Question

在加载修改后的模型时，使用EnityRuler标志向管道中添加before='ner'时遇到了一个奇怪的问题。

所以我创建了模式，将它们添加到我的entity_ruler，并将标尺添加到我的管道：

nlp = spacy.load('de_core_news_sm')
ruler = EntityRuler(nlp)
ruler.add_patterns(patterns)
nlp.add_pipe(ruler, before='ner')

然后我检查目录，如有必要，将其创建并将模型保存到磁盘：

output_dir = Path('custom_model')
if not output_dir.exists():
  output_dir.mkdir()
nlp.to_disk(output_dir)

现在我用...加载模型

spacy.load(Path('custom_model'))

它引发以下错误：

ValueError: [E109] Model for component 'ner' not initialized.
Did you forget to load a model, or forget to call begin_training()?

没有before='ner' ...

我试图整天弄清楚这个问题，但是失败了：/
我正在使用Spacy v2.1.4

扩展错误：

Traceback (most recent call last):
  File "/spacy/__init__.py", line 27, in load
    return util.load_model(name, **overrides)
  File "/spacy/util.py", line 135, in load_model
    return load_model_from_path(name, **overrides)
  File "/spacy/util.py", line 173, in load_model_from_path
    return nlp.from_disk(model_path)
  File "/spacy/language.py", line 791, in from_disk
    util.from_disk(path, deserializers, exclude)
  File "/spacy/util.py", line 630, in from_disk
    reader(path / key)
  File "/spacy/language.py", line 787, in <lambda>
    deserializers[name] = lambda p, proc=proc: proc.from_disk(p, exclude=["vocab"])
  File "/spacy/pipeline/entityruler.py", line 183, in from_disk
    self.add_patterns(patterns)
  File "/spacy/pipeline/entityruler.py", line 138, in add_patterns
    self.phrase_patterns[label].append(self.nlp(pattern))
  File "/spacy/language.py", line 390, in __call__
    doc = proc(doc, **component_cfg.get(name, {}))
  File "nn_parser.pyx", line 205, in spacy.syntax.nn_parser.Parser.__call__
  File "nn_parser.pyx", line 238, in spacy.syntax.nn_parser.Parser.predict
  File "nn_parser.pyx", line 235, in spacy.syntax.nn_parser.Parser.require_model
ValueError: [E109] Model for component 'ner' not initialized.
Did you forget to load a model, or forget to call begin_training()?

在ner之前添加EntityRuler并将模型保存到磁盘崩溃导致加载模型

0 个答案: