48 KiB
title | teaser | source | menu | ||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Model Architectures | Pre-defined model architectures included with the core library | spacy/ml/models |
|
TODO: intro and how architectures work, link to
registry
,
custom functions usage etc.
Tok2Vec architectures
spacy.HashEmbedCNN.v1
Example Config
[model] @architectures = "spacy.HashEmbedCNN.v1" pretrained_vectors = null width = 96 depth = 4 embed_size = 2000 window_size = 1 maxout_pieces = 3 subword_features = true
Build spaCy's "standard" embedding layer, which uses hash embedding with subword features and a CNN with layer-normalized maxout.
Name | Description |
---|---|
width |
The width of the input and output. These are required to be the same, so that residual connections can be used. Recommended values are 96 , 128 or 300 . |
depth |
The number of convolutional layers to use. Recommended values are between 2 and 8 . |
embed_size |
The number of rows in the hash embedding tables. This can be surprisingly small, due to the use of the hash embeddings. Recommended values are between 2000 and 10000 . |
window_size |
The number of tokens on either side to concatenate during the convolutions. The receptive field of the CNN will be depth * (window_size * 2 + 1) , so a 4-layer network with a window size of 2 will be sensitive to 17 words at a time. Recommended value is 1 . |
maxout_pieces |
The number of pieces to use in the maxout non-linearity. If 1 , the Mish non-linearity is used instead. Recommended values are 1 -3 . |
subword_features |
Whether to also embed subword features, specifically the prefix, suffix and word shape. This is recommended for alphabetic languages like English, but not if single-character tokens are used for a language such as Chinese. |
pretrained_vectors |
Whether to also use static vectors. |
CREATES | The model using the architecture. |
spacy.Tok2Vec.v1
Example config
[model] @architectures = "spacy.Tok2Vec.v1" [model.embed] @architectures = "spacy.CharacterEmbed.v1" # ... [model.encode] @architectures = "spacy.MaxoutWindowEncoder.v1" # ...
Construct a tok2vec model out of embedding and encoding subnetworks. See the "Embed, Encode, Attend, Predict" blog post for background.
Name | Description |
---|---|
embed |
Embed tokens into context-independent word vector representations. For example, CharacterEmbed or MultiHashEmbed. |
encode |
Encode context into the embeddings, using an architecture such as a CNN, BiLSTM or transformer. For example, MaxoutWindowEncoder. |
CREATES | The model using the architecture. |
spacy.Tok2VecListener.v1
Example config
[components.tok2vec] factory = "tok2vec" [components.tok2vec.model] @architectures = "spacy.HashEmbedCNN.v1" width = 342 [components.tagger] factory = "tagger" [components.tagger.model] @architectures = "spacy.Tagger.v1" [components.tagger.model.tok2vec] @architectures = "spacy.Tok2VecListener.v1" width = ${components.tok2vec.model.width}
A listener is used as a sublayer within a component such as a
DependencyParser
,
EntityRecognizer
or
TextCategorizer
. Usually you'll have multiple
listeners connecting to a single upstream Tok2Vec
component
that's earlier in the pipeline. The listener layers act as proxies, passing
the predictions from the Tok2Vec
component into downstream components, and
communicating gradients back upstream.
Instead of defining its own Tok2Vec
instance, a model architecture like
Tagger can define a listener as its tok2vec
argument that connects to the shared tok2vec
component in the pipeline.
Name | Description |
---|---|
width |
The width of the vectors produced by the "upstream" Tok2Vec component. |
upstream |
A string to identify the "upstream" Tok2Vec component to communicate with. The upstream name should either be the wildcard string "*" , or the name of the Tok2Vec component. You'll almost never have multiple upstream Tok2Vec components, so the wildcard string will almost always be fine. |
CREATES | The model using the architecture. |
spacy.MultiHashEmbed.v1
Example config
[model] @architectures = "spacy.MultiHashEmbed.v1" width = 64 rows = 2000 also_embed_subwords = false also_use_static_vectors = false
Construct an embedding layer that separately embeds a number of lexical
attributes using hash embedding, concatenates the results, and passes it through
a feed-forward subnetwork to build a mixed representations. The features used
are the NORM
, PREFIX
, SUFFIX
and SHAPE
, which can have varying
definitions depending on the Vocab
of the Doc
object passed in. Vectors from
pretrained static vectors can also be incorporated into the concatenated
representation.
