Commit Graph

41 Commits (1ce997daaf7ffdf6c93a7bbd179d7be18928ba41)

Author SHA1 Message Date
shenggan 798cb72907 [NFC] polish applications/Chat/coati/trainer/base.py code style (#4260) 2023-07-26 14:12:57 +08:00
Zheng Zangwei (Alex Zheng) b2debdc09b [NFC] polish applications/Chat/coati/dataset/sft_dataset.py code style (#4259) 2023-07-26 14:12:57 +08:00
Junming Wu 77c469e1ba [NFC] polish applications/Chat/coati/models/base/actor.py code style (#4248) 2023-07-26 14:12:57 +08:00
Frank Lee f447ca1811 [chat] removed cache file (#4155) 2023-07-04 16:05:01 +08:00
wukong1992 c1c672d0f0 [shardformer] shardformer support t5 model (#3994)
test t5
2023-07-04 16:05:01 +08:00
Wenhao Chen 3d8d5d0d58
[chat] use official transformers and fix some issues (#4117)
* feat: remove on_learn_epoch fn as not used

* revert: add _on_learn_epoch fn

* feat: remove NaiveStrategy

* test: update train_prompts tests

* fix: remove prepare_llama_tokenizer_and_embedding

* test: add lora arg

* feat: remove roberta support in train_prompts due to runtime errs

* feat: remove deberta & roberta in rm as not used

* test: remove deberta and roberta tests

* feat: remove deberta and roberta models as not used

* fix: remove calls to roberta

* fix: remove prepare_llama_tokenizer_and_embedding

* chore: update transformers version

* docs: update transformers version

* fix: fix actor inference

* fix: fix ci

* feat: change llama pad token to unk

* revert: revert ddp setup_distributed

* fix: change llama pad token to unk

* revert: undo unnecessary changes

* fix: use pip to install transformers
2023-07-04 13:49:09 +08:00
Wenhao Chen edd75a59ea
[chat] remove naive strategy and split colossalai strategy (#4094)
* feat: remove on_learn_epoch fn as not used

* revert: add _on_learn_epoch fn

* to: remove the use of NaiveStrategy

* test: remove NaiveStrategy tests

* feat: remove NaiveStrategy

* style: modify comments and params

* feat: split ColossalAIStrategy into LowLevelZeroStrategy and GeminiStrategy

* fix: remove naive

* fix: align with modified colossal strategy

* fix: fix ddp _try_init_dist arg
2023-06-29 18:11:00 +08:00
Wenhao Chen b03d64d010
[chat] refactor trainer class (#4080)
* to: add SLTrainer

* refactor: refactor RMTrainer and SFTTrainer

* fix: fix init file

* feat: remove on_learn_epoch fn as not used

* fix: align with modified gemini arguments

* to: add OnPolicyTrainer

* revert: add _on_learn_epoch fn

* refactor: refactor PPOTrainer

* style: rename PPOTrainer argument

* fix: align with modified PPO arguments

* test: align with modified train_prompts arguments

* chore: modify train_prompts

* docs: align with modified arguments

* fix: remove unnecessary output

* fix: move dataloader to fit fn of SLTrainer

* fix: move dataloader to fit fn of OnPolicyTrainer

* fix: modify usage of prompt and pretrain dataloader
2023-06-29 10:48:09 +08:00
Baizhou Zhang 4da324cd60
[hotfix]fix argument naming in docs and examples (#4083) 2023-06-26 23:50:04 +08:00
Wenhao Chen 153b957a1b
[chat] refactor strategy class with booster api (#3987)
* refactor: adapt boost API in base and naive strategies

