Getting KeyError [E022]

[Mayurdeo]

I'm trying to use perform custom named entity recognition, but I'm facing an error (shown above). This has never happened previously. Any help would be appreciated.

Error: "[E022] Could not find a transition with the name 'B-Overcast Sky Cover' in the NER model."

How to reproduce the behaviour

labels=['3 letter identifier', 'UA/UUA', 'Location','Time','Flight level', 'Aircraft type', 'Temperature','Wind',
        'Sky Cover', 'Broken Sky Cover', 'Scattered Sky Cover', 'Overcast Sky Cover'
        'Rain Flight visibility', 'Fog Flight visibility', 'Haze Flight visibility',
        'No Turbulence', 'Light Chop Turbulence','Moderate Turbulence','Severe Turbulence',
        'No icing', 'Light rime icing', 'Moderate clear icing',
        'Remarks', 'LLWS Gain', 'LLWS Loss']

@plac.annotations(
    model=("Model name. Defaults to blank 'en' model.", "option", "m", str),
    new_model_name=("New model name for model meta.", "option", "nm", str),
    output_dir=("Optional output directory", "option", "o", Path),
    n_iter=("Number of training iterations", "option", "n", int),
)


def main(model=None, new_model_name="animal", output_dir=None, n_iter=30):
    """Set up the pipeline and entity recognizer, and train the new entity."""
    random.seed(0)
    if model is not None:
        nlp = spacy.load(model)  # load existing spaCy model
        print("Loaded model '%s'" % model)
    else:
        nlp = spacy.blank("en")  # create blank Language class
        print("Created blank 'en' model")
    # Add entity recognizer to model if it's not in the pipeline
    # nlp.create_pipe works for built-ins that are registered with spaCy
    if "ner" not in nlp.pipe_names:
        ner = nlp.create_pipe("ner")
        nlp.add_pipe(ner)
    # otherwise, get it, so we can add labels to it
    else:
        ner = nlp.get_pipe("ner")

    for i in labels:  
      ner.add_label(i)  # add new entity label to entity recognizer

    if model is None:
        optimizer = nlp.begin_training()
    else:
        optimizer = nlp.resume_training()
    move_names = list(ner.move_names)
    # get names of other pipes to disable them during training
    pipe_exceptions = ["ner", "trf_wordpiecer", "trf_tok2vec"]
    other_pipes = [pipe for pipe in nlp.pipe_names if pipe not in pipe_exceptions]
    # only train NER
    with nlp.disable_pipes(*other_pipes), warnings.catch_warnings():
        # show warnings for misaligned entity spans once
        warnings.filterwarnings("once", category=UserWarning, module='spacy')

        sizes = compounding(1.0, 4.0, 1.001)
        # batch up the examples using spaCy's minibatch
        for itn in range(n_iter):
            random.shuffle(TRAIN_DATA)
            batches = minibatch(TRAIN_DATA, size=sizes)
            losses = {}
            for batch in batches:
                texts, annotations = zip(*batch)
                # try:
                #   nlp.update(text, annotations, sgd=optimizer, drop=0.35, losses=losses) 
                # except:
                #   import pdb;pdb.set_trace() 
                nlp.update(texts, annotations, sgd=optimizer, drop=0.35, losses=losses)
            print("Losses", losses)

    # save model to output directory
    if output_dir is not None:
        output_dir = Path(output_dir)
        if not output_dir.exists():
            output_dir.mkdir()
        nlp.meta["name"] = new_model_name  # rename model
        nlp.to_disk(output_dir)
        print("Saved model to", output_dir)

main(model='en', new_model_name="airplane_final_1", output_dir='/content/gdrive/My Drive/Airplane/Final_1', n_iter=2000)        

Your Environment

• Operating System: Mac OS
• Python Version Used: 3.6.9
• spaCy Version Used: 2.2.4
• Environment Information: Google Colab

[Mayurdeo]

Hello,
Please do help to solve this issue. I'm stuck.

[svlandeg]

labels=['3 letter identifier', 'UA/UUA', 'Location','Time','Flight level', 'Aircraft type', 'Temperature','Wind',
        'Sky Cover', 'Broken Sky Cover', 'Scattered Sky Cover', 'Overcast Sky Cover'
        'Rain Flight visibility', 'Fog Flight visibility', 'Haze Flight visibility',
        'No Turbulence', 'Light Chop Turbulence','Moderate Turbulence','Severe Turbulence',
        'No icing', 'Light rime icing', 'Moderate clear icing',
        'Remarks', 'LLWS Gain', 'LLWS Loss']

It looks like there's a comma missing at the end of your second line ;-)

[Naveen9182]

KeyError: "[E022] Could not find a transition with the name 'B-Deed_type' in the NER model."

This is my script:
#Function to train NER blank model
def train_spacy(TRAIN_DATA,iterations):
import en_core_web_sm

nlp = en_core_web_sm.load()
#nlp=spacy.load('en_core_web_sm')

ner=nlp.create_pipe("ner")
LABEL=['Deed type','Legal Description','PA Number','Instrument/Book&Page Number']
for item in LABEL:
    print(item)
    ner.add_label(item)
nlp.add_pipe(ner,name='cpi_ner')
other_pipes = [pipe for pipe in nlp.pipe_names if pipe!='cpi_ner']
with nlp.disable_pipes(*other_pipes):
    optimizer = nlp.begin_training()
    for itn in range(iterations):
        print(f"Starting iteration{str(itn)}")
        random.shuffle(TRAIN_DATA)
        losses={}
        for text,annotations in TRAIN_DATA:
            nlp.update([text],[annotations],drop=0.2,sgd=optimizer,losses=losses)
        print(losses)
return(nlp)

random.shuffle(data)
trained=train_spacy(data,100)

Your Environment
Operating System: Windows OS
Python Version Used: 3.7
spaCy Version Used: 2.3.7
Environment Information: Jupyter Notebook

Please look into this issue and help me in this case, I'm eagerly waiting for your reply.

[svlandeg]

Hi @Naveen9182 : You'll need to double check your data & NER annotations. You're adding the label "Deed type" but it looks like your data contains a label "Deed_type" - mind the underscore.

[Naveen9182]

Yes, I am able to resolve my issue.
Thank you