Kirigaya Kazuto
3b2a59b0ba
[pipeline/rank_recorder] fix bug when process data before backward | add a tool for multiple ranks debug ( #1681 )
...
* [pipeline/tuning] improve dispatch performance both time and space cost
* [pipeline/converge] add interface for testing convergence
* [NFC] polish colossalai/utils/multi_tensor_apply/multi_tensor_apply.py code style
* Update PipelineBase.py
* [pipeline/chimera] reconstruct PipelineBase and Worker to support more feasible custom schedule | finish Chimera
* [pipeline/chimera] test chimera | fix bug of initializing
* [pipeline/pytree] add pytree to process args and kwargs | provide to process args and kwargs after forward
2 years ago
YuliangLiu0306
517b63939a
[autoparallel] add unary element wise handler v2 ( #1674 )
2 years ago
YuliangLiu0306
f6c6a932b8
[autoparallel] add following node generator ( #1673 )
...
* [autoparallel] add following node generator
* polish code
* polish code
* update name of arguments
2 years ago
YuliangLiu0306
52fda88796
[autoparallel] add layer norm handler v2 ( #1671 )
...
* [autoparallel] add layer norm handler v2
* polish code
* polish code
2 years ago
Fazzie-Maqianli
87c5ad352a
update version to 0.1.10 ( #1676 )
2 years ago
HELSON
b28991dd0a
[feature] A new ZeRO implementation ( #1644 )
2 years ago
Boyuan Yao
b1be5b88bd
[autoparallel] fix insecure subprocess ( #1680 )
...
* [autoparallel] fix insecure subprocess
* [fx] fix insecure subprocess
2 years ago
Boyuan Yao
d8420f81a4
[hotfix] fix wrong type name in profiler ( #1678 )
2 years ago
Boyuan Yao
132b4306b7
[fx] Add concrete info prop ( #1677 )
...
* [fx] concreteinfoprop
* [fx] add concreteinfoprop
* [fx] modify docstring of ConcreteInfoProp
* [fx] fix device error
* [fx] modify parameter calculation
* [fx] modify parameters calculation
2 years ago
Boyuan Yao
1df98d5b66
[autoparallel] add rotor C version ( #1658 )
...
* [autoparallel] add rotor c version
* [fx] remove metainfoprop in rotor solver
* [autoparallel] modify C
code format
* [autoparallel] remove build.py
* [autoparallel] fix C extension build
* [autoparallel] add C solver consistency test
* [autoparallel] remove some unused imports
* [autoparallel] refactor rotor solver code
* [autoparallel] replace print with colossalai logger
* [autoparallel] ranks fixed
2 years ago
YuliangLiu0306
11ec070e53
[hotfix]unit test ( #1670 )
2 years ago
Frank Lee
a60024e77a
[autoparallel] added utils for broadcast operation ( #1665 )
...
* [autoparallel] added utils for broadcast operation
* polish code
2 years ago
YuliangLiu0306
3f068d1409
[autoparallel] update CommSpec ( #1667 )
2 years ago
Frank Lee
247a9dbca9
[autoparallel] added bias comm spec to matmul strategy ( #1664 )
2 years ago
YuliangLiu0306
746f8f979d
[autoparallel] add batch norm handler v2 ( #1666 )
2 years ago
Kirigaya Kazuto
9708638ded
[pipeline/pytree] add pytree to process args and kwargs | provide `data_process_func` to process args and kwargs after forward ( #1642 )
...
* [pipeline/tuning] improve dispatch performance both time and space cost
* [pipeline/converge] add interface for testing convergence
* [NFC] polish colossalai/utils/multi_tensor_apply/multi_tensor_apply.py code style
* Update PipelineBase.py
* [pipeline/chimera] reconstruct PipelineBase and Worker to support more feasible custom schedule | finish Chimera
* [pipeline/chimera] test chimera | fix bug of initializing
* [pipeline/pytree] add pytree to process args and kwargs | provide to process args and kwargs after forward
2 years ago
YuliangLiu0306
c27e701cb2
[autoparallel] remove no strategy nodes ( #1652 )
...
* [autoparallel] remove no strategy nodes
* fix none object iteration issue
2 years ago
Frank Lee
50f16a2850
[autoparallel] added compute resharding costs for node handler ( #1662 )
2 years ago
Frank Lee
9ec401a722
[autoparallel] added new strategy constructor template ( #1661 )
...
