import torch
from transformers import AutoTokenizer, AutoModelForQuestionAnswering

tokenizer = AutoTokenizer.from_pretrained("Intel/dynamic_tinybert")
model = AutoModelForQuestionAnswering.from_pretrained("Intel/dynamic_tinybert")

context = "记住数字123456，我稍后会问你。"
question = "我告诉你的数字是多少？"

# 对上下文和问题进行编码
tokens = tokenizer.encode_plus(question, context, return_tensors="pt", truncation=True)

# 获取输入ID和注意力掩码
input_ids = tokens["input_ids"]
attention_mask = tokens["attention_mask"]

# 执行问答任务
outputs = model(input_ids, attention_mask=attention_mask)
start_scores = outputs.start_logits
end_scores = outputs.end_logits

# 找到答案的起始和结束位置
answer_start = torch.argmax(start_scores)
answer_end = torch.argmax(end_scores) + 1
answer = tokenizer.convert_tokens_to_string(tokenizer.convert_ids_to_tokens(input_ids[0][answer_start:answer_end]))

# 打印答案
print("答案:", answer)