ncbi_disease_ner

This model is a fine-tuned version of bert-base-cased on an ncbi_disease dataset. It achieves the following results on the evaluation set:

• Loss: 0.0921
• Precision: 0.8082
• Recall: 0.8564
• F1: 0.8316
• Accuracy: 0.9835

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.1149	1.0	680	0.0580	0.7475	0.8501	0.7955	0.9810
0.0436	2.0	1360	0.0687	0.7430	0.8412	0.7890	0.9825
0.0164	3.0	2040	0.0702	0.7901	0.8513	0.8196	0.9830
0.0076	4.0	2720	0.0829	0.7884	0.8666	0.8257	0.9826
0.0047	5.0	3400	0.0921	0.8082	0.8564	0.8316	0.9835

Run the model

from transformers import pipeline

model_checkpoint = "manibt1993/ncbi_disease_ner"
token_classifier = pipeline(
    "token-classification", model=model_checkpoint, aggregation_strategy="simple"
)
token_classifier("patient has diabtes, anemia, hypertension with ckd which hurts the patient since 6 years. Patient today experience with right leg pain, fever and cough.")

Model output

[{'entity_group': 'Disease',
  'score': 0.69145554,
  'word': 'diabtes',
  'start': 12,
  'end': 19},
 {'entity_group': 'Disease',
  'score': 0.9955915,
  'word': 'anemia',
  'start': 21,
  'end': 27},
 {'entity_group': 'Disease',
  'score': 0.99971104,
  'word': 'hypertension',
  'start': 29,
  'end': 41},
 {'entity_group': 'Disease',
  'score': 0.9249976,
  'word': 'right leg pain',
  'start': 120,
  'end': 134},
 {'entity_group': 'Disease',
  'score': 0.9983512,
  'word': 'fever',
  'start': 136,
  'end': 141},
 {'entity_group': 'Disease',
  'score': 0.99849665,
  'word': 'cough',
  'start': 146,
  'end': 151}]

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1