Name | Description |
---|---|
width |
The output width. Also used as the width of the embedding tables. Recommended values are between 64 and 300 . |
rows |
The number of rows for the embedding tables. Can be low, due to the hashing trick. Embeddings for prefix, suffix and word shape use half as many rows. Recommended values are between 2000 and 10000 . |
also_embed_subwords |
Whether to use the PREFIX , SUFFIX and SHAPE features in the embeddings. If not using these, you may need more rows in your hash embeddings, as there will be increased chance of collisions. |
also_use_static_vectors |
Whether to also use static word vectors. Requires a vectors table to be loaded in the Doc objects' vocab. |
CREATES | The model using the architecture. |
spacy.CharacterEmbed.v1
Example config
[model] @architectures = "spacy.CharacterEmbed.v1" width = 128 rows = 7000 nM = 64 nC = 8
Construct an embedded representations based on character embeddings, using a feed-forward network. A fixed number of UTF-8 byte characters are used for each word, taken from the beginning and end of the word equally. Padding is used in the center for words that are too short.
For instance, let's say nC=4
, and the word is "jumping". The characters used
will be "jung"
(two from the start, two from the end). If we had nC=8
, the
characters would be "jumpping"
: 4 from the start, 4 from the end. This ensures
that the final character is always in the last position, instead of being in an
arbitrary position depending on the word length.
The characters are embedded in a embedding table with 256 rows, and the vectors
concatenated. A hash-embedded vector of the NORM
of the word is also
concatenated on, and the result is then passed through a feed-forward network to
construct a single vector to represent the information.
Name | Description |
---|---|
width |
The width of the output vector and the NORM hash embedding. |
rows |
The number of rows in the NORM hash embedding table. |
nM |
The dimensionality of the character embeddings. Recommended values are between 16 and 64 . |
nC |
The number of UTF-8 bytes to embed per word. Recommended values are between 3 and 8 , although it may depend on the length of words in the language. |
CREATES | The model using the architecture. |
spacy.MaxoutWindowEncoder.v1
Example config
[model] @architectures = "spacy.MaxoutWindowEncoder.v1" width = 128 window_size = 1 maxout_pieces = 3 depth = 4
Encode context using convolutions with maxout activation, layer normalization and residual connections.
Name | Description |
---|---|
width |
The input and output width. These are required to be the same, to allow residual connections. This value will be determined by the width of the inputs. Recommended values are between 64 and 300 . |
window_size |
The number of words to concatenate around each token to construct the convolution. Recommended value is 1 . |
maxout_pieces |
The number of maxout pieces to use. Recommended values are 2 or 3 . |
depth |
The number of convolutional layers. Recommended value is 4 . |
CREATES | The model using the architecture. |
spacy.MishWindowEncoder.v1
Example config
[model] @architectures = "spacy.MishWindowEncoder.v1" width = 64 window_size = 1 depth = 4
Encode context using convolutions with
Mish
activation, layer normalization
and residual connections.
Name | Description |
---|---|
width |
The input and output width. These are required to be the same, to allow residual connections. This value will be determined by the width of the inputs. Recommended values are between 64 and 300 . |
window_size |
The number of words to concatenate around each token to construct the convolution. Recommended value is 1 . |
depth |
The number of convolutional layers. Recommended value is 4 . |
CREATES | The model using the architecture. |
spacy.TorchBiLSTMEncoder.v1
Example config
[model] @architectures = "spacy.TorchBiLSTMEncoder.v1" width = 64 window_size = 1 depth = 4
Encode context using bidirectional LSTM layers. Requires PyTorch.
Name | Description |
---|---|
width |
The input and output width. These are required to be the same, to allow residual connections. This value will be determined by the width of the inputs. Recommended values are between 64 and 300 . |
window_size |
The number of words to concatenate around each token to construct the convolution. Recommended value is 1 . |
depth |
The number of convolutional layers. Recommended value is 4 . |
CREATES | The model using the architecture. |
spacy.StaticVectors.v1
Example config
[model] @architectures = "spacy.StaticVectors.v1" nO = null nM = null dropout = 0.2 key_attr = "ORTH" [model.init_W] @initializers = "glorot_uniform_init.v1"
Embed Doc
objects with their vocab's vectors table, applying a
learned linear projection to control the dimensionality. See the documentation
on static vectors for details.
