Keep in mind that only a limited number of optimizers support
sparse gradients: currently it's :class:`optim.SGD` (`CUDA` and `CPU`),
:class:`optim.SparseAdam` (`CUDA` and `CPU`) and :class:`optim.Adagrad` (`CPU`)
.. note::
When :attr:`max_norm` is not ``None``, :class:`Embedding`'s forward method will modify the
:attr:`weight` tensor in-place. Since tensors needed for gradient computations cannot be
modified in-place, performing a differentiable operation on ``Embedding.weight`` before
calling :class:`Embedding`'s forward method requires cloning ``Embedding.weight`` when
:attr:`max_norm` is not ``None``. For example::
n, d, m = 3, 5, 7
embedding = nn.Embedding(n, d, max_norm=1.0)
W = torch.randn((m, d), requires_grad=True)
idx = torch.tensor([1, 2])
a = embedding.weight.clone() @ W.t() # weight must be cloned for this to be differentiable
b = embedding(idx) @ W.t() # modifies weight in-place
out = (a.unsqueeze(0) + b.unsqueeze(1))
loss = out.sigmoid().prod()
loss.backward()
Examples::
>>> # an Embedding module containing 10 tensors of size 3
>>> embedding = nn.Embedding(10, 3)
>>> # a batch of 2 samples of 4 indices each
>>> input = torch.LongTensor([[1, 2, 4, 5], [4, 3, 2, 9]])
>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> embedding(input)
tensor([[[-0.0251, -1.6902, 0.7172],
[-0.6431, 0.0748, 0.6969],
[ 1.4970, 1.3448, -0.9685],
[-0.3677, -2.7265, -0.1685]],
[[ 1.4970, 1.3448, -0.9685],
[ 0.4362, -0.4004, 0.9400],
[-0.6431, 0.0748, 0.6969],
[ 0.9124, -2.3616, 1.1151]]])
>>> # example with padding_idx
>>> embedding = nn.Embedding(10, 3, padding_idx=0)
>>> input = torch.LongTensor([[0, 2, 0, 5]])
>>> embedding(input)
tensor([[[ 0.0000, 0.0000, 0.0000],
[ 0.1535, -2.0309, 0.9315],
[ 0.0000, 0.0000, 0.0000],
[-0.1655, 0.9897, 0.0635]]])
>>> # example of changing `pad` vector
>>> padding_idx = 0
>>> embedding = nn.Embedding(3, 3, padding_idx=padding_idx)
>>> embedding.weight
Parameter containing:
tensor([[ 0.0000, 0.0000, 0.0000],
[-0.7895, -0.7089, -0.0364],
[ 0.6778, 0.5803, 0.2678]], requires_grad=True)
>>> with torch.no_grad():
... embedding.weight[padding_idx] = torch.ones(3)
>>> embedding.weight
Parameter containing:
tensor([[ 1.0000, 1.0000, 1.0000],
[-0.7895, -0.7089, -0.0364],
[ 0.6778, 0.5803, 0.2678]], requires_grad=True)
)num_embeddings
embedding_dim padding_idxmax_norm norm_typescale_grad_by_freqsparser r r r r r weightfreezer N_weight_freezereturnc > XS.n [ T
U ] 5 Xl X l UbI US: a X0R : d S5 eO,US: a& X0R * : d S5 eU R U- nX0l X@l XPl X`l Uc; [ [ R " X440 U D6U ( + S9U l U R 5 O5[ UR 5 UU/:X d S5 e[ X( + S9U l Xpl g )Ndevicedtyper z)Padding_idx must be within num_embeddings)
requires_grad?Shape of weight does not match num_embeddings and embedding_dim)super__init__r r r r r r r torchemptyr reset_parameterslistshaper )selfr r r r r r r r r r r factory_kwargs __class__s O/var/www/auris/envauris/lib/python3.13/site-packages/torch/nn/modules/sparse.pyr! Embedding.__init__ s %+;
