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

Use FTRL training in parser

This commit is contained in:
Matthew Honnibal 2017-03-08 01:38:51 +01:00
parent d108534dc2
commit 40703988bc
1 changed files with 15 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
spacy/syntax/parser.pyx
View File

@ -124,6 +124,8 @@ cdef class Parser:
elif 'features' not in cfg: elif 'features' not in cfg:
cfg['features'] = self.feature_templates cfg['features'] = self.feature_templates
self.model = ParserModel(cfg['features']) self.model = ParserModel(cfg['features'])
self.model.l1_penalty = 1e-7
self.cfg = cfg self.cfg = cfg
def __reduce__(self): def __reduce__(self):
@ -258,15 +260,20 @@ cdef class Parser:
self.model.set_featuresC(&eg.c, stcls.c) self.model.set_featuresC(&eg.c, stcls.c)
self.moves.set_costs(eg.c.is_valid, eg.c.costs, stcls, gold) self.moves.set_costs(eg.c.is_valid, eg.c.costs, stcls, gold)
self.model.set_scoresC(eg.c.scores, eg.c.features, eg.c.nr_feat) self.model.set_scoresC(eg.c.scores, eg.c.features, eg.c.nr_feat)
self.model.updateC(&eg.c) self.model.time += 1
guess = VecVec.arg_max_if_true(eg.c.scores, eg.c.is_valid, eg.c.nr_class) guess = VecVec.arg_max_if_true(eg.c.scores, eg.c.is_valid, eg.c.nr_class)
if eg.c.costs[guess] > 0:
best = VecVec.arg_max_if_zero(eg.c.scores, eg.c.costs, eg.c.nr_class)
for feat in eg.c.features[:eg.c.nr_feat]:
self.model.update_weight_ftrl(feat.key, best, -feat.value * eg.costs[guess])
self.model.update_weight_ftrl(feat.key, guess, feat.value * eg.costs[guess])
action = self.moves.c[eg.guess] action = self.moves.c[guess]
action.do(stcls.c, action.label) action.do(stcls.c, action.label)
loss += eg.costs[eg.guess] loss += eg.costs[guess]
eg.fill_scores(0, eg.nr_class) eg.fill_scores(0, eg.c.nr_class)
eg.fill_costs(0, eg.nr_class) eg.fill_costs(0, eg.c.nr_class)
eg.fill_is_valid(1, eg.nr_class) eg.fill_is_valid(1, eg.c.nr_class)
return loss return loss
def step_through(self, Doc doc): def step_through(self, Doc doc):