Commit Graph

1049 Commits (6c4c6a0409d12eb9adfa84c49516148f960d5887)

Author SHA1 Message Date
ver217 a109225bc2 add sharded adam 2022-03-11 15:50:28 +08:00
Jiarui Fang e17e92c54d Polish sharded parameter (#297)
* init shard param from shape tuple

* add more unitest for shard param

* add more unittests to shareded param
2022-03-11 15:50:28 +08:00
ver217 7aef75ca42 [zero] add sharded grad and refactor grad hooks for ShardedModel (#287) 2022-03-11 15:50:28 +08:00
Frank Lee 9afb5c8b2d fixed typo in ShardParam (#294) 2022-03-11 15:50:28 +08:00
Frank Lee e17e54e32a added buffer sync to naive amp model wrapper (#291) 2022-03-11 15:50:28 +08:00
Jiarui Fang 8d653af408 add a common util for hooks registered on parameter. (#292) 2022-03-11 15:50:28 +08:00
Jie Zhu f867365aba bug fix: pass hook_list to engine (#273)
* bug fix: pass hook_list to engine

* change parameter name
2022-03-11 15:50:28 +08:00
Jiarui Fang 5a560a060a Feature/zero (#279)
* add zero1 (#209)

* add zero1

* add test zero1

* update zero stage 1 develop (#212)

* Implement naive zero3 (#240)

* naive zero3 works well

* add zero3 param manager

* add TODOs in comments

* add gather full param ctx

* fix sub module streams

* add offload

* fix bugs of hook and add unit tests

* fix bugs of hook and add unit tests (#252)

* add gather full param ctx

* fix sub module streams

* add offload

* fix bugs of hook and add unit tests

* polish code and add state dict hook

* fix bug

* update unit test

* refactor reconstructed zero code

* clip_grad support zero3 and add unit test

* add unit test for Zero3ParameterManager

* [WIP] initialize the shard param class

* [WIP] Yet another sharded model implementation (#274)

* [WIP] initialize the shard param class

* [WIP] Yes another implementation of shardModel. Using a better hook method.

* torch.concat -> torch.cat

* fix test_zero_level_1.py::test_zero_level_1 unitest

* remove deepspeed implementation and refactor for the reconstructed zero module

* polish zero dp unittests

Co-authored-by: ver217 <lhx0217@gmail.com>
Co-authored-by: Frank Lee <somerlee.9@gmail.com>
2022-03-11 15:50:28 +08:00
1SAA 82023779bb Added TPExpert for special situation 2022-03-11 15:50:28 +08:00
HELSON 36b8477228 Fixed parameter initialization in FFNExpert (#251) 2022-03-11 15:50:28 +08:00
アマデウス e13293bb4c fixed CI dataset directory; fixed import error of 2.5d accuracy (#255) 2022-03-11 15:50:28 +08:00
1SAA 219df6e685 Optimized MoE layer and fixed some bugs;
Decreased moe tests;

Added FFNExperts and ViTMoE model
2022-03-11 15:50:28 +08:00
zbian 3dba070580 fixed padding index issue for vocab parallel embedding layers; updated 3D linear to be compatible with examples in the tutorial 2022-03-11 15:50:28 +08:00
Frank Lee f5ca88ec97 fixed apex import (#227) 2022-02-15 11:31:13 +08:00
Frank Lee 3a1a9820b0 fixed mkdir conflict and align yapf config with flake (#220) 2022-02-15 11:31:13 +08:00
アマデウス 9ee197d0e9 moved env variables to global variables; (#215)
added branch context;
added vocab parallel layers;
moved split_batch from load_batch to tensor parallel embedding layers;
updated gpt model;
updated unit test cases;
fixed few collective communicator bugs
2022-02-15 11:31:13 +08:00
Frank Lee 812357d63c
fixed utils docstring and add example to readme (#200) 2022-02-03 11:37:17 +08:00
Frank Lee 765db512b5
fixed ddp bug on torch 1.8 (#194) 2022-01-28 15:14:04 +08:00
Jiarui Fang 569357fea0
add pytorch hooks (#179)
* add pytorch hooks
fix #175