* fix: initialize plugin after setup_distributed

* fix: fix save_pretrained fn

* refactor: adapt boost API in DDPStrategy

* to: add _post_init check

* to: fix ddp backward, modify ddp dataloader and unwrap

* feat: adapt boost API in ColossalAIStrategy

* fix: call setup_distributed before use get_current_device

* fix: fix save_model and save_optimizer

* test: remove save_sharded_optimizer test

* style: apply formatter

* fix: fix stage check and add comments

* feat: allow dict type arg in strategy.prepare

* to: temporarily remove lr_scheduler for testing

* style: simplify init of ColossalAIStrategy

* fix: fix lr_scheduler in sft and rm

* style: modify comments

* test: add train_prompts tests

* fix: fix inference only case and use in train_prompts

* test: skip failed tests in ci

* style: fix CodeFactor check

* fix: do not use model.to('cpu') with GeminiPlugin

* test: enable colossalai_gemini tests

* test: set CUDA_VISIBLE_DEVICES in ci

* docs: add note
2023-06-25 17:36:21 +08:00
digger yu d4fb7bfda7
fix typo applications/Chat/coati/ (#3947) 2023-06-15 10:43:11 +08:00
Wenhao Chen 9d02590c9a
[chat] refactor actor class (#3968)
* refactor: separate log_probs fn from Actor forward fn

* refactor: separate generate fn from Actor class

* feat: update unwrap_model and get_base_model
* unwrap_model returns model not wrapped by Strategy
* get_base_model returns HF model for Actor, Critic and RewardModel

* feat: simplify Strategy.prepare

* style: remove get_base_model method of Actor

* perf: tokenize text in batches

* refactor: move calc_action_log_probs to utils of model

* test: update test with new forward fn

* style: rename forward fn args

* fix: do not unwrap model in save_model fn of naive strategy

* test: add gemini test for train_prompts

* fix: fix _set_default_generate_kwargs
2023-06-13 13:31:56 +08:00
Hongxin Liu b5f0566363
[chat] add distributed PPO trainer (#3740)
* Detached ppo (#9)

* run the base

* working on dist ppo

* sync

* detached trainer

* update detached trainer. no maker update function

* facing init problem

* 1 maker 1 trainer detached run. but no model update

* facing cuda problem

* fix save functions

* verified maker update

* nothing

* add ignore

* analyize loss issue

* remove some debug codes

* facing 2m1t stuck issue

* 2m1t verified

* do not use torchrun

* working on 2m2t

* working on 2m2t

* initialize strategy in ray actor env

* facing actor's init order issue

* facing ddp model update issue (need unwarp ddp)

* unwrap ddp actor

* checking 1m2t stuck problem

* nothing

* set timeout for trainer choosing. It solves the stuck problem!

* delete some debug output

* rename to sync with upstream

* rename to sync with upstream

* coati rename

* nothing

* I am going to detach the replaybuffer from trainer and make it a Ray Actor. Two benefits: 1. support TP trainer. 2. asynchronized buffer operations

* experience_maker_holder performs target-revolving _send_experience() instead of length comparison.

* move code to ray subfolder

* working on pipeline inference

* apply comments

* working on pipeline strategy. in progress.

* remove pipeline code. clean this branch

* update remote parameters by state_dict. no test

* nothing

* state_dict sharding transfer

* merge debug branch

* gemini _unwrap_model fix

* simplify code

* simplify code & fix LoRALinear AttributeError

* critic unwrapped state_dict

---------

Co-authored-by: csric <richcsr256@gmail.com>

* [chat] add perfomance evaluator and fix bugs (#10)

* [chat] add performance evaluator for ray

* [chat] refactor debug arg

* [chat] support hf config

* [chat] fix generation

* [chat] add 1mmt dummy example

* [chat] fix gemini ckpt

* split experience to send (#11)

Co-authored-by: csric <richcsr256@gmail.com>

* [chat] refactor trainer and maker (#12)

* [chat] refactor experience maker holder

* [chat] refactor model init

* [chat] refactor trainer args

* [chat] refactor model init

* [chat] refactor trainer

* [chat] refactor experience sending logic and training loop args (#13)

* [chat] refactor experience send logic

* [chat] refactor trainer

* [chat] refactor trainer

* [chat] refactor experience maker

* [chat] refactor pbar

* [chat] refactor example folder (#14)

* [chat] support quant (#15)

* [chat] add quant

* [chat] add quant example

* prompt example (#16)

* prompt example

* prompt load csv data

* remove legacy try

---------

Co-authored-by: csric <richcsr256@gmail.com>

* [chat] add mmmt dummy example and refactor experience sending (#17)

* [chat] add mmmt dummy example

* [chat] refactor naive strategy

* [chat] fix struck problem

* [chat] fix naive strategy

* [chat] optimize experience maker sending logic

* [chat] refactor sending assignment

* [chat] refactor performance evaluator (#18)

* Prompt Example & requires_grad state_dict & sharding state_dict (#19)

* prompt example

* prompt load csv data

* remove legacy try

* maker models require_grad set to False

* working on zero redundancy update

* mmmt_prompt example; naive strategy requires_grad state_dict & sharding; maker model requires_no_grad.

