在地平线horizon_xj3_open_explorer_v1.14.3_20220727 SDK包的文档中有如下一段原话：
浮点模型训练
数据集准备好之后，就可以开始训练浮点的检测网络了。在网络训练开始之前，你可以使用以下命令先测试一下网络的计算量和参数数量：
python3 tools/calops.py --config configs/detection/fcos/fcos_efficientnetb0_mscoco.py --input-shape “1,3,512,512”
执行上述代码后报告如下错误：
Traceback (most recent call last):
File “/root/projects/horizon_model_train_samples/tools/calops.py”, line 6, in
from hatbc.utils import deprecated_warning
ModuleNotFoundError: No module named ‘hatbc’
既然缺失了相关包，那么我应该安装哪个whl文件？
以下是从SDK中能找到的所有whl文件，全部安装了，无效。
horizon_xj3_open_explorer_v1.14.3_20220727> find .| grep .whl
./ddk/samples/ai_toolchain/horizon_model_train_samples/hemat-1.0.3-py3-none-any.whl
./ddk/samples/ai_toolchain/horizon_model_train_samples/horizon_hat-1.5.5-py3-none-any.whl
./ddk/package/host/ddk_vcs-0.4.1-py3-none-any.whl
./ddk/package/host/ai_toolchain/hbdk-3.37.2-cp36-cp36m-linux_x86_64.whl
./ddk/package/host/ai_toolchain/hbdk_model_verifier-3.37.2-py3-none-linux_x86_64.whl
./ddk/package/host/ai_toolchain/horizon_nn-0.14.0-py3-none-any.whl
./ddk/package/host/ai_toolchain/horizon_nn_gpu-0.14.0-py3-none-any.whl
./ddk/package/host/ai_toolchain/horizon_tc_ui-1.9.9-py3-none-any.whl
torch版本为1.9.1
python版本为：3.9)