explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2024-12-26 01:46:28 +03:00
Code Issues Packages Projects Releases Wiki Activity

Divide d_loss by batch size

Browse Source
This commit is contained in:
Matthew Honnibal 2017-05-27 18:32:46 -05:00
parent b082f76494
commit 9e711c3476
2 changed files with 3 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
spacy/pipeline.pyx
View File

@ -228,6 +228,7 @@ class NeuralTagger(object):
idx += 1 idx += 1
correct = self.model.ops.xp.array(correct, dtype='i') correct = self.model.ops.xp.array(correct, dtype='i')
d_scores = scores - to_categorical(correct, nb_classes=scores.shape[1]) d_scores = scores - to_categorical(correct, nb_classes=scores.shape[1])
d_scores /= d_scores.shape[0]
loss = (d_scores**2).sum() loss = (d_scores**2).sum()
d_scores = self.model.ops.unflatten(d_scores, [len(d) for d in docs]) d_scores = self.model.ops.unflatten(d_scores, [len(d) for d in docs])
return float(loss), d_scores return float(loss), d_scores
@ -292,6 +293,7 @@ class NeuralLabeller(NeuralTagger):
idx += 1 idx += 1
correct = self.model.ops.xp.array(correct, dtype='i') correct = self.model.ops.xp.array(correct, dtype='i')
d_scores = scores - to_categorical(correct, nb_classes=scores.shape[1]) d_scores = scores - to_categorical(correct, nb_classes=scores.shape[1])
d_scores /= d_scores.shape[0]
loss = (d_scores**2).sum() loss = (d_scores**2).sum()
d_scores = self.model.ops.unflatten(d_scores, [len(d) for d in docs]) d_scores = self.model.ops.unflatten(d_scores, [len(d) for d in docs])
return float(loss), d_scores return float(loss), d_scores

2
spacy/syntax/nn_parser.pyx
View File

@ -450,7 +450,7 @@ cdef class Parser:
scores, bp_scores = vec2scores.begin_update(vector, drop=drop) scores, bp_scores = vec2scores.begin_update(vector, drop=drop)
d_scores = self.get_batch_loss(states, golds, scores) d_scores = self.get_batch_loss(states, golds, scores)
d_vector = bp_scores(d_scores, sgd=sgd) d_vector = bp_scores(d_scores / d_scores.shape[0], sgd=sgd)
if drop != 0: if drop != 0:
d_vector *= mask d_vector *= mask