Name | Description |
---|---|
nO |
The output width of the layer, after the linear projection. |
nM |
The width of the static vectors. |
dropout |
Optional dropout rate. If set, it's applied per dimension over the whole batch. Defaults to None . |
init_W |
The initialization function. Defaults to glorot_uniform_init . |
key_attr |
Defaults to "ORTH" . |
CREATES | The model using the architecture. |
Transformer architectures
The following architectures are provided by the package
spacy-transformers
. See the
usage documentation for how to integrate the
architectures into your training config.
spacy-transformers.TransformerModel.v1
Example Config
[model] @architectures = "spacy-transformers.TransformerModel.v1" name = "roberta-base" tokenizer_config = {"use_fast": true} [model.get_spans] @span_getters = "strided_spans.v1" window = 128 stride = 96
Load and wrap a transformer model from the
HuggingFace transformers
library. You
can any transformer that has pretrained weights and a PyTorch implementation.
The name
variable is passed through to the underlying library, so it can be
either a string or a path. If it's a string, the pretrained weights will be
downloaded via the transformers library if they are not already available
locally.
In order to support longer documents, the
TransformerModel layer allows you to pass
in a get_spans
function that will divide up the Doc
objects
before passing them through the transformer. Your spans are allowed to overlap
or exclude tokens. This layer is usually used directly by the
Transformer
component, which allows you to share the
transformer weights across your pipeline. For a layer that's configured for use
in other components, see
Tok2VecTransformer.
Name | Description |
---|---|
name |
Any model name that can be loaded by transformers.AutoModel . |
get_spans |
Function that takes a batch of Doc object and returns lists of Span objects to process by the transformer. See here for built-in options and examples. |
tokenizer_config |
Tokenizer settings passed to transformers.AutoTokenizer . |
CREATES | The model using the architecture. |
spacy-transformers.Tok2VecListener.v1
Example Config
[model] @architectures = "spacy-transformers.Tok2VecListener.v1" grad_factor = 1.0 [model.pooling] @layers = "reduce_mean.v1"
Create a TransformerListener
layer, which will connect to a
Transformer
component earlier in the pipeline. The layer
takes a list of Doc
objects as input, and produces a list of
2-dimensional arrays as output, with each array having one row per token. Most
spaCy models expect a sublayer with this signature, making it easy to connect
them to a transformer model via this sublayer. Transformer models usually
operate over wordpieces, which usually don't align one-to-one against spaCy
tokens. The layer therefore requires a reduction operation in order to calculate
a single token vector given zero or more wordpiece vectors.
Name | Description |
---|---|
pooling |
A reduction layer used to calculate the token vectors based on zero or more wordpiece vectors. If in doubt, mean pooling (see reduce_mean ) is usually a good choice. |
grad_factor |
Reweight gradients from the component before passing them upstream. You can set this to 0 to "freeze" the transformer weights with respect to the component, or use it to make some components more significant than others. Leaving it at 1.0 is usually fine. |
CREATES | The model using the architecture. |
spacy-transformers.Tok2VecTransformer.v1
Example Config
[model] @architectures = "spacy.Tok2VecTransformer.v1" name = "albert-base-v2" tokenizer_config = {"use_fast": false} grad_factor = 1.0
Use a transformer as a Tok2Vec
layer directly. This does
not allow multiple components to share the transformer weights, and does
not allow the transformer to set annotations into the Doc
object, but it's a simpler solution if you only need the transformer within
one component.
Name | Description |
---|---|
get_spans |
Function that takes a batch of Doc object and returns lists of Span objects to process by the transformer. See here for built-in options and examples. |
tokenizer_config |
Tokenizer settings passed to transformers.AutoTokenizer . |
pooling |
A reduction layer used to calculate the token vectors based on zero or more wordpiece vectors. If in doubt, mean pooling (see reduce_mean ) is usually a good choice. |
grad_factor |
Reweight gradients from the component before passing them upstream. You can set this to 0 to "freeze" the transformer weights with respect to the component, or use it to make some components more significant than others. Leaving it at 1.0 is usually fine. |
CREATES | The model using the architecture. |
Parser & NER architectures
spacy.TransitionBasedParser.v1
Example Config
[model] @architectures = "spacy.TransitionBasedParser.v1" nr_feature_tokens = 6 hidden_width = 64 maxout_pieces = 2 [model.tok2vec] @architectures = "spacy.HashEmbedCNN.v1" pretrained_vectors = null width = 96 depth = 4 embed_size = 2000 window_size = 1 maxout_pieces = 3 subword_features = true
Build a transition-based parser model. Can apply to NER or dependency parsing. Transition-based parsing is an approach to structured prediction where the task of predicting the structure is mapped to a series of state transitions. You might find this tutorial helpful for background information. The neural network state prediction model consists of either two or three subnetworks:
- tok2vec: Map each token into a vector representations. This subnetwork is run once for each batch.
