Ziyue Jiang
|
1d0aba4153
|
[tensor] add ColoTensor 1Dcol (#888)
|
3 years ago |
Jiarui Fang
|
72cdc06875
|
[Tensor] make ColoTensor more robust for getattr (#886)
* [Tensor] make ColoTensor more robust for getattr
* polish
* polish
|
3 years ago |
Ziyue Jiang
|
9bc5a77c31
|
[tensor] wrap function in the torch_tensor to ColoTensor (#881)
|
3 years ago |
Jiarui Fang
|
7f76517a85
|
[Tensor] make a simple net works with 1D row TP (#879)
|
3 years ago |
Jiarui Fang
|
909211453b
|
[Tensor] Add some attributes to ColoTensor (#877)
* [Tensor] add some function to ColoTensor
* torch.allclose
* rm torch.add
|
3 years ago |
Ziyue Jiang
|
26d4ab8b03
|
[Tensor] Add function to spec and update linear 1Drow and unit tests (#869)
|
3 years ago |
Jiarui Fang
|
d01d3b8cb0
|
colo init context add device attr. (#866)
|
3 years ago |
Jiarui Fang
|
126ba573a8
|
[Tensor] add layer norm Op (#852)
|
3 years ago |
YuliangLiu0306
|
c6930d8ddf
|
[pipelinable]use ColoTensor to replace dummy tensor. (#853)
|
3 years ago |
Ziyue Jiang
|
bcc8655021
|
[Tensor ] Add 1Drow weight reshard by spec (#854)
|
3 years ago |
Jiarui Fang
|
62f059251b
|
[Tensor] init a tp network training unittest (#849)
|
3 years ago |
Ziyue Jiang
|
2a0a427e04
|
[tensor]add assert for colo_tensor 1Drow (#846)
|
3 years ago |
Ziyue Jiang
|
05023ecfee
|
[Tensor] TP Linear 1D row (#843)
|
3 years ago |
Jiarui Fang
|
ea0a2ed25f
|
[hotfix] the bug of numel() in ColoTensor (#845)
|
3 years ago |
Jiarui Fang
|
8789850eea
|
Init Conext supports lazy allocate model memory (#842)
|
3 years ago |
Jiarui Fang
|
4575a3298b
|
[hotfix] ColoTensor pin_memory (#840)
|
3 years ago |
Jiarui Fang
|
cb5a4778e1
|
Revert "[WIP] Applying ColoTensor on TP-1D-row Linear. (#831)" (#835)
This reverts commit ac88de6dfc .
|
3 years ago |
Jiarui Fang
|
ac88de6dfc
|
[WIP] Applying ColoTensor on TP-1D-row Linear. (#831)
* revert zero tensors back
* [tensor] init row 1d linear
|
3 years ago |
Jiarui Fang
|
294a6060d0
|
[tensor] ZeRO use ColoTensor as the base class. (#828)
* [refactor] moving InsertPostInitMethodToModuleSubClasses to utils.
* [tensor] ZeRO use ColoTensor as the base class.
* polish
|
3 years ago |
Ziyue Jiang
|
1a9e2c2dff
|
[tensor] fix kwargs in colo_tensor torch_funtion (#825)
|
3 years ago |
Jiarui Fang
|
2ecc3d7a55
|
[tensor] lazy init (#823)
|
3 years ago |
Jiarui Fang
|
68dcd51d41
|
[Tensor] update ColoTensor torch_function (#822)
* Revert "[zero] add ZeroTensorShardStrategy (#793)"
This reverts commit 88759e289e .
* [gemini] set cpu memory capacity
* [log] local throughput collecting
* polish
* polish
* polish
* polish code
* polish
* polish code
* add a new tensor structure and override linear for it
* polish
* polish
* polish
* polish
* polish
* polish
* polish
* polish
* polish
* polish
* polish
* [tensor] renaming and reorganize directory structure.
* rm useless dir
* polish
* polish
* [tensor] hander the function not wrapped
* polish
|
3 years ago |
Jiarui Fang
|
0ce8924ceb
|
[tensor] reorganize files (#820)
|
3 years ago |