* [autoparallel] added new strategy constructor template
* polish code
2 years ago
Frank Lee
3a4d6f63a8
[autoparallel] added node handler for bmm ( #1655 )
2 years ago
YuliangLiu0306
095854477f
[autoparallel] add conv handler v2 ( #1663 )
2 years ago
YuliangLiu0306
1e7816a460
[autoparallel] adapt solver with gpt ( #1653 )
2 years ago
Jiarui Fang
c638bec028
[embedding] polish async copy ( #1657 )
2 years ago
Jiarui Fang
988570e4a6
[embedding] add more detail profiling ( #1656 )
2 years ago
Jiarui Fang
e1f97fd2b8
[embedding] print profiling results ( #1654 )
2 years ago
Frank Lee
30e50c8b4a
[autoparallel] implemented all matmul strategy generator ( #1650 )
2 years ago
YuliangLiu0306
03978aad45
[autoparallel] change the following nodes strategies generation logic ( #1636 )
...
* [autoparallel] change the following nodes strategies generation logic
* fix unit test
2 years ago
YuliangLiu0306
59f100510a
[autoparallel] where handler ( #1651 )
...
* [autoparallel] where handler
* fix unit test
2 years ago
Super Daniel
6135e178b3
[fx] refactor code for profiler / enable fake tensor movement. ( #1646 )
...
* [fx/profiling] provide summary for MetaInfoProp.
* [fx/profiler] provide a table of summary.
* [fx/profiler] provide a table of summary.
* [fx/profiler] provide a table of summary.
* [fx/profiler] provide a table of summary.
* [fx] optimize table repr.
* [fx] optimize table repr.
* [fx] refactor code for profiler.
* [fx] add docstring.
* [fx] add docstring.
* [fx] skip test.
* [fx] redo.
* [fx] redo.
* [fx] fix import error for torch11.
* [fx] fix import error for torch11.
2 years ago
Boyuan Yao
5d0fdb9cb4
[fx] fix offload codegen test ( #1648 )
...
* [fx] fix offload codegen test
* [fx] modify typing
2 years ago
Frank Lee
45b39a692a
[autoparallel] implemented linear projection strategy generator ( #1639 )
2 years ago
Frank Lee
154d3ef432
[fix] fixed the collective pattern name for consistency ( #1649 )
...
* [fix] fixed the collective pattern name for consistency
* polish code
2 years ago
YuliangLiu0306
b2b2a4af98
[autoparallel] adapt solver with mlp ( #1638 )
2 years ago
Jiarui Fang
04443605a5
[embedding] non-blocking cpu-gpu copy ( #1647 )
2 years ago
CsRic
0767f67a0f
[embedding] isolate cache_op from forward ( #1645 )
...
Co-authored-by: ric <mkkt_bkkt@mail.ustc.edu.cn>
2 years ago
Jiarui Fang
c5d39215f6
Revert "[feature] new zero implementation ( #1623 )" ( #1643 )
...
This reverts commit 5be118f405
.
2 years ago
HELSON
5be118f405
[feature] new zero implementation ( #1623 )
2 years ago
Boyuan Yao
f921733621
[autoparallel] Add pofo sequence annotation ( #1637 )
...
* [autoparallel] annotate pofo sequence
* [autoparallel] remove unused print
* [autoparallel] fix some code
2 years ago
Super Daniel
04bbabeea8
[fx/profiler] provide a table of summary. ( #1634 )
...
* [fx/profiling] provide summary for MetaInfoProp.
* [fx/profiler] provide a table of summary.
* [fx] optimize table repr.
2 years ago
HELSON
95c35f73bd
[moe] initialize MoE groups by ProcessGroup ( #1640 )
2 years ago
Jiarui Fang
e57df80325
[embeddings] cache option ( #1635 )
2 years ago
HELSON
a088022efc
[moe] fix moe bugs ( #1633 )
2 years ago
YuliangLiu0306
702dbc5288
[tensor] use communication autograd func ( #1617 )
...
* [tensor] use communication autograd func
* change all to all comm spec info
* rename pattern and distinguish fwd/bwd
* polish code
2 years ago
YuliangLiu0306
c7ac0f4ab2
[autoparallel] add elementwise handler ( #1622 )
...
* [autoparallel] add elementwise handler
* polish code
* polish code
* reduce skipped strategies range
* polish code
2 years ago
YuliangLiu0306
3a46215135
[autoparallel] add embedding handler ( #1620 )
2 years ago
YuliangLiu0306
69448f64c4
[autoparallel] protect bcast handler from invalid strategies ( #1631 )
2 years ago
YuliangLiu0306
0c703189b9
[autoparallel] add layernorm handler ( #1629 )
2 years ago
YuliangLiu0306
bf77d3ab65
[autoparallel] recover the merged node strategy index ( #1613 )
2 years ago
Boyuan Yao
d6b01feb66
[fx] Modify offload codegen ( #1618 )
...
* [fx] modify offload codegen
* [fx] remove repeated hook definitions
* [fx] modify offload test
2 years ago
YuliangLiu0306
9eae855408
[hotfix] add recompile after graph manipulatation ( #1621 )
2 years ago