train模式下指定的 --id_file不生效

你好，我想利用部分数据进行模型的训练，使用如下命令后
``` sh
python -m tf_euler --data_dir test_data --id_file test_data/train_id.csv \
--max_id 999 \
--feature_idx 0 --feature_dim 3 -—label_idx 1 --label_dim 1 \
--num_classes 2 \
--model graphsage_supervised --mode save_embedding \
--batch_size 10 --num_epochs 100
```
发现，embdeding中的id并没有只使用指定的部分id。当然，查看文档中的描述是说 `--id_file是用于测试集id文件，evaluate必须；`，请问在训练模式下，有什么办法指定训练、校验、和测试集的参数吗？

Start at the `python -m tf_euler` entry point and trace how `--id_file` is consumed in `train`, `save_embedding`, and `evaluate` modes. Compare the documented test-set behavior with the reported command, then determine how training, validation, and test IDs are currently selected. Done means the supported behavior is explicit and the requested ID selection is applied consistently.

Written by the indexing model from the issue text.