What is the spaCy 3 equivalent of the following code?

[romanRegmi]

I am trying to follow a tutorial that was made using spaCy v2. It turns out that with the update the whole training process changed. Now I cannot figure out the changes I have to make in the code.

import spacy

nlp = spacy.blank("en")
ner = nlp.create_pipe('ner')
ner.add_label('CONC_CAMP')

nlp.add_pipe(ner, name='conc_camp_ner')
nlp.to_disk('holocaust_ner')

Here is what I have tried.

import spacy

nlp = spacy.blank("en")


nlp.add_pipe("ner", name='conc_camp_ner')

nlp.to_disk("holocaust_ner")

ner = nlp.get_pipe('conc_camp_ner')

ner.add_label('CONC_CAMP')

The code is supposed to create a label in the meta.json file. But it fails to do so.

"labels":{
    "conc_camp_ner":[
      "CONC_CAMP"
    ]
  }

But the label is not created in my case.

"labels":{
    "conc_camp_ner":[
    ]
  }

[ljvmiranda921]

It turns out that with the update the whole training process changed

Hi @romanRegmi, if you're training an NER model in v3, then you should do it through the spaCy config. That means you don't have to write any code for training, although you have to convert your text into a serialized format. You can check a few examples in the spaCy projects repo.