fix : 修复策略相关文件的类型注解与抽象方法体

- 修复 strategy.py 单元素 Union 与缺失的参数/返回类型注解
- 修复 train_context.py 8 个 default=None 字段缺 Optional 标记
- 修复 sample.py/packing.py/position_id.py 方法缺参数及返回类型注解
- 修复 factory.py _resolve_type/list_registered 缺类型注解
- 修复 train_config.py 裸 dict/list 缺泛型参数
- abstractmethod body 从 ... 改为 raise NotImplementedError
- feat : checkpoint meta.json 保存 TrainConfig 超参供人工查阅

astrai/preprocessing/packing.py

@@ -12,7 +12,7 @@ from typing import Dict, List, Tuple
 from astrai.factory import BaseFactory
 
 
-def _truncate(seq: list, max_len: int, mode: str) -> list:
+def _truncate(seq: List[int], max_len: int, mode: str) -> List[int]:
     if len(seq) <= max_len:
         return seq
     if mode == "keep_end":
@@ -26,10 +26,11 @@ class PackingStrategy(ABC):
     @abstractmethod
     def apply(
         self,
-        keys: Dict[str, List[list]],
+        keys: Dict[str, List[List[int]]],
         max_packed_len: int,
         truncation_mode: str,
-    ) -> Dict[str, List[list]]: ...
+    ) -> Dict[str, List[List[int]]]:
+        raise NotImplementedError
 
 
 class PackingStrategyFactory(BaseFactory["PackingStrategy"]):
@@ -38,7 +39,12 @@ class PackingStrategyFactory(BaseFactory["PackingStrategy"]):
 
 @PackingStrategyFactory.register("simple")
 class SimplePacking(PackingStrategy):
-    def apply(self, keys, max_packed_len, truncation_mode):
+    def apply(
+        self,
+        keys: Dict[str, List[List[int]]],
+        max_packed_len: int,
+        truncation_mode: str,
+    ) -> Dict[str, List[List[int]]]:
         return {
             k: [_truncate(v, max_packed_len, truncation_mode) for v in vals]
             for k, vals in keys.items()
@@ -47,7 +53,12 @@ class SimplePacking(PackingStrategy):
 
 @PackingStrategyFactory.register("bfd")
 class BFDPacking(PackingStrategy):
-    def apply(self, keys, max_packed_len, truncation_mode):
+    def apply(
+        self,
+        keys: Dict[str, List[List[int]]],
+        max_packed_len: int,
+        truncation_mode: str,
+    ) -> Dict[str, List[List[int]]]:
         sequences = keys.get("sequence", [])
         if not sequences:
             return keys
@@ -61,7 +72,7 @@ class BFDPacking(PackingStrategy):
         return dict(reordered)
 
     @staticmethod
-    def _plan(sequences: List[list], max_packed_len: int) -> List[Tuple[int, int]]:
+    def _plan(sequences: List[List[int]], max_packed_len: int) -> List[Tuple[int, int]]:
         n = len(sequences)
         order = sorted(range(n), key=lambda i: len(sequences[i]), reverse=True)
         bins: List[List[int]] = []

astrai/preprocessing/position_id.py

@@ -16,7 +16,8 @@ class PositionIdStrategy(ABC):
     """Generate ``position_ids`` for packed sequences."""
 
     @abstractmethod
-    def generate(self, sequences: List[list]) -> List[int]: ...
+    def generate(self, sequences: List[List[int]]) -> List[int]:
+        raise NotImplementedError
 
 
 class PositionIdStrategyFactory(BaseFactory["PositionIdStrategy"]):
@@ -25,13 +26,13 @@ class PositionIdStrategyFactory(BaseFactory["PositionIdStrategy"]):
 
 @PositionIdStrategyFactory.register("none")
 class NoPositionId(PositionIdStrategy):
-    def generate(self, sequences):
+    def generate(self, sequences: List[List[int]]) -> List[int]:
         return []
 
 
 @PositionIdStrategyFactory.register("doc_reset")
 class DocResetPositionId(PositionIdStrategy):
-    def generate(self, sequences):
+    def generate(self, sequences: List[List[int]]) -> List[int]:
         pos_ids = []
         for seq in sequences:
             pos_ids.extend(range(len(seq)))
@@ -40,6 +41,6 @@ class DocResetPositionId(PositionIdStrategy):
 
 @PositionIdStrategyFactory.register("continuous")
 class ContinuousPositionId(PositionIdStrategy):
-    def generate(self, sequences):
+    def generate(self, sequences: List[List[int]]) -> List[int]:
         total = sum(len(seq) for seq in sequences)
         return list(range(total))