* remove legacy examples

* remove legacy examples

* remove replay buffer tp state. bad design

---------

Co-authored-by: csric <richcsr256@gmail.com>

* state_dict sending adapts to new unwrap function (#20)

* prompt example

* prompt load csv data

* remove legacy try

* maker models require_grad set to False

* working on zero redundancy update

* mmmt_prompt example; naive strategy requires_grad state_dict & sharding; maker model requires_no_grad.

* remove legacy examples

* remove legacy examples

* remove replay buffer tp state. bad design

* opt benchmark

* better script

* nothing

* [chat] strategy refactor unwrap model

* [chat] strategy refactor save model

* [chat] add docstr

* [chat] refactor trainer save model

* [chat] fix strategy typing

* [chat] refactor trainer save model

* [chat] update readme

* [chat] fix unit test

* working on lora reconstruction

* state_dict sending adapts to new unwrap function

* remove comments

---------

Co-authored-by: csric <richcsr256@gmail.com>
Co-authored-by: ver217 <lhx0217@gmail.com>

* [chat-ray] add readme (#21)

* add readme

* transparent graph

* add note background

---------

Co-authored-by: csric <richcsr256@gmail.com>

* [chat] get images from url (#22)

* Refactor/chat ray (#23)

* [chat] lora add todo

* [chat] remove unused pipeline strategy

* [chat] refactor example structure

* [chat] setup ci for ray

* [chat-ray] Support LoRA trainer. LoRA weights reconstruction. (#24)

* lora support prototype

* lora support

* 1mmt lora & remove useless code

---------

Co-authored-by: csric <richcsr256@gmail.com>

* [chat] fix test ci for ray

* [chat] fix test ci requirements for ray

* [chat] fix ray runtime env

* [chat] fix ray runtime env

* [chat] fix example ci docker args

* [chat] add debug info in trainer

* [chat] add nccl debug info

* [chat] skip ray test

* [doc] fix typo

---------

Co-authored-by: csric <59389055+CsRic@users.noreply.github.com>
Co-authored-by: csric <richcsr256@gmail.com>
2023-06-07 10:41:16 +08:00
digger yu e2d81eba0d
[nfc] fix typo colossalai/ applications/ (#3831)
* fix typo colossalai/autochunk auto_parallel amp

* fix typo colossalai/auto_parallel nn utils etc.

* fix typo colossalai/auto_parallel autochunk fx/passes  etc.

* fix typo docs/

* change placememt_policy to placement_policy in docs/ and examples/

* fix typo colossalai/ applications/
2023-05-25 16:19:41 +08:00
digger yu 9265f2d4d7
[NFC]fix typo colossalai/auto_parallel nn utils etc. (#3779)
* fix typo colossalai/autochunk auto_parallel amp

* fix typo colossalai/auto_parallel nn utils etc.
2023-05-23 15:28:20 +08:00
Yuanchen 05759839bd
[chat] fix bugs in stage 3 training (#3759)
Co-authored-by: Yuanchen Xu <yuanchen.xu00@gmail.com>
2023-05-17 17:44:05 +08:00
digger-yu b7141c36dd
[CI] fix some spelling errors (#3707)
* fix spelling error with examples/comminity/

* fix spelling error with tests/

* fix some spelling error with tests/ colossalai/ etc.
2023-05-10 17:12:03 +08:00
Hongxin Liu 7bd0bee8ea
[chat] add opt attn kernel (#3655)
* [chat] add opt attn kernel

* [chat] disable xformer during fwd
2023-05-04 16:03:33 +08:00
tanitna 1a60dc07a8
[chat] typo accimulation_steps -> accumulation_steps (#3662) 2023-04-28 15:42:57 +08:00
Hongxin Liu 842768a174
[chat] refactor model save/load logic (#3654)
* [chat] strategy refactor unwrap model

