Commit Graph

53 Commits

Author SHA1 Message Date
Matthew Honnibal
34bbad8e0e Add __reduce__ methods on parser subclasses. Fixes pickling. 2017-05-27 15:46:06 -05:00
Matthew Honnibal
467bbeadb8 Add hidden layers for tagger 2017-05-24 20:09:51 -05:00
Matthew Honnibal
5b67bcbee0 Increase default embed size to 7500 2017-05-23 15:20:16 -05:00
Matthew Honnibal
3959d778ac Revert "Revert "WIP on improving parser efficiency""
This reverts commit 532afef4a8.
2017-05-23 03:06:53 -05:00
Matthew Honnibal
532afef4a8 Revert "WIP on improving parser efficiency"
This reverts commit bdaac7ab44.
2017-05-23 03:05:25 -05:00
Matthew Honnibal
bdaac7ab44 WIP on improving parser efficiency 2017-05-23 02:59:31 -05:00
Matthew Honnibal
a7ee63c0ac Fix labeller loss for unseen labels 2017-05-22 10:41:20 -05:00
Matthew Honnibal
83ffd16474 Fix offset calculation for other negative values 2017-05-22 08:00:53 -05:00
Matthew Honnibal
b45b4aa392 PseudoProjectivity --> nonproj 2017-05-22 05:17:44 -05:00
Matthew Honnibal
8d1e64be69 Add experimental NeuralLabeller 2017-05-22 04:51:08 -05:00
Matthew Honnibal
9b1b0742fd Fix prediction for tok2vec 2017-05-22 04:51:08 -05:00
Matthew Honnibal
5db89053aa Merge docstrings 2017-05-21 13:46:23 -05:00
Matthew Honnibal
180e5afede Fix tokvecs flattening in pipeline 2017-05-21 09:05:34 -05:00
ines
99b631617d Reformat docstrings 2017-05-21 13:32:15 +02:00
ines
d82ae9a585 Change "function" to "callable" in docs 2017-05-21 13:17:40 +02:00
Matthew Honnibal
3b7c108246 Pass tokvecs through as a list, instead of concatenated. Also fix padding 2017-05-20 13:23:32 -05:00
Matthew Honnibal
d52b65aec2 Revert "Move to contiguous buffer for token_ids and d_vectors"
This reverts commit 3ff8c35a79.
2017-05-20 11:26:23 -05:00
Matthew Honnibal
3ff8c35a79 Move to contiguous buffer for token_ids and d_vectors 2017-05-20 04:17:30 -05:00
Matthew Honnibal
c12ab47a56 Remove state argument in pipeline. Other changes 2017-05-19 13:26:36 -05:00
ines
0fc05e54e4 Document TokenVectorEncoder 2017-05-19 00:00:02 +02:00
Matthew Honnibal
c2c825127a Fix use_params and pipe methods 2017-05-18 08:30:59 -05:00
Matthew Honnibal
b460533827 Bug fixes to pipeline 2017-05-18 04:29:51 -05:00
Matthew Honnibal
692bd2a186 Bug fix to tagger: wasnt backproping to token vectors 2017-05-17 13:13:14 +02:00
Matthew Honnibal
793430aa7a Get spaCy train command working with neural network
* Integrate models into pipeline
* Add basic serialization (maybe incorrect)
* Fix pickle on vocab
2017-05-17 12:04:50 +02:00
Matthew Honnibal
8cf097ca88 Redesign training to integrate NN components
* Obsolete .parser, .entity etc names in favour of .pipeline
* Components no longer create models on initialization
* Models created by loading method (from_disk(), from_bytes() etc), or
    .begin_training()
* Add .predict(), .set_annotations() methods in components
* Pass state through pipeline, to allow components to share information
    more flexibly.
2017-05-16 16:17:30 +02:00
Matthew Honnibal
5211645af3 Get data flowing through pipeline. Needs redesign 2017-05-16 11:21:59 +02:00
Matthew Honnibal
a9edb3aa1d Improve integration of NN parser, to support unified training API 2017-05-15 21:53:27 +02:00
Matthew Honnibal
4b9d69f428 Merge branch 'v2' into develop
* Move v2 parser into nn_parser.pyx
* New TokenVectorEncoder class in pipeline.pyx
* New spacy/_ml.py module

Currently the two parsers live side-by-side, until we figure out how to
organize them.
2017-05-14 01:10:23 +02:00
Matthew Honnibal
5cac951a16 Move new parser to nn_parser.pyx, and restore old parser, to make tests pass. 2017-05-14 00:55:01 +02:00
Matthew Honnibal
613ba79e2e Fiddle with sizings for parser 2017-05-13 17:20:23 -05:00
Matthew Honnibal
827b5af697 Update draft of parser neural network model
Model is good, but code is messy. Currently requires Chainer, which may cause the build to fail on machines without a GPU.

