Rooba
/
spaCy
mirror of https://github.com/explosion/spaCy.git
1
1
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Let beam forward use minibatches

This commit is contained in:
Matthew Honnibal 2017-11-15 00:51:42 +01:00
parent 855872f872
commit d274d3a3b9
1 changed files with 32 additions and 23 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

55
spacy/syntax/nn_parser.pyx
View File

@ -17,7 +17,7 @@ from cpython.ref cimport PyObject, Py_XDECREF
from cpython.exc cimport PyErr_CheckSignals, PyErr_SetFromErrno from cpython.exc cimport PyErr_CheckSignals, PyErr_SetFromErrno
from libc.math cimport exp from libc.math cimport exp
from libcpp.vector cimport vector from libcpp.vector cimport vector
from libc.string cimport memset from libc.string cimport memset, memcpy
from libc.stdlib cimport calloc, free from libc.stdlib cimport calloc, free
from cymem.cymem cimport Pool from cymem.cymem cimport Pool
from thinc.typedefs cimport weight_t, class_t, hash_t from thinc.typedefs cimport weight_t, class_t, hash_t
@ -485,14 +485,14 @@ cdef class Parser:
cdef np.ndarray scores cdef np.ndarray scores
cdef Doc doc cdef Doc doc
cdef int nr_class = self.moves.n_moves cdef int nr_class = self.moves.n_moves
cdef StateClass stcls, output
cuda_stream = util.get_cuda_stream() cuda_stream = util.get_cuda_stream()
(tokvecs, bp_tokvecs), state2vec, vec2scores = self.get_batch_model( (tokvecs, bp_tokvecs), state2vec, vec2scores = self.get_batch_model(
docs, cuda_stream, 0.0) docs, cuda_stream, 0.0)
beams = []
cdef int offset = 0 cdef int offset = 0
cdef int j = 0 cdef int j = 0
cdef int k cdef int k
beams = []
for doc in docs: for doc in docs:
beam = Beam(nr_class, beam_width, min_density=beam_density) beam = Beam(nr_class, beam_width, min_density=beam_density)
beam.initialize(self.moves.init_beam_state, doc.length, doc.c) beam.initialize(self.moves.init_beam_state, doc.length, doc.c)
@ -501,34 +501,43 @@ cdef class Parser:
state.offset = offset state.offset = offset
offset += len(doc) offset += len(doc)
beam.check_done(_check_final_state, NULL) beam.check_done(_check_final_state, NULL)
while not beam.is_done: beams.append(beam)
states = [] cdef np.ndarray token_ids
token_ids = numpy.zeros((len(docs) * beam_width, self.nr_feature),
dtype='i', order='C')
todo = [beam for beam in beams if not beam.is_done]
cdef int* c_ids
cdef int nr_feature = self.nr_feature
cdef int n_states
while todo:
todo = [beam for beam in beams if not beam.is_done]
token_ids.fill(-1)
c_ids = <int*>token_ids.data
n_states = 0
for beam in todo:
for i in range(beam.size): for i in range(beam.size):
stcls = StateClass.borrow(<StateC*>beam.at(i)) state = <StateC*>beam.at(i)
# This way we avoid having to score finalized states # This way we avoid having to score finalized states
# We do have to take care to keep indexes aligned, though # We do have to take care to keep indexes aligned, though
if not stcls.is_final(): if not state.is_final():
states.append(stcls) state.set_context_tokens(c_ids, nr_feature)
token_ids = self.get_token_ids(states) c_ids += nr_feature
vectors = state2vec(token_ids) n_states += 1
if self.cfg.get('hist_size', 0): if n_states == 0:
hists = numpy.asarray([st.history[:self.cfg['hist_size']] break
for st in states], dtype='i') vectors = state2vec(token_ids[:n_states])
scores = vec2scores((vectors, hists)) scores = vec2scores(vectors)
else: c_scores = <float*>scores.data
scores = vec2scores(vectors) for beam in todo:
j = 0
c_scores = <float*>scores.data
for i in range(beam.size): for i in range(beam.size):
state = <StateC*>beam.at(i) state = <StateC*>beam.at(i)
if not state.is_final(): if not state.is_final():
self.moves.set_valid(beam.is_valid[i], state) self.moves.set_valid(beam.is_valid[i], state)
for k in range(nr_class): memcpy(beam.scores[i], c_scores, nr_class * sizeof(float))
beam.scores[i][k] = c_scores[j * scores.shape[1] + k] c_scores += nr_class
j += 1
beam.advance(_transition_state, NULL, <void*>self.moves.c) beam.advance(_transition_state, NULL, <void*>self.moves.c)
beam.check_done(_check_final_state, NULL) beam.check_done(_check_final_state, NULL)
beams.append(beam)
tokvecs = self.model[0].ops.unflatten(tokvecs, tokvecs = self.model[0].ops.unflatten(tokvecs,
[len(doc) for doc in docs]) [len(doc) for doc in docs])
return beams, tokvecs return beams, tokvecs
@ -536,7 +545,7 @@ cdef class Parser:
def update(self, docs, golds, drop=0., sgd=None, losses=None): def update(self, docs, golds, drop=0., sgd=None, losses=None):
if not any(self.moves.has_gold(gold) for gold in golds): if not any(self.moves.has_gold(gold) for gold in golds):
return None return None
if self.cfg.get('beam_width', 1) >= 2 and numpy.random.random() >= 0.5: if self.cfg.get('beam_width', 1) >= 2 and numpy.random.random() >= 0.0:
return self.update_beam(docs, golds, return self.update_beam(docs, golds,
self.cfg['beam_width'], self.cfg['beam_density'], self.cfg['beam_width'], self.cfg['beam_density'],
drop=drop, sgd=sgd, losses=losses) drop=drop, sgd=sgd, losses=losses)