explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2025-10-31 16:07:41 +03:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #1415 from IamJeffG/fix-alpha-example-train-ner-standalone

Browse Source 
Bugfix example script train_ner_standalone.py, fails after training
This commit is contained in:
Matthew Honnibal 2017-10-12 21:44:28 +02:00 committed by GitHub
commit cb0e727c54
1 changed files with 4 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
examples/training/train_ner_standalone.py
View File

@ -6,7 +6,7 @@ To achieve that, it duplicates some of spaCy's internal functionality.
Specifically, in this example, we don't use spaCy's built-in Language class to
wire together the Vocab, Tokenizer and EntityRecognizer. Instead, we write
our own simle Pipeline class, so that it's easier to see how the pieces
our own simple Pipeline class, so that it's easier to see how the pieces
interact.
Input data:
@ -142,16 +142,15 @@ def train(nlp, train_examples, dev_examples, nr_epoch=5):
inputs, annots = zip(*batch)
nlp.update(list(inputs), list(annots), sgd, losses=losses)
scores = nlp.evaluate(dev_examples)
report_scores(i, losses['ner'], scores)
scores = nlp.evaluate(dev_examples)
report_scores(channels, i+1, loss, scores)
report_scores(i+1, losses['ner'], scores)
def report_scores(i, loss, scores):
precision = '%.2f' % scores['ents_p']
recall = '%.2f' % scores['ents_r']
f_measure = '%.2f' % scores['ents_f']
print('%d %s %s %s' % (int(loss), precision, recall, f_measure))
print('Epoch %d: %d %s %s %s' % (
i, int(loss), precision, recall, f_measure))
def read_examples(path):