Weibo Datase處理

[elaine35002]

作者您好，我想使用ZEN做embedding並提取相應的詞向量，因此執行了以下程式碼:

 python3 run_token_level_classification.py --data_dir data/weibo --bert_model models --task_name msra --do_train --do_ev
al --train_batch_size 32 --eval_batch_size 32 --learning_rate 2e-5 --num_train_epochs 3

但出現如下報錯:

raceback (most recent call last):
  File "run_token_level_classification.py", line 438, in <module>
    main()
  File "run_token_level_classification.py", line 425, in main
    model = ZenForTokenClassification.from_pretrained(args.bert_model,
  File "/home/code-server/SANER/ZEN/modeling.py", line 838, in from_pretrained
    raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for ZenForTokenClassification:
        size mismatch for classifier.weight: copying a param with shape torch.Size([10, 768]) from checkpoint, the shape in current model is torch.Size([20, 768]).
        size mismatch for classifier.bias: copying a param with shape torch.Size([10]) from checkpoint, the shape in current model is torch.Size([20]).

請問可以怎麼解決呢?謝謝 我用的模型是下載的預訓練NER模型

[GuiminChen]

您好，您的来信已收到，我会尽快回复您的邮件。 ============ 祝您生活愉快！