- 数据集:https://www.kaggle.com/datasets/kuanghueilee/scan-features
- bert模型:https://huggingface.co/google-bert/bert-base-uncased/resolve/main/pytorch_model.bin?download=true
- 参数配置文件:config.py
- 安装环境
conda env create -n crossmodal -f environment.yml
- Punkt Sentence Tokenizer:
import nltk
nltk.download()
> d punkt
- 开始训练
python train.py