Named Entity Recognition predictions

Course Q&A Deep Learning Specialization Sequence Models
1

Hello! I am trying to use the pre-trained model to make some predictions. After some googling around I managed to ‘read’ the prediction outputs:

from transformers import TokenClassificationPipeline

pipe = TokenClassificationPipeline(model=model, tokenizer=tokenizer)
pipe("Michael Jordan basketball player - defense, offense")

The output though only uses LABEL_11 (Unknown) for all words. What is it missing?

[{'word': 'michael',
  'score': 0.2715179920196533,
  'entity': 'LABEL_11',
  'index': 1,
  'start': 0,
  'end': 7},
 {'word': 'jordan',
  'score': 0.3014214038848877,
  'entity': 'LABEL_11',
  'index': 2,
  'start': 8,
  'end': 14},
 {'word': 'basketball',
  'score': 0.31579136848449707,
  'entity': 'LABEL_11',
  'index': 3,
  'start': 15,
  'end': 25},
 {'word': 'player',
  'score': 0.34259045124053955,
  'entity': 'LABEL_11',
  'index': 4,
  'start': 26,
  'end': 32},
 {'word': '-',
  'score': 0.3144441246986389,
  'entity': 'LABEL_11',
  'index': 5,
  'start': 33,
  'end': 34},
 {'word': 'defense',
  'score': 0.3268392086029053,
  'entity': 'LABEL_11',
  'index': 6,
  'start': 35,
  'end': 42},
 {'word': ',',
  'score': 0.29030275344848633,
  'entity': 'LABEL_11',
  'index': 7,
  'start': 42,
  'end': 43},
 {'word': 'offense',
  'score': 0.2715778350830078,
  'entity': 'LABEL_11',
  'index': 8,
  'start': 44,
  'end': 51}]
2

Were you able to find an answer to your question?