GitLab

* [zero] refactor low-level zero folder structure

* [zero] fix legacy zero import path

* [zero] fix legacy zero import path

* [zero] remove useless import

* [zero] refactor gemini folder structure

* [zero] refactor gemini folder structure

* [zero] refactor legacy zero import path

* [zero] refactor gemini folder structure

* [zero] refactor gemini folder structure

* [zero] refactor gemini folder structure

* [zero] refactor legacy zero import path

* [zero] fix test import path

* [zero] fix test

* [zero] fix circular import

* [zero] update import

* [hotfix] skip torchaudio tracing test

* fix lazy init test issue

* [lazyinit] lazy tensor add distribute

* [lazyinit] refactor distribute

* [lazyinit] add test dist lazy init

* [lazyinit] add verbose info for dist lazy init

* [lazyinit] fix rnn flatten weight op

* [lazyinit] polish test

* [lazyinit] polish test

* [lazyinit] fix lazy tensor data setter

* [lazyinit] polish test

* [lazyinit] fix clean

* [lazyinit] make materialize inplace

* [lazyinit] refactor materialize

* [lazyinit] refactor test distribute

* [lazyinit] fix requires_grad

* [lazyinit] fix tolist after materialization

* [lazyinit] refactor distribute module

* [lazyinit] polish docstr

* [lazyinit] polish lazy init context

* [lazyinit] temporarily skip test

* [lazyinit] polish test

* [lazyinit] add docstr

* [lazyinit] fix shared module

* [tests] add lazy init test utils

* [tests] add torchvision for lazy init

* [lazyinit] fix pre op fn

* [lazyinit] handle legacy constructor

* [tests] refactor lazy init test models

* [tests] refactor lazy init test utils

* [lazyinit] fix ops don't support meta

* [tests] lazy init test timm models

* [lazyinit] fix set data

* [lazyinit] handle apex layers

* [tests] lazy init test transformers models

* [tests] lazy init test torchaudio models

* [lazyinit] fix import path

* [tests] lazy init test torchrec models

* [tests] update torch version in CI

* [tests] revert torch version in CI

* [tests] skip lazy init test

* [utils] Add use_reetrant=False into colossalai checkpoint

* [utils] add some annotation in utils.activaion_checkpoint

* [test] add reset_seed at the beginning of tests in test_actiavion_checkpointing.py

* [test] modify test_activation_checkpoint.py

* [test] modify test for reentrant=False

[colotensor] add megatron initialization for gpt2

* [utils] integrated colotensor with lazy init context

* polish code

* polish code

* polish code

* [CLI] add CLI launcher

* Revert "[CLI] add CLI launcher"

This reverts commit df7e6506d4500af6a9220ef7fe4d3c7b1daebd4c.

* [context]support arbitary module materialization.

* [test]add numerical check for lazy init context.

* [CLI] add CLI launcher

* Revert "[CLI] add CLI launcher"

This reverts commit df7e6506d4500af6a9220ef7fe4d3c7b1daebd4c.

* [context]use meta tensor to init model lazily.

* polish

* make module with device kwargs bypass the normal init.

* change unit test to adapt updated context.

* add save/load state dict for ColoDDP

* add unit test

* refactor unit test folder

* polish unit test

* rename unit test

* add set_params_to_ignore for ColoDDP

* polish code

* fix zero hook v2

* add unit test

* polish docstr

* [pipeline] refactor the pipeline module

* polish code

* [context] support lazy init of module

* polish code

* [test] skip tests when not enough GPUs are detected

* polish code

* polish code

* [test] ignore 8 gpu test

* polish code

* polish workflow

* polish workflow