, * "Q "5"55?>?5 q $7$7#77?>?7 " 1 1K ? &
""4 ?# ^ ;N~N")kDK
! ! #
& +
Q Q
Q $G;GDK c d [ R " U R 5 U R 5 g Nr normal_r _fill_padding_idx_with_zeror' s r* r$ Embedding.reset_parameters T[[! ((*r, c U R bG [ R " 5 U R U R R S5 S S S 5 g g ! , ( d f g = fNr r r" no_gradr fill_r2 s r* r1 %Embedding._fill_padding_idx_with_zero F ' D , ,-33A6 ! ( )A
A$inputc [ R " UU R U R U R U R
U R U R 5 $ r. )F embeddingr r r r r r )r' r= s r* forwardEmbedding.forward sD {{ KK MM NN # # KK
r, c SnU R b US-
nU R b US-
nU R S:w a US-
nU R SLa US-
nU R SLa US-
nUR
" S 0 U R D6$ )
N!{num_embeddings}, {embedding_dim}, padding_idx={padding_idx}, max_norm={max_norm} , norm_type={norm_type}F), scale_grad_by_freq={scale_grad_by_freq}z
, sparse=True )r r r r r format__dict__)r' ss r*
extra_reprEmbedding.extra_repr s / '
. .A == $
( (A >>Q
* *A " "% /
< >> # FloatTensor containing pretrained weights
>>> weight = torch.FloatTensor([[1, 2.3, 3], [4, 5.1, 6.3]])
>>> embedding = nn.Embedding.from_pretrained(weight)
>>> # Get embeddings for index 1
>>> input = torch.LongTensor([1])
>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> embedding(input)
tensor([[ 4.0000, 5.1000, 6.3000]])
rG 4Embeddings parameter is expected to be 2-dimensional) r r r r r r r r r )dimr& ) cls
embeddingsr r r r r r rowscolsr@ s r* from_pretrainedEmbedding.from_pretrained s` L
NN ! B A B ! %%
# 1
r, )r r r r r r r r ) NN @FFNFNNr N)TNNrY FF)__name__
__module__ __qualname____firstlineno____doc__
__constants__int__annotations__r floatboolr r! r$ r1 rA strrN classmethodrW __static_attributes__
__classcell__r) s @r* r
r
s6 _BM # uo N L L &*$( #( $( - - - c] -
5/ -
- !- - &!- -
- - ^+7
V
)C ) 3 3r, c ^ \ rS rSr% Sr/ SQr\\S' \\S' \\ \S' \ \S' \
\S' \ \S ' \ \S
' \
\S ' \
\S ' \\ \S
' S S\S\S\\ S\ S\
S
\ S \
S\\ S \
S
\\ SS4U 4S jjjr
S S jrS S jr S S\ S\\ S\\ S\ 4S jjrS\ 4S jr\ S S\ S\
S\\ S\ S\
S
\ S \
S \
S
\\ SS 4S jj5 rS rU =r$ )!r i
aH Compute sums or means of 'bags' of embeddings, without instantiating the intermediate embeddings.
For bags of constant length, no :attr:`per_sample_weights`, no indices equal to :attr:`padding_idx`,
and with 2D inputs, this class
* with ``mode="sum"`` is equivalent to :class:`~torch.nn.Embedding` followed by ``torch.sum(dim=1)``,
* with ``mode="mean"`` is equivalent to :class:`~torch.nn.Embedding` followed by ``torch.mean(dim=1)``,
* with ``mode="max"`` is equivalent to :class:`~torch.nn.Embedding` followed by ``torch.max(dim=1)``.
However, :class:`~torch.nn.EmbeddingBag` is much more time and memory efficient than using a chain of these
operations.
EmbeddingBag also supports per-sample weights as an argument to the forward
pass. This scales the output of the Embedding before performing a weighted
reduction as specified by ``mode``. If :attr:`per_sample_weights` is passed, the
only supported ``mode`` is ``"sum"``, which computes a weighted sum according to
:attr:`per_sample_weights`.
