wulikai1993

extract_xgbooost_cluster_env() and xgb.rabit.get_rank() get different rank number

2

I ran distributed training on k8s. The `rank` number was got by `extract_xgbooost_cluster_env()` as in https://github.com/kubeflow/xgboost-operator/blob/master/config/samples/xgboost-dist/train.py#L29 However, `xgb.rabit.get_rank()` got another `rank` number as in https://github.com/kubeflow/xgboost-operator/blob/master/config/samples/xgboost-dist/train.py#L57. There are two things confusing...

本TF版本的输入 `token_ids, segment_ids` 和原版的模型输入 `input_ids, token_type_ids` 是否一样？

9

本TF版predict 输入为 https://github.com/bojone/CDial-GPT-tf/blob/master/example.py#L37 ```bash return model.predict([token_ids, segment_ids])[:, -1] ``` 原版模型输入为 https://github.com/thu-coai/CDial-GPT/blob/master/interact.py#L81 ```bash logits, *_ = model(input_ids, token_type_ids=token_type_ids) ``` 我使用 pytorch 版的输入 `input_ids, token_type_ids` 作为 TF版的输入 `token_ids, segment_ids`，但两个模型的输出结果虽然shape相同，但数值差别很大。 请问什么原因呢？感谢！