Commits · d16671da75ccb6aecacb2a15826a6d34367373d1 · OpenDAS / ColossalAI

28 Apr, 2022 3 commits
- [Tensor] initialize the ColoOptimizer (#898) · d16671da
  Jiarui Fang authored Apr 28, 2022
```
* [Tensor] activation is an attr of ColoTensor

* [Tensor] add optimizer

* only detach parameters in context

* polish code
```
  d16671da
- [Tensor] test parameters() as member function (#896) · e76f76c0
  Jiarui Fang authored Apr 28, 2022
  
  e76f76c0
- [tensor] refine linear and add gather for laynorm (#893) · cb182da7
  Ziyue Jiang authored Apr 28, 2022
```
* refine linear and add function to ColoTensor

* add gather for layernorm

* polish

* polish
```
  cb182da7
27 Apr, 2022 4 commits
- [Tensor] overriding paramters() for Module using ColoTensor (#889) · 26c49639
  Jiarui Fang authored Apr 27, 2022
  
  26c49639
- [tensor] add ColoTensor 1Dcol (#888) · 1d0aba41
  Ziyue Jiang authored Apr 27, 2022
  
  1d0aba41
- [Tensor] test model check results for a simple net (#887) · a0e59716
  Jiarui Fang authored Apr 27, 2022
  
  a0e59716
- [Tensor] make ColoTensor more robust for getattr (#886) · 72cdc068
  Jiarui Fang authored Apr 27, 2022
```
* [Tensor] make ColoTensor more robust for getattr

* polish

* polish
```
  72cdc068
26 Apr, 2022 7 commits
- [tensor] wrap function in the torch_tensor to ColoTensor (#881) · 9bc5a77c
  Ziyue Jiang authored Apr 26, 2022
  
  9bc5a77c
- [Tensor] make a simple net works with 1D row TP (#879) · 7f76517a
  Jiarui Fang authored Apr 26, 2022
  
  7f76517a
- [gemini] accelerate adjust_layout() (#878) · c4d903e6
  ver217 authored Apr 26, 2022
```
* add lru cache

* polish code

* update unit test

* fix sharded optim
```
  c4d903e6
- [Tensor] Add some attributes to ColoTensor (#877) · 90921145
  Jiarui Fang authored Apr 26, 2022
```
* [Tensor] add some function to ColoTensor

* torch.allclose

* rm torch.add
```
  90921145
- [Tensor] get named parameters for model using ColoTensors (#874) · e43f83aa
  Jiarui Fang authored Apr 26, 2022
  
  e43f83aa
- [tensor] customized op returns ColoTensor (#875) · 96211c2c
  Jiarui Fang authored Apr 26, 2022
```
* [tensor] customized op returns ColoTensor

* polish

* polish code
```
  96211c2c
- [Tensor] Add function to spec and update linear 1Drow and unit tests (#869) · 26d4ab8b
  Ziyue Jiang authored Apr 26, 2022
  
  26d4ab8b
25 Apr, 2022 5 commits
- [tensor] add cross_entrophy_loss (#868) · 1190b2c4
  Jiarui Fang authored Apr 25, 2022
  
  1190b2c4
- [gemini] add stateful tensor container (#867) · 31078171
  HELSON authored Apr 25, 2022
  
  31078171
- colo init context add device attr. (#866) · d01d3b8c
  Jiarui Fang authored Apr 25, 2022
  
  d01d3b8c
- [Tensor] add layer norm Op (#852) · 126ba573
  Jiarui Fang authored Apr 25, 2022
  
  126ba573
- [ci] cache cuda extension (#860) · 1258af71
  Frank Lee authored Apr 25, 2022
  
  1258af71
24 Apr, 2022 7 commits
- [Tensor ] Add 1Drow weight reshard by spec (#854) · bcc86550
  Ziyue Jiang authored Apr 24, 2022
  
  bcc86550
- [Tensor] init a tp network training unittest (#849) · 62f05925
  Jiarui Fang authored Apr 24, 2022
  
  62f05925
- [tensor]add assert for colo_tensor 1Drow (#846) · 2a0a427e
  Ziyue Jiang authored Apr 24, 2022
  
