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

* Upd train script

This commit is contained in:
Matthew Honnibal 2015-01-31 13:44:37 +11:00
parent 08ca5c8970
commit df749f257d
1 changed files with 2 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
bin/parser/train.py
View File

@ -9,7 +9,6 @@ import codecs
import random import random
import time import time
import gzip import gzip
import nltk
import plac import plac
import cProfile import cProfile
@ -219,10 +218,12 @@ def train(Language, paragraphs, model_dir, n_iter=15, feat_set=u'basic', seed=0,
def evaluate(Language, dev_loc, model_dir, gold_preproc=False): def evaluate(Language, dev_loc, model_dir, gold_preproc=False):
global loss
nlp = Language() nlp = Language()
n_corr = 0 n_corr = 0
total = 0 total = 0
skipped = 0 skipped = 0
loss = 0
with codecs.open(dev_loc, 'r', 'utf8') as file_: with codecs.open(dev_loc, 'r', 'utf8') as file_:
paragraphs = read_docparse_gold(file_) paragraphs = read_docparse_gold(file_)
for tokens, tag_strs, heads, labels in iter_data(paragraphs, nlp.tokenizer, for tokens, tag_strs, heads, labels in iter_data(paragraphs, nlp.tokenizer,