feat(training): 实现 Trainer 模块化重构 (Trainer V2)

- 新增 FactorManager 组件:统一管理多种来源因子
- 新增 DataPipeline 组件:完整数据处理流程(注册、过滤、划分、预处理)
- 新增 Task 策略组件:BaseTask 抽象基类、RegressionTask、RankTask
- 新增 ResultAnalyzer 组件:特征重要性分析和结果组装
- 新增 TrainerV2:作为纯调度引擎协调各组件
- 支持回归和排序学习两种训练模式
- 采用组合模式解耦训练流程,消除代码重复
This commit is contained in:
2026-03-24 23:35:31 +08:00
parent bace4cc5f4
commit e41a128ca3
13 changed files with 4045 additions and 1509 deletions

View File

@@ -43,6 +43,12 @@ from src.training.utils import check_data_quality
# 配置
from src.training.config import TrainingConfig
# 新增:模块化 Trainer 组件
from src.training.factor_manager import FactorManager
from src.training.pipeline import DataPipeline
from src.training.result_analyzer import ResultAnalyzer
from src.training.tasks import BaseTask, RegressionTask, RankTask
__all__ = [
# 基础抽象类
"BaseModel",
@@ -74,4 +80,11 @@ __all__ = [
"check_data_quality",
# 配置
"TrainingConfig",
# 新增:模块化 Trainer 组件
"FactorManager",
"DataPipeline",
"ResultAnalyzer",
"BaseTask",
"RegressionTask",
"RankTask",
]