refactor: 重构打包模块,新增 BFD/FFD/Greedy 三种 bin-packing 算法,默认 BFD

- 将 pipeline/packing.py 拆分为 packing/ 子包 (base/stream/binpack)
- 新增 BfdPacker(默认)/FfDPacker/GreedyPacker,移除 StreamingPacker
- 超长序列直接截断至 pack_size
- group_size 语义改为"每 N 个 chunk 合并为一块",默认 1000
- 新增 AutoTokenizer.token_to_id(),修复 ChatML 中 hacky 的 nl_id 获取
- pad_value 默认改为 2(pad_token_id),position_ids pad=0, loss_mask pad=False
- 新增 position_ids 打包后归零一致性测试
- scripts/cache_h5.py 新增 --pack-algo 参数
This commit is contained in:
2026-07-03 16:17:27 +08:00
parent 2f919e9243
commit 598e1ce4ae
14 changed files with 715 additions and 295 deletions
+6
View File
@@ -266,6 +266,12 @@ class AutoTokenizer:
return self._tokenizer.decode(tokens, skip_special_tokens=skip_special_tokens)
def token_to_id(self, token: str) -> Optional[int]:
"""Convert a token string to its integer ID."""
if self._tokenizer is None:
raise RuntimeError("Tokenizer not initialized.")
return self._tokenizer.token_to_id(token)
def __len__(self) -> int:
if self._tokenizer is None:
return 0