  2a0a427e
- [Tensor] TP Linear 1D row (#843) · 05023ecf
  Ziyue Jiang authored Apr 24, 2022
  
  05023ecf
- [gemini] add GeminiMemoryManger (#832) · e5ea3fde
  HELSON authored Apr 24, 2022
```
* refactor StatefulTensor, tensor utilities

* add unitest for GeminiMemoryManager
```
  e5ea3fde
- [pipelinable]use pipelinable context to initialize non-pipeline model (#816) · 35ea6e10
  YuliangLiu0306 authored Apr 24, 2022
```
* [CLI] add CLI launcher

* Revert "[CLI] add CLI launcher"

This reverts commit df7e6506d4500af6a9220ef7fe4d3c7b1daebd4c.

* [pipeline]add module lazy init feature to support large model initization.

* [pipeline]add to_layer_list and partition method to support arbitrary non-pp model

* refactor the module structure

* polish

* [pipelinable]add unit test for pipelinable

* polish

* polish

* Fix CodeFactor issues.
```
  35ea6e10
- [hotfix] the bug of numel() in ColoTensor (#845) · ea0a2ed2
  Jiarui Fang authored Apr 24, 2022
  
  ea0a2ed2
22 Apr, 2022 6 commits
- Init Conext supports lazy allocate model memory (#842) · 8789850e
  Jiarui Fang authored Apr 22, 2022
  
  8789850e
- [unittest] refactored unit tests for change in dependency (#838) · 943982d2
  Frank Lee authored Apr 22, 2022
  
  943982d2
- [dependency] removed torchvision (#833) · 01e9f834
  Frank Lee authored Apr 22, 2022
```
* [dependency] removed torchvision

* fixed transforms
```
  01e9f834
- Revert "[WIP] Applying ColoTensor on TP-1D-row Linear. (#831)" (#835) · cb5a4778
  Jiarui Fang authored Apr 22, 2022
```
This reverts commit ac88de6d.
```
  cb5a4778
- [WIP] Applying ColoTensor on TP-1D-row Linear. (#831) · ac88de6d
  Jiarui Fang authored Apr 22, 2022
```
* revert zero tensors back

* [tensor] init row 1d linear
```
  ac88de6d
- [tensor] ZeRO use ColoTensor as the base class. (#828) · 294a6060
  Jiarui Fang authored Apr 22, 2022
```
* [refactor] moving InsertPostInitMethodToModuleSubClasses to utils.

* [tensor] ZeRO use ColoTensor as the base class.

* polish
```
  294a6060
21 Apr, 2022 6 commits

[tensor]fix test_linear (#826) · 8e6fdb4f
Ziyue Jiang authored Apr 21, 2022

8e6fdb4f
[tensor] fix kwargs in colo_tensor torch_funtion (#825) · 1a9e2c2d
Ziyue Jiang authored Apr 21, 2022

1a9e2c2d
[tensor] lazy init (#823) · 2ecc3d7a
Jiarui Fang authored Apr 21, 2022

2ecc3d7a

[Tensor] apply ColoTensor on Torch functions (#821) · 660d2d1f

Jiarui Fang authored Apr 21, 2022

* Revert "[zero] add ZeroTensorShardStrategy (#793)"

This reverts commit 88759e28.

* [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

660d2d1f

[tensor] reorganize files (#820) · 0ce8924c
Jiarui Fang authored Apr 21, 2022

0ce8924c

[gemini] a new tensor structure (#818) · ab962b97

Jiarui Fang authored Apr 21, 2022

* Revert "[zero] add ZeroTensorShardStrategy (#793)"

This reverts commit 88759e28.

* [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

ab962b97

19 Apr, 2022 2 commits
- Revert "[zero] add ZeroTensorShardStrategy (#793)" (#806) · e761ad2c
  Jiarui Fang authored Apr 19, 2022
  
  e761ad2c
- [zero] add ZeroTensorShardStrategy (#793) · 88759e28
  HELSON authored Apr 19, 2022
  
  88759e28