Args:
num_embeddings (int): size of the dictionary of embeddings
embedding_dim (int): the size of each embedding vector
max_norm (float, optional): If given, each embedding vector with norm larger than :attr:`max_norm`
is renormalized to have norm :attr:`max_norm`.
norm_type (float, optional): The p of the p-norm to compute for the :attr:`max_norm` option. Default ``2``.
scale_grad_by_freq (bool, optional): if given, this will scale gradients by the inverse of frequency of
the words in the mini-batch. Default ``False``.
Note: this option is not supported when ``mode="max"``.
mode (str, optional): ``"sum"``, ``"mean"`` or ``"max"``. Specifies the way to reduce the bag.
``"sum"`` computes the weighted sum, taking :attr:`per_sample_weights`
into consideration. ``"mean"`` computes the average of the values
in the bag, ``"max"`` computes the max value over each bag.
Default: ``"mean"``
sparse (bool, optional): if ``True``, gradient w.r.t. :attr:`weight` matrix will be a sparse tensor. See
Notes for more details regarding sparse gradients. Note: this option is not
supported when ``mode="max"``.
include_last_offset (bool, optional): if ``True``, :attr:`offsets` has one additional element, where the last element
is equivalent to the size of `indices`. This matches the CSR format.
padding_idx (int, optional): If specified, the entries at :attr:`padding_idx` do not contribute to the
gradient; therefore, the embedding vector at :attr:`padding_idx` is not updated
during training, i.e. it remains as a fixed "pad". For a newly constructed
EmbeddingBag, the embedding vector at :attr:`padding_idx` will default to all
zeros, but can be updated to another value to be used as the padding vector.
Note that the embedding vector at :attr:`padding_idx` is excluded from the
reduction.
Attributes:
weight (Tensor): the learnable weights of the module of shape `(num_embeddings, embedding_dim)`
initialized from :math:`\mathcal{N}(0, 1)`.
Examples::
>>> # an EmbeddingBag module containing 10 tensors of size 3
>>> embedding_sum = nn.EmbeddingBag(10, 3, mode='sum')
>>> # a batch of 2 samples of 4 indices each
>>> input = torch.tensor([1, 2, 4, 5, 4, 3, 2, 9], dtype=torch.long)
>>> offsets = torch.tensor([0, 4], dtype=torch.long)
>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> embedding_sum(input, offsets)
tensor([[-0.8861, -5.4350, -0.0523],
[ 1.1306, -2.5798, -1.0044]])
>>> # Example with padding_idx
>>> embedding_sum = nn.EmbeddingBag(10, 3, mode='sum', padding_idx=2)
>>> input = torch.tensor([2, 2, 2, 2, 4, 3, 2, 9], dtype=torch.long)
>>> offsets = torch.tensor([0, 4], dtype=torch.long)
>>> embedding_sum(input, offsets)
tensor([[ 0.0000, 0.0000, 0.0000],
[-0.7082, 3.2145, -2.6251]])
>>> # An EmbeddingBag can be loaded from an Embedding like so
>>> embedding = nn.Embedding(10, 3, padding_idx=2)
>>> embedding_sum = nn.EmbeddingBag.from_pretrained(
embedding.weight,
padding_idx=embedding.padding_idx,
mode='sum')
) r r r r r moder include_last_offsetr r r r r r r rk r rl r Nr r c
> XS.n
[ TU ] 5 Xl X l X0l X@l XPl U
bI U
S: a XR : d S5 eO,U
S: a& XR * : d S5 eU R U
- n
Xl Uc7 [ [ R " X440 U
D65 U l U R 5 O2[ UR 5 UU/:X d S5 e[ U5 U l X`l Xpl Xl g )Nr r z)padding_idx must be within num_embeddingsr )r r! r r r r r r r r" r# r r$ r% r& rk r rl )r' r r r r r rk r r rl r r r r( r) s r* r! EmbeddingBag.__init__r s %+;
, *
""4 "Q "5"55?>?5 q $7$7#77?>?7 " 1 1K ? & ?# ^ ;N~NDK
! ! #
& +
Q Q
Q $G,DK #6 r, c d [ R " U R 5 U R 5 g r. r/ r2 s r* r$ EmbeddingBag.reset_parameters r4 r, c U R bG [ R " 5 U R U R R S5 S S S 5 g g ! , ( d f g = fr6 r7 r2 s r* r1 (EmbeddingBag._fill_padding_idx_with_zero r; r< r= offsetsper_sample_weightsc
[ R " UU R UU R U R U R
U R U R UU R U R 5 $ )a+ Forward pass of EmbeddingBag.
