Transformers와 TensorFlow의 겉 핥아보기

아래의 것들을 설치하자.

pip3이 없다면 pip으로 하면 됩니다.

cGlwMyBpbnN0YWxsIC0tdXBncmFkZSBwaXANCnBpcDMgaW5zdGFsbCB0ZW5zb3JmbG93LWNwdQ0KcGlwMyBpbnN0YWxsIHRyYW5zZm9ybWVycw0KcGlwMyBpbnN0YWxsIHRlbnNvcmZsb3c9PTIuMy4wDQo=


     pip3 install --upgrade pip
     pip3 install tensorflow-cpu
     pip3 install transformers
     pip3 install tensorflow==2.3.0

다 설치했으면 아래 코드를 test.py 파일에 담아서 저장하고

python3 test.py 해서 실행해봅니다

코드상에 넣은 text 의 내용에 대해서 questions 배열에 질문을 담으면 애가 글을 읽고 맞는 답을 내어줍니다

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


     from transformers import AutoTokenizer, TFAutoModelForQuestionAnswering
     import tensorflow as tf
     tokenizer = AutoTokenizer.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
     model = TFAutoModelForQuestionAnswering.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
     text = r"""
     The Empire State Building is a 102-story Art Deco skyscraper in Midtown Manhattan in New York City.
     It was designed by Shreve, Lamb & Harmon and built from 1930 to 1931.
     Its name is derived from "Empire State", the nickname of the state of New York.
     The building has a roof height of 1,250 feet (380 m) and stands a total of 1,454 feet (443.2 m) tall, including its antenna.
     """
     questions = [
     "What's the location of the building?",
     "What's the name of the building?",
     "How tall is this?"
     ]
     for question in questions:
     inputs = tokenizer(question, text, add_special_tokens=True, return_tensors="tf")
     input_ids = inputs["input_ids"].numpy()[0]
     text_tokens = tokenizer.convert_ids_to_tokens(input_ids)
     answer_start_scores, answer_end_scores = model(inputs)
     answer_start = tf.argmax(
     answer_start_scores, axis=1
     ).numpy()[0] # Get the most likely beginning of answer with the argmax of the score
     answer_end = (
     tf.argmax(answer_end_scores, axis=1) + 1
     ).numpy()[0] # Get the most likely end of answer with the argmax of the score
     answer = tokenizer.convert_tokens_to_string(tokenizer.convert_ids_to_tokens(input_ids[answer_start:answer_end]))
     print(f"Question: {question}")
     print(f"Answer: {answer}")

애의 답은 다음과 같다

UXVlc3Rpb246IFdoYXQncyB0aGUgbG9jYXRpb24gb2YgdGhlIGJ1aWxkaW5nPw0KQW5zd2VyOiBtaWR0b3duIG1hbmhhdHRhbg0KUXVlc3Rpb246IFdoYXQncyB0aGUgbmFtZSBvZiB0aGUgYnVpbGRpbmc/DQpBbnN3ZXI6IHRoZSBlbXBpcmUgc3RhdGUgYnVpbGRpbmcNClF1ZXN0aW9uOiBIb3cgdGFsbCBpcyB0aGlzPw0KQW5zd2VyOiAxICwgNDU0IGZlZXQ=


     Question: What's the location of the building?
     Answer: midtown manhattan
     Question: What's the name of the building?
     Answer: the empire state building
     Question: How tall is this?
     Answer: 1 , 454 feet