pytorch question-answering

bert-base-finetuned-squad2

Model description

This model is based on bert-base-uncased and was finetuned on SQuAD2.0. The corresponding papers you can found here (model) and here (data).

How to use

from transformers.pipelines import pipeline

model_name = "phiyodr/bert-base-finetuned-squad2"
nlp = pipeline('question-answering', model=model_name, tokenizer=model_name)
inputs = {
    'question': 'What discipline did Winkelmann create?',
    'context': 'Johann Joachim Winckelmann was a German art historian and archaeologist. He was a pioneering Hellenist who first articulated the difference between Greek, Greco-Roman and Roman art. "The prophet and founding hero of modern archaeology", Winckelmann was one of the founders of scientific archaeology and first applied the categories of style on a large, systematic basis to the history of art. '
}
nlp(inputs)

Training procedure

{
	"base_model": "bert-base-uncased",
	"do_lower_case": True,
	"learning_rate": 3e-5,
	"num_train_epochs": 4,
	"max_seq_length": 384,
	"doc_stride": 128,
	"max_query_length": 64,
	"batch_size": 96 
}

Eval results

{
  "exact": 70.3950138970774,
  "f1": 73.90527661873521,
  "total": 11873,
  "HasAns_exact": 71.4574898785425,
  "HasAns_f1": 78.48808186475087,
  "HasAns_total": 5928,
  "NoAns_exact": 69.33557611438184,
  "NoAns_f1": 69.33557611438184,
  "NoAns_total": 5945
}