- lower: Construct a feature-specific vector for each
(token, feature)
pair. This is also run once for each batch. Constructing the state representation is then simply a matter of summing the component features and applying the non-linearity. - upper (optional): A feed-forward network that predicts scores from the state representation. If not present, the output from the lower model is used as action scores directly.
Name | Description |
---|---|
tok2vec |
Subnetwork to map tokens into vector representations. |
nr_feature_tokens |
The number of tokens in the context to use to construct the state vector. Valid choices are 1 , 2 , 3 , 6 , 8 and 13 . The 2 , 8 and 13 feature sets are designed for the parser, while the 3 and 6 feature sets are designed for the entity recognizer. The recommended feature sets are 3 for NER, and 8 for the dependency parser. |
hidden_width |
The width of the hidden layer. |
maxout_pieces |
How many pieces to use in the state prediction layer. Recommended values are 1 , 2 or 3 . If 1 , the maxout non-linearity is replaced with a Relu non-linearity if use_upper is True , and no non-linearity if False . |
use_upper |
Whether to use an additional hidden layer after the state vector in order to predict the action scores. It is recommended to set this to False for large pretrained models such as transformers, and True for smaller networks. The upper layer is computed on CPU, which becomes a bottleneck on larger GPU-based models, where it's also less necessary. |
nO |
The number of actions the model will predict between. Usually inferred from data at the beginning of training, or loaded from disk. |
CREATES | The model using the architecture. |
spacy.BILUOTagger.v1
Example Config
[model] @architectures = "spacy.BILUOTagger.v1 " [model.tok2vec] @architectures = "spacy.HashEmbedCNN.v1" # etc.
Construct a simple NER tagger that predicts
BILUO tag scores for each token and
uses greedy decoding with transition-constraints to return a valid BILUO tag
sequence. A BILUO tag sequence encodes a sequence of non-overlapping labelled
spans into tags assigned to each token. The first token of a span is given the
tag B-LABEL
, the last token of the span is given the tag L-LABEL
, and tokens
within the span are given the tag U-LABEL
. Single-token spans are given the
tag U-LABEL
. All other tokens are assigned the tag O
. The BILUO tag scheme
generally results in better linear separation between classes, especially for
non-CRF models, because there are more distinct classes for the different
situations (Ratinov et al., 2009).
Name | Description |
---|---|
tok2vec |
Subnetwork to map tokens into vector representations. |
CREATES | The model using the architecture. |
spacy.IOBTagger.v1
Example Config
[model] @architectures = "spacy.IOBTagger.v1 " [model.tok2vec] @architectures = "spacy.HashEmbedCNN.v1" # etc.
Construct a simple NER tagger, that predicts IOB tag scores for each token and uses greedy decoding with transition-constraints to return a valid IOB tag sequence. An IOB tag sequence encodes a sequence of non-overlapping labeled spans into tags assigned to each token. The first token of a span is given the tag B-LABEL, and subsequent tokens are given the tag I-LABEL. All other tokens are assigned the tag O.
Name | Description |
---|---|
tok2vec |
Subnetwork to map tokens into vector representations. |
CREATES | The model using the architecture. |
Tagging architectures
spacy.Tagger.v1
Example Config
[model] @architectures = "spacy.Tagger.v1" nO = null [model.tok2vec] # ...
Build a tagger model, using a provided token-to-vector component. The tagger model simply adds a linear layer with softmax activation to predict scores given the token vectors.
Name | Description |
---|---|
tok2vec |
Subnetwork to map tokens into vector representations. |
nO |
The number of tags to output. Inferred from the data if None . |
CREATES | The model using the architecture. |
Text classification architectures
A text classification architecture needs to take a Doc
as input,
and produce a score for each potential label class. Textcat challenges can be
binary (e.g. sentiment analysis) or involve multiple possible labels.
Multi-label challenges can either have mutually exclusive labels (each example
has exactly one label), or multiple labels may be applicable at the same time.
As the properties of text classification problems can vary widely, we provide several different built-in architectures. It is recommended to experiment with different architectures and settings to determine what works best on your specific data and challenge.
spacy.TextCatEnsemble.v1
Example Config
[model] @architectures = "spacy.TextCatEnsemble.v1" exclusive_classes = false pretrained_vectors = null width = 64 embed_size = 2000 conv_depth = 2 window_size = 1 ngram_size = 1 dropout = null nO = null
Stacked ensemble of a bag-of-words model and a neural network model. The neural network has an internal CNN Tok2Vec layer and uses attention.
