Description of the new feature / enhancement | 您所需的新功能或能力
Baselines to be upgraded from v0.5.8 to v1.0
Contribution Guidelines
Welcome to become a contributor and help make this project better! When upgrading the aforementioned models, please adhere to the following specifications:
-
Code Standards
- Variable Naming Conventions: If a variable's meaning is unclear, it should be renamed to be more descriptive.
- Fields and Local Variables: Use all lowercase letters, connected by
_.
- Constants and Global Variables: Use all uppercase letters, connected by
_.
- Class Names: Use CamelCase (capitalize the first letter of each word) and should not use
_.
- Function and Class Comments: Add comments for each class and major function to ensure readability. Refer to the implemented baselines for examples.
- Function Parameter Type Annotations: Annotate the parameter types and return type for each function. For example:
def func(a: int, b: float, c: torch.Tensor) -> torch.Tensor:
- Pass local checks using isort and pylint. Usage:
pip install pylint
pip install isort
isort src/basicts
pylint src/basicts
-
Code Style Standards
- Model configuration classes should be dataclasses inheriting from
BasicTSModelConfig. Try to keep field names consistent with other baselines (e.g., use hidden_size instead of d_model). Refer to the configurations of already implemented baselines.
- Strive to improve code efficiency by removing unnecessary modules and inefficient implementations.
- Prefer using components from the BasicTS module library to replace original components where possible (common examples include Transformer, RevIN, Embedding, etc.).
-
Please first open a pull request to beta branch to submit your updates.
贡献指南
欢迎成为贡献者,一起使这个项目变得更好!在升级上述模型时,请遵守下面的规范:
-
代码规范
- 变量命名规范:若变量语义不明确,应该重命名为语义明确的名字。
- 字段与临时变量:全部小写,用
_连接。
- 常量与全局变量:全部大写,用
_连接。
- 类名:首字母大写,不应该使用
_。
- 函数与类的注释:为每个类和主要的函数添加注释以保证可读性,可参考已实现的baseline。
- 函数参数类型注解:为每个函数的参数注释类型并标明返回值类型,例如:
def func(a: int, b: float, c: torch.Tensor) -> torch.Tensor:
- 在本地通过isor、pylint检查。使用方法:
pip install pylint
pip install isort
isort src/basicts
pylint src/basicts
-
代码风格规范
- 模型配置类应该为继承
BasicTSModelConfig的数据类,并尽量保持字段命名与其他baseline一致(例如:使用hidden_size 而不是 d_model)。可以参考已实现的baseline配置。
- 尽可能提高代码效率,移除不必要的模块和低效实现。
- 尽可能使用BasicTS module模块中的组件替换原有组件(例如常用的包括Transformer、RevIN、Embedding等)。
-
请向beta分支发起PR来提交您的更改。
Scenario when this would be used? | 使用场景
NA
Supporting information | 附加信息
No response
Description of the new feature / enhancement | 您所需的新功能或能力
Baselines to be upgraded from v0.5.8 to v1.0
Contribution Guidelines
Welcome to become a contributor and help make this project better! When upgrading the aforementioned models, please adhere to the following specifications:
Code Standards
_._._.Code Style Standards
BasicTSModelConfig. Try to keep field names consistent with other baselines (e.g., usehidden_sizeinstead ofd_model). Refer to the configurations of already implemented baselines.Please first open a pull request to beta branch to submit your updates.
贡献指南
欢迎成为贡献者,一起使这个项目变得更好!在升级上述模型时,请遵守下面的规范:
代码规范
_连接。_连接。_。代码风格规范
BasicTSModelConfig的数据类,并尽量保持字段命名与其他baseline一致(例如:使用hidden_size而不是d_model)。可以参考已实现的baseline配置。请向beta分支发起PR来提交您的更改。
Scenario when this would be used? | 使用场景
NA
Supporting information | 附加信息
No response