Args:
input (Tensor): Tensor containing bags of indices into the embedding matrix.
offsets (Tensor, optional): Only used when :attr:`input` is 1D. :attr:`offsets` determines
the starting index position of each bag (sequence) in :attr:`input`.
per_sample_weights (Tensor, optional): a tensor of float / double weights, or None
to indicate all weights should be taken to be ``1``. If specified, :attr:`per_sample_weights`
must have exactly the same shape as input and is treated as having the same
:attr:`offsets`, if those are not ``None``. Only supported for ``mode='sum'``.
Returns:
Tensor output shape of `(B, embedding_dim)`.
.. note::
A few notes about ``input`` and ``offsets``:
- :attr:`input` and :attr:`offsets` have to be of the same type, either int or long
- If :attr:`input` is 2D of shape `(B, N)`, it will be treated as ``B`` bags (sequences)
each of fixed length ``N``, and this will return ``B`` values aggregated in a way
depending on the :attr:`mode`. :attr:`offsets` is ignored and required to be ``None`` in this case.
- If :attr:`input` is 1D of shape `(N)`, it will be treated as a concatenation of
multiple bags (sequences). :attr:`offsets` is required to be a 1D tensor containing the
starting index positions of each bag in :attr:`input`. Therefore, for :attr:`offsets` of shape `(B)`,
:attr:`input` will be viewed as having ``B`` bags. Empty bags (i.e., having 0-length) will have
returned vectors filled by zeros.
)
r?
embedding_bagr r r r rk r rl r )r' r= rs rt s r* rA EmbeddingBag.forward s] H KK MM NN # # II KK $ $
r, c H SnU R b US-
nU R S:w a US-
nU R SLa US-
nUS-
nU R b US-
nUR " S 0 U R
R
5 VVs0 s H u p#U[ U5 _M snnD6$ s snnf )
NrD rF rG rH FrI z
, mode={mode}rE rJ )r r r r rK rL itemsrepr)r' rM kvs r* rN EmbeddingBag.extra_repr s / == $
( (A >>Q
* *A " "% /
< >> # FloatTensor containing pretrained weights
>>> weight = torch.FloatTensor([[1, 2.3, 3], [4, 5.1, 6.3]])
>>> embeddingbag = nn.EmbeddingBag.from_pretrained(weight)
>>> # Get embeddings for index 1
>>> input = torch.LongTensor([[1, 0]])
>>> # xdoctest: +IGNORE_WANT("non-deterministic")
>>> embeddingbag(input)
tensor([[ 2.5000, 3.7000, 4.6500]])
rG rQ )
r r r r r r rk r rl r )rR r& r r )
rS rT r r r r rk r rl r rU rV embeddingbags
r* rW EmbeddingBag.from_pretrained ss P
NN ! B A B ! %%
1 3#
17J )r, )
r rl r rk r r r r r r )
NrY FmeanFNFNNNrZ )NN)TNrY Fr FFN)r[ r\ r] r^ r_ r` ra rb r rc rd r re r! r$ r1 rA rN rf rW rg rh ri s @r* r r
s KZ
M uo N
I L # %) #( $($)%).7.7 .7 5/ .7
.7 !
.7 .7 .7 &!.7 ".7 c].7
.7 .7`+7 %)/3 0
0
&!0
%V , 0
0
d JC J $( #( $)%)7 7 7 5/ 7
7 !
7 7 7 "7 c]7
7 7 r, )typingr r" r torch.nnr r? r torch.nn.parameterr moduler __all__r
r rJ r, r* r s@ * (
'{ {|U 6 U r,