Name | Description |
---|---|
exclusive_classes |
Whether or not categories are mutually exclusive. |
pretrained_vectors |
Whether or not pretrained vectors will be used in addition to the feature vectors. |
width |
Output dimension of the feature encoding step. |
embed_size |
Input dimension of the feature encoding step. |
conv_depth |
Depth of the tok2vec layer. |
window_size |
The number of contextual vectors to concatenate from the left and from the right. |
ngram_size |
Determines the maximum length of the n-grams in the BOW model. For instance, ngram_size=3 would give unigram, trigram and bigram features. |
dropout |
The dropout rate. |
nO |
Output dimension, determined by the number of different labels. If not set, the TextCategorizer component will set it when begin_training is called. |
CREATES | The model using the architecture. |
spacy.TextCatCNN.v1
Example Config
[model] @architectures = "spacy.TextCatCNN.v1" exclusive_classes = false nO = null [model.tok2vec] @architectures = "spacy.HashEmbedCNN.v1" pretrained_vectors = null width = 96 depth = 4 embed_size = 2000 window_size = 1 maxout_pieces = 3 subword_features = true
A neural network model where token vectors are calculated using a CNN. The vectors are mean pooled and used as features in a feed-forward network. This architecture is usually less accurate than the ensemble, but runs faster.
Name | Description |
---|---|
exclusive_classes |
Whether or not categories are mutually exclusive. |
tok2vec |
The tok2vec layer of the model. |
nO |
Output dimension, determined by the number of different labels. If not set, the TextCategorizer component will set it when begin_training is called. |
CREATES | The model using the architecture. |
spacy.TextCatBOW.v1
Example Config
[model] @architectures = "spacy.TextCatBOW.v1" exclusive_classes = false ngram_size = 1 no_output_layer = false nO = null
An ngram "bag-of-words" model. This architecture should run much faster than the others, but may not be as accurate, especially if texts are short.
Name | Description |
---|---|
exclusive_classes |
Whether or not categories are mutually exclusive. |
ngram_size |
Determines the maximum length of the n-grams in the BOW model. For instance, ngram_size=3 would give unigram, trigram and bigram features. |
no_output_layer |
Whether or not to add an output layer to the model (Softmax activation if exclusive_classes is True , else Logistic . |
nO |
Output dimension, determined by the number of different labels. If not set, the TextCategorizer component will set it when begin_training is called. |
CREATES | The model using the architecture. |
Entity linking architectures
An EntityLinker
component disambiguates textual mentions
(tagged as named entities) to unique identifiers, grounding the named entities
into the "real world". This requires 3 main components:
- A
KnowledgeBase
(KB) holding the unique identifiers, potential synonyms and prior probabilities. - A candidate generation step to produce a set of likely identifiers, given a certain textual mention.
- A Machine learning
Model
that picks the most plausible ID from the set of candidates.
spacy.EntityLinker.v1
Example Config
[model] @architectures = "spacy.EntityLinker.v1" nO = null [model.tok2vec] @architectures = "spacy.HashEmbedCNN.v1" pretrained_vectors = null width = 96 depth = 2 embed_size = 300 window_size = 1 maxout_pieces = 3 subword_features = true [kb_loader] @assets = "spacy.EmptyKB.v1" entity_vector_length = 64 [get_candidates] @assets = "spacy.CandidateGenerator.v1"
The EntityLinker
model architecture is a Thinc Model
with a
Linear
output layer.
Name | Description |
---|---|
tok2vec |
The tok2vec layer of the model. |
nO |
Output dimension, determined by the length of the vectors encoding each entity in the KB. If the nO dimension is not set, the entity linking component will set it when begin_training is called. |
CREATES | The model using the architecture. |
spacy.EmptyKB.v1
A function that creates a default, empty KnowledgeBase
from a
Vocab
instance.
Name | Description |
---|---|
entity_vector_length |
The length of the vectors encoding each entity in the KB. Defaults to 64 . |
spacy.CandidateGenerator.v1
A function that takes as input a KnowledgeBase
and a
Span
object denoting a named entity, and returns a list of
plausible Candidate
objects. The default
CandidateGenerator
simply uses the text of a mention to find its potential
aliases in the KnowledgeBase
. Note that this function is case-dependent.