ColossalAI/colossalai/legacy
Edenzzzz f5c84af0b0
[Feature] Zigzag Ring attention (#5905)
* halfway

* fix cross-PP-stage position id length diff bug

* fix typo

* fix typo

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* unified cross entropy func for all shardformer models

* remove redundant lines

* add basic ring attn; debug cross entropy

* fwd bwd logic complete

* fwd bwd logic complete; add experimental triton rescale

* precision tests passed

* precision tests passed

* fix typos and remove misc files

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* add sp_mode to benchmark; fix varlen interface

* update softmax_lse shape by new interface

* change tester name

* remove buffer clone; support packed seq layout

* add varlen tests

* fix typo

* all tests passed

* add dkv_group; fix mask

* remove debug statements

---------

Co-authored-by: Edenzzzz <wtan45@wisc.edu>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2024-08-16 13:56:38 +08:00
..
amp [feat] refactored extension module (#5298) 2024-01-25 17:01:48 +08:00
builder [misc] update pre-commit and run all files (#4752) 2023-09-19 14:20:26 +08:00
communication [npu] change device to accelerator api (#5239) 2024-01-09 10:20:05 +08:00
context [pre-commit.ci] pre-commit autoupdate (#5572) 2024-07-01 17:16:41 +08:00
engine [npu] change device to accelerator api (#5239) 2024-01-09 10:20:05 +08:00
inference [pre-commit.ci] pre-commit autoupdate (#5572) 2024-07-01 17:16:41 +08:00
moe [Feature] Zigzag Ring attention (#5905) 2024-08-16 13:56:38 +08:00
nn [Feature] Zigzag Ring attention (#5905) 2024-08-16 13:56:38 +08:00
pipeline [misc] refactor launch API and tensor constructor (#5666) 2024-04-29 10:40:11 +08:00
registry [misc] update pre-commit and run all files (#4752) 2023-09-19 14:20:26 +08:00
tensor Fix ColoTensorSpec for py11 (#5440) 2024-03-26 15:56:49 +08:00
trainer [npu] change device to accelerator api (#5239) 2024-01-09 10:20:05 +08:00
utils [feat] refactored extension module (#5298) 2024-01-25 17:01:48 +08:00
zero Merge branch 'main' into sync/npu 2024-01-18 12:05:21 +08:00
__init__.py [bug] fix get_default_parser in examples (#4764) 2023-09-21 10:42:25 +08:00
constants.py [misc] update pre-commit and run all files (#4752) 2023-09-19 14:20:26 +08:00
core.py [misc] update pre-commit and run all files (#4752) 2023-09-19 14:20:26 +08:00
global_variables.py [misc] update pre-commit and run all files (#4752) 2023-09-19 14:20:26 +08:00
initialize.py [npu] change device to accelerator api (#5239) 2024-01-09 10:20:05 +08:00