generate_answers.py

from transformers import BertForQuestionAnswering, BertTokenizer
import torch

model_name = "bert-large-uncased-whole-word-masking-finetuned-squad"
tokenizer = BertTokenizer.from_pretrained(model_name)
model = BertForQuestionAnswering.from_pretrained(model_name)

def get_answer(question, context):
    inputs = tokenizer(question, context, return_tensors="pt", truncation=True, max_length=512)
    with torch.no_grad():
        outputs = model(**inputs)

    start_idx = torch.argmax(outputs.start_logits)
    end_idx = torch.argmax(outputs.end_logits) + 1

    answer_tokens = inputs["input_ids"][0][start_idx:end_idx]
    answer = tokenizer.decode(answer_tokens, skip_special_tokens=True)
    return answer