ViperEkura 4e508afa2d fix : SFT pipeline position_ids default & doc boundary preservation ...

- change position_ids_mode default from "none" to "doc_reset" so SFT preprocessing always generates position_ids (was causing dataset load KeyError)
- generate per-doc position_ids before packing (doc_reset mode), preserving document boundaries for BFD packing (cross-doc attention leak fix)
- change _align_bucket padding from [1] to [0] to avoid accidentally training on loss_mask padding

2026-07-04 15:59:11 +08:00

..

config

fix : SFT pipeline position_ids default & doc boundary preservation

2026-07-04 15:59:11 +08:00

dataset

feat: add JSONL dataset store with on-the-fly tokenization

2026-07-04 15:42:33 +08:00

inference

fix: 使用 threading.Event 替代裸 bool，补全公共 API

2026-06-18 15:38:35 +08:00

model

feat : GPT-2 residual scaling weight init

2026-06-25 15:08:31 +08:00

parallel

refactor : grad_norm 指标简化，clip_grad_norm 移至 executor

2026-06-30 14:59:43 +08:00

preprocessing

fix : SFT pipeline position_ids default & doc boundary preservation

2026-07-04 15:59:11 +08:00

serialization

feat: add JSONL dataset store with on-the-fly tokenization

2026-07-04 15:42:33 +08:00

tokenize

refactor : Pipeline 去除去重，ids 重命名为 sequence，泛型透传

2026-05-31 15:14:27 +08:00

trainer

fix: scheduler division-by-zero, loss_mask bool

2026-07-03 22:04:55 +08:00

__init__.py

chore : 升级 torch 2.11.0+cu128，移除自定义 Muon，修复 gloo device_id

2026-06-27 16:10:37 +08:00

factory.py

fix: ForwardRef._evaluate Python 3.12 compatibility

2026-07-03 18:41:19 +08:00

protocols.py

refactor: 重构训练后端为 Executor 模式

2026-05-24 20:35:44 +08:00