* remove licenses in src code

* add gpu memory tracer

* replacing print with logger in ophooks.
2022-01-25 22:20:54 +08:00
ver217 708404d5f8
fix pipeline forward return tensors (#176) 2022-01-21 15:46:02 +08:00
HELSON 0f8c7f9804
Fixed docstring in colossalai (#171) 2022-01-21 10:44:30 +08:00
Frank Lee e2089c5c15
adapted for sequence parallel (#163) 2022-01-20 13:44:51 +08:00
puck_WCR 9473a1b9c8
AMP docstring/markdown update (#160) 2022-01-18 18:33:36 +08:00
Frank Lee f3802d6b06
fixed jit default setting (#154) 2022-01-18 13:37:20 +08:00
ver217 7bf1e98b97
pipeline last stage supports multi output (#151) 2022-01-17 15:57:47 +08:00
ver217 f68eddfb3d
refactor kernel (#142) 2022-01-13 16:47:17 +08:00
BoxiangW 4a3d3446b0
Update layer integration documentations (#108)
Update the documentations of layer integration

Update _log_hook.py

Update _operation.py
2022-01-10 18:05:58 +08:00
ver217 9ef05ed1fc
try import deepspeed when using zero (#130) 2022-01-07 17:24:57 +08:00
HELSON dceae85195
Added MoE parallel (#127) 2022-01-07 15:08:36 +08:00
ver217 293fb40c42
add scatter/gather optim for pipeline (#123) 2022-01-07 13:22:22 +08:00
Jiarui Fang 2c0c85d3d3
fix a bug in timer (#114) 2022-01-05 16:07:06 +08:00
ver217 7904baf6e1
fix layers/schedule for hybrid parallelization (#111) (#112) 2022-01-04 20:52:31 +08:00
ver217 a951bc6089
update default logger (#100) (#101) 2022-01-04 20:03:26 +08:00
ver217 96780e6ee4
Optimize pipeline schedule (#94)
* add pipeline shared module wrapper and update load batch

* added model parallel process group for amp and clip grad (#86)

* added model parallel process group for amp and clip grad

* update amp and clip with model parallel process group

* remove pipeline_prev/next group (#88)

* micro batch offload

* optimize pipeline gpu memory usage

* pipeline can receive tensor shape (#93)

* optimize pipeline gpu memory usage

* fix grad accumulation step counter

* rename classes and functions

Co-authored-by: Frank Lee <somerlee.9@gmail.com>
2021-12-30 15:56:46 +08:00
アマデウス 01a80cd86d
Hotfix/Colossalai layers (#92)
* optimized 1d layer apis; reorganized nn.layer modules; fixed tests

* fixed 2.5d runtime issue

* reworked split batch, now called in trainer.schedule.load_batch

Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-29 23:32:10 +08:00
アマデウス 0fedef4f3c
Layer integration (#83)
* integrated parallel layers for ease of building models

* integrated 2.5d layers

* cleaned codes and unit tests

* added log metric by step hook; updated imagenet benchmark; fixed some bugs

* reworked initialization; cleaned codes

Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-27 15:04:32 +08:00
shenggan 5c3843dc98
add colossalai kernel module (#55) 2021-12-21 12:19:52 +08:00
ver217 8f02a88db2
add interleaved pipeline, fix naive amp and update pipeline model initializer (#80) 2021-12-20 23:26:19 +08:00
Frank Lee 91c327cb44
fixed zero level 3 dtype bug (#76) 2021-12-20 17:00:53 +08:00
HELSON 632e622de8
overlap computation and communication in 2d operations (#75) 2021-12-16 16:05:15 +08:00
Frank Lee cd9c28e055
added CI for unit testing (#69) 2021-12-16 10:32:08 +08:00
Frank Lee 35813ed3c4
update examples and sphnix docs for the new api (#63) 2021-12-13 22:07:01 +08:00
ver217 7d3711058f
fix zero3 fp16 and add zero3 model context (#62) 2021-12-10 17:48:50 +08:00
Frank Lee 9a0466534c
update markdown docs (english) (#60) 2021-12-10 14:37:33 +08:00
Frank Lee da01c234e1
Develop/experiments (#59)
* Add gradient accumulation, fix lr scheduler

* fix FP16 optimizer and adapted torch amp with tensor parallel (#18)

* fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes

* fixed trainer

* Revert "fixed trainer"

This reverts commit 2e0b0b7699.