* [chat] strategy refactor save model

* [chat] add docstr

* [chat] refactor trainer save model

* [chat] fix strategy typing

* [chat] refactor trainer save model

* [chat] update readme

* [chat] fix unit test
2023-04-27 18:41:49 +08:00
Hongxin Liu 6ef7011462
[chat] remove lm model class (#3653)
* [chat] refactor lora

* [chat] remove lm class

* [chat] refactor save model

* [chat] refactor train sft

* [chat] fix ci

* [chat] fix ci
2023-04-27 15:37:38 +08:00
Hongxin Liu 2a951955ad
[chat] refactor trainer (#3648)
* [chat] ppo trainer remove useless args

* [chat] update examples

* [chat] update benchmark

* [chat] update examples

* [chat] fix sft training with wandb

* [chat] polish docstr
2023-04-26 18:11:49 +08:00
Hongxin Liu f8288315d9
[chat] polish performance evaluator (#3647) 2023-04-26 17:34:59 +08:00
Hongxin Liu 50793b35f4
[gemini] accelerate inference (#3641)
* [gemini] support don't scatter after inference

* [chat] update colossalai strategy

* [chat] fix opt benchmark

* [chat] update opt benchmark

* [gemini] optimize inference

* [test] add gemini inference test

* [chat] fix unit test ci

* [chat] fix ci

* [chat] fix ci

* [chat] skip checkpoint test
2023-04-26 16:32:40 +08:00
ddobokki df309fc6ab
[Chat] Remove duplicate functions (#3625) 2023-04-24 12:23:15 +08:00
digger-yu d7bf284706
[chat] polish code note typo (#3612) 2023-04-20 17:22:15 +08:00
Yuanchen 1ec0d386a9
reconstruct chat trainer and fix training script (#3588)
Co-authored-by: Yuanchen Xu <yuanchen.xu00@gmail.com>
2023-04-18 16:44:03 +08:00
Camille Zhong 36a519b49f Update test_ci.sh
update

Update test_ci.sh

Update test_ci.sh

Update test_ci.sh

Update test_ci.sh

Update test_ci.sh

Update test_ci.sh

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update test_ci.sh

Update test_ci.sh

update

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

update ci

Update test_ci.sh

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml

Update test_ci.sh

Update test_ci.sh

Update run_chatgpt_examples.yml

Update test_ci.sh

Update test_ci.sh

Update test_ci.sh

update test ci

RoBERTa for RLHF Stage 2 & 3 (still in testing)

Revert "Add RoBERTa for RLHF Stage 2 & 3 (test)"

This reverts commit 06741d894d.

Add RoBERTa for RLHF stage 2 & 3

1. add roberta folder under model folder
2. add  roberta option in train_reward_model.py
3. add some test in testci

Update test_ci.sh

Revert "Update test_ci.sh"

This reverts commit 9c7352b81766f3177d31eeec0ec178a301df966a.

Add RoBERTa for RLHF Stage 2 & 3 (test)

RoBERTa for RLHF Stage 2 & 3 (still in testing)

Revert "Add RoBERTa for RLHF Stage 2 & 3 (test)"

This reverts commit 06741d894d.

Add RoBERTa for RLHF stage 2 & 3

1. add roberta folder under model folder
2. add  roberta option in train_reward_model.py
3. add some test in testci

Update test_ci.sh

Revert "Update test_ci.sh"

This reverts commit 9c7352b81766f3177d31eeec0ec178a301df966a.

update roberta with coati

chat ci update

Revert "chat ci update"

This reverts commit 17ae7ae01fa752bd3289fc39069868fde99cf846.

[test]chat_update_ci

Update test_ci.sh

Update test_ci.sh

test

Update gpt_critic.py

Update gpt_critic.py

Update run_chatgpt_unit_tests.yml

update test ci

update

update

update

update

Update test_ci.sh

update

Update test_ci.sh

Update test_ci.sh

Update run_chatgpt_examples.yml

Update run_chatgpt_examples.yml
2023-04-18 14:33:12 +08:00
tingfeng cao 7788e0b0a5
fix: fix sft (#3568) 2023-04-17 16:47:44 +08:00
csric e355144375
[chatgpt] Detached PPO Training (#3195)
* run the base

* working on dist ppo

* sync

* detached trainer

* update detached trainer. no maker update function

* facing init problem

* 1 maker 1 trainer detached run. but no model update

* facing cuda problem

* fix save functions

* verified maker update

* nothing

* add ignore

* analyize loss issue

* remove some debug codes

* facing 2m1t stuck issue

* 2m1t verified

* do not use torchrun

* working on 2m2t

* working on 2m2t

* initialize strategy in ray actor env

* facing actor's init order issue

* facing ddp model update issue (need unwarp ddp)

* unwrap ddp actor

* checking 1m2t stuck problem

* nothing

* set timeout for trainer choosing. It solves the stuck problem!