Outline of the model:

We first predict context-sensitive vectors for each word in the input:

(embed_lower | embed_prefix | embed_suffix | embed_shape)
>> Maxout(token_width)
>> convolution ** 4

This convolutional layer is shared between the tagger and the parser. This prevents the parser from needing tag features.
To boost the representation, we make a "super tag" with POS, morphology and dependency label. The tagger predicts this
by adding a softmax layer onto the convolutional layer --- so, we're teaching the convolutional layer to give us a
representation that's one affine transform from this informative lexical information. This is obviously good for the
parser (which backprops to the convolutions too).

The parser model makes a state vector by concatenating the vector representations for its context tokens. Current
results suggest few context tokens works well. Maybe this is a bug.

The current context tokens:

* S0, S1, S2: Top three words on the stack
* B0, B1: First two words of the buffer
* S0L1, S0L2: Leftmost and second leftmost children of S0
* S0R1, S0R2: Rightmost and second rightmost children of S0
* S1L1, S1L2, S1R2, S1R, B0L1, B0L2: Likewise for S1 and B0

This makes the state vector quite long: 13*T, where T is the token vector width (128 is working well). Fortunately,
there's a way to structure the computation to save some expense (and make it more GPU friendly).

The parser typically visits 2*N states for a sentence of length N (although it may visit more, if it back-tracks
with a non-monotonic transition). A naive implementation would require 2*N (B, 13*T) @ (13*T, H) matrix multiplications
for a batch of size B. We can instead perform one (B*N, T) @ (T, 13*H) multiplication, to pre-compute the hidden
weights for each positional feature wrt the words in the batch. (Note that our token vectors come from the CNN
-- so we can't play this trick over the vocabulary. That's how Stanford's NN parser works --- and why its model
is so big.)

This pre-computation strategy allows a nice compromise between GPU-friendliness and implementation simplicity.
The CNN and the wide lower layer are computed on the GPU, and then the precomputed hidden weights are moved
to the CPU, before we start the transition-based parsing process. This makes a lot of things much easier.
We don't have to worry about variable-length batch sizes, and we don't have to implement the dynamic oracle
in CUDA to train.

Currently the parser's loss function is multilabel log loss, as the dynamic oracle allows multiple states to
be 0 cost. This is defined as:

(exp(score) / Z) - (exp(score) / gZ)

Where gZ is the sum of the scores assigned to gold classes. I'm very interested in regressing on the cost directly,
but so far this isn't working well.

Machinery is in place for beam-search, which has been working well for the linear model. Beam search should benefit
greatly from the pre-computation trick.
2017-05-12 16:09:15 -05:00
Matthew Honnibal
b16ae75824 Remove serializer hacks from pipeline classes 2017-05-09 18:16:40 +02:00
Matthew Honnibal
bef89ef23d Mergery 2017-05-08 08:29:36 -05:00
Matthew Honnibal
94e86ae00a Predict tags with encoder 2017-05-08 07:53:45 -05:00
Matthew Honnibal
a66a4a4d0f Replace einsums 2017-05-08 14:46:50 +02:00
Matthew Honnibal
6782eedf9b Tmp GPU code 2017-05-07 11:04:24 -05:00
Matthew Honnibal
f99f5b75dc working residual net 2017-05-07 03:57:26 +02:00
Matthew Honnibal
7e04260d38 Data running through, likely errors in model 2017-05-06 14:22:20 +02:00
ines
d24589aa72 Clean up imports, unused code, whitespace, docstrings 2017-04-15 12:05:47 +02:00
ines
561f2a3eb4 Use consistent formatting for docstrings 2017-04-15 11:59:21 +02:00
Matthew Honnibal
354458484c WIP on add_label bug during NER training
Currently when a new label is introduced to NER during training,
it causes the labels to be read in in an unexpected order. This
invalidates the model.
2017-04-14 23:52:17 +02:00
Matthew Honnibal
2f63806ddb Update config when adding label. Re #910 2017-03-25 22:35:44 +01:00
Raphaël Bournhonesque
f332bf05be Remove unused import statements 2017-03-21 21:08:54 +01:00
Matthew Honnibal
7769bc31e3 Add beam-search classes 2017-03-15 09:27:41 -05:00
Matthew Honnibal
fa23278ee3 Add classes for beam parser and beam NER 2017-03-11 12:45:37 -06:00
Matthew Honnibal
f77a5bb60a Switch back to greedy parser 2017-03-11 11:11:30 -06:00
Matthew Honnibal
dcce9ca3f3 Use beam parser 2017-03-11 07:00:20 -06:00
Matthew Honnibal
b86f8af0c1 Fix doc strings 2016-11-01 12:25:36 +01:00
Matthew Honnibal
3e688e6d4b Fix issue #514 -- serializer fails when new entity type has been added. The fix here is quite ugly. It's best to add the entities ASAP after loading the NLP pipeline, to mitigate the brittleness. 2016-10-23 17:45:44 +02:00
Matthew Honnibal
f787cd29fe Refactor the pipeline classes to make them more consistent, and remove the redundant blank() constructor. 2016-10-16 21:34:57 +02:00