AstrAI/astrai
ViperEkura fd65b9bc23 feat: support HEAD_DIM=32 and split extension into loader/ops
- add case 32 to decode/prefill dispatch switch
- fix swiz_col out-of-bounds for HEAD_DIM=32: XOR mask now limited to chunk count (3 for 32, 7 for >=64) instead of always 7, which produced column offsets >= LD=32 and corrupted shared memory
- restructure decode dispatch to #ifndef/#else/#endif matching prefill
- split astrai/extension/__init__.py into loader.py (kernel .so discovery) and ops.py (wrapper functions + torch SDPA fallback); __init__.py now re-exports the public API
2026-07-08 14:14:11 +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
extension feat: support HEAD_DIM=32 and split extension into loader/ops 2026-07-08 14:14:11 +08:00
inference refactor: clean up inference design patterns 2026-07-05 11:41:54 +08:00
model perf: replace paged KV cache with contiguous ContiguousCache, decode all groups 2026-07-05 11:34:36 +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 fix: support single .h5 file loading in load_h5 2026-07-07 00:11:16 +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: set tqdm postfix before update so first step shows metrics 2026-07-07 00:14:13 +08:00
__init__.py refactor: extension dispatch layer with CUDA/torch fallback 2026-07-06 21:07:16 +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