* delete some debug output

* rename to sync with upstream

* rename to sync with upstream

* coati rename

* nothing

* I am going to detach the replaybuffer from trainer and make it a Ray Actor. Two benefits: 1. support TP trainer. 2. asynchronized buffer operations

* experience_maker_holder performs target-revolving _send_experience() instead of length comparison.

* move code to ray subfolder

* working on pipeline inference

* apply comments

---------

Co-authored-by: csric <richcsr256@gmail.com>
2023-04-17 14:46:50 +08:00
zhang-yi-chi e6a132a449
[chat]: add vf_coef argument for PPOTrainer (#3318) 2023-04-11 09:54:59 +08:00
gongenlei a7ca297281
[coati] Fix LlamaCritic (#3475)
* mv LlamaForCausalLM to LlamaModel

* rm unused imports

---------

Co-authored-by: gongenlei <gongenlei@baidu.com>
2023-04-07 11:39:09 +08:00
YY Lin 62f4e2eb07
[Chat]Add Peft support & fix the ptx bug (#3433)
* Update ppo.py

Fix the bug of fetching wrong batch data

* Add peft model support in SFT and Prompts training

In stage-1 and stage-3, the peft model supports are added. So the trained artifacts will be only a small lora additions instead of the whole bunch of files.

* Delete test_prompts.txt

* Delete test_pretrained.txt

* Move the peft stuffs to a community folder.

* Move the demo sft to community

* delete dirty files

* Add instructions to install peft using source

* Remove Chinese comments

* remove the Chinese comments
2023-04-06 11:54:52 +08:00
Dr-Corgi 73afb63594
[chat]fix save_model(#3377)
The function save_model should be a part of PPOTrainer.
2023-04-06 11:19:14 +08:00
Camille Zhong 72cb4dd433
[Chat] fix the tokenizer "int too big to convert" error in SFT training (#3453)
* Add RoBERTa for RLHF Stage 2 & 3 (test)

RoBERTa for RLHF Stage 2 & 3 (still in testing)

* Revert "Add RoBERTa for RLHF Stage 2 & 3 (test)"

This reverts commit 06741d894d.

* Add RoBERTa for RLHF stage 2 & 3

1. add roberta folder under model folder
2. add  roberta option in train_reward_model.py
3. add some test in testci

* Update test_ci.sh

* Revert "Update test_ci.sh"

This reverts commit 9c7352b81766f3177d31eeec0ec178a301df966a.

* Add RoBERTa for RLHF Stage 2 & 3 (test)

RoBERTa for RLHF Stage 2 & 3 (still in testing)

* Revert "Add RoBERTa for RLHF Stage 2 & 3 (test)"

This reverts commit 06741d894d.

* Add RoBERTa for RLHF stage 2 & 3

1. add roberta folder under model folder
2. add  roberta option in train_reward_model.py
3. add some test in testci

* Update test_ci.sh

* Revert "Update test_ci.sh"

This reverts commit 9c7352b81766f3177d31eeec0ec178a301df966a.

* update roberta with coati

* chat ci update

* Revert "chat ci update"

This reverts commit 17ae7ae01fa752bd3289fc39069868fde99cf846.

* [Chat] fix the tokenizer "int too big to convert" error in SFT training

fix the tokenizer error during SFT training using Bloom and OPT
2023-04-06 09:30:28 +08:00
Yuanchen b92313903f
fix save_model indent error in ppo trainer (#3450)
Co-authored-by: Yuanchen Xu <yuanchen.xu00@gmail.com>
2023-04-05 09:45:42 +08:00
Yuanchen 773955abfa
fix save_model inin naive and ddp strategy (#3436)
Co-authored-by: Yuanchen Xu <yuanchen