* improved consistency between trainer, engine and schedule (#23)

Co-authored-by: 1SAA <c2h214748@gmail.com>

* Split conv2d, class token, positional embedding in 2d, Fix random number in ddp
Fix convergence in cifar10, Imagenet1000

* Integrate 1d tensor parallel in Colossal-AI (#39)

* fixed 1D and 2D convergence (#38)

* optimized 2D operations

* fixed 1D ViT convergence problem

* Feature/ddp (#49)

* remove redundancy func in setup (#19) (#20)

* use env to control the language of doc (#24) (#25)

* Support TP-compatible Torch AMP and Update trainer API (#27)

* Add gradient accumulation, fix lr scheduler

* fix FP16 optimizer and adapted torch amp with tensor parallel (#18)

* fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes

* fixed trainer

* Revert "fixed trainer"

This reverts commit 2e0b0b7699.

* improved consistency between trainer, engine and schedule (#23)

Co-authored-by: 1SAA <c2h214748@gmail.com>

Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>

* add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29)

* add explanation for ViT example (#35) (#36)

* support torch ddp

* fix loss accumulation

* add log for ddp

* change seed

* modify timing hook

Co-authored-by: Frank Lee <somerlee.9@gmail.com>
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: binmakeswell <binmakeswell@gmail.com>

* Feature/pipeline (#40)

* remove redundancy func in setup (#19) (#20)

* use env to control the language of doc (#24) (#25)

* Support TP-compatible Torch AMP and Update trainer API (#27)

* Add gradient accumulation, fix lr scheduler

* fix FP16 optimizer and adapted torch amp with tensor parallel (#18)

* fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes

* fixed trainer

* Revert "fixed trainer"

This reverts commit 2e0b0b7699.

* improved consistency between trainer, engine and schedule (#23)

Co-authored-by: 1SAA <c2h214748@gmail.com>

Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>

* add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29)

* add explanation for ViT example (#35) (#36)

* optimize communication of pipeline parallel

* fix grad clip for pipeline

Co-authored-by: Frank Lee <somerlee.9@gmail.com>
Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: binmakeswell <binmakeswell@gmail.com>

* optimized 3d layer to fix slow computation ; tested imagenet performance with 3d; reworked lr_scheduler config definition; fixed launch args; fixed some printing issues; simplified apis of 3d layers (#51)

* Update 2.5d layer code to get a similar accuracy on imagenet-1k dataset

* update api for better usability (#58)

update api for better usability

Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>
Co-authored-by: puck_WCR <46049915+WANG-CR@users.noreply.github.com>
Co-authored-by: binmakeswell <binmakeswell@gmail.com>
Co-authored-by: アマデウス <kurisusnowdeng@users.noreply.github.com>
Co-authored-by: BoxiangW <45734921+BoxiangW@users.noreply.github.com>
2021-12-09 15:08:29 +08:00
ver217 dbe62c67b8
add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) 2021-11-18 23:45:09 +08:00
Frank Lee 3defa32aee
Support TP-compatible Torch AMP and Update trainer API (#27)
* Add gradient accumulation, fix lr scheduler

* fix FP16 optimizer and adapted torch amp with tensor parallel (#18)

* fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes

* fixed trainer

* Revert "fixed trainer"

This reverts commit 2e0b0b7699.

* improved consistency between trainer, engine and schedule (#23)

Co-authored-by: 1SAA <c2h214748@gmail.com>

Co-authored-by: 1SAA <c2h214748@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>
2021-11-18 19:45:06 +08:00
ver217 3c7604ba30 update documentation 2021-10-29 09:29:20 +08:00
zbian 404ecbdcc6 Migrated project 2021-10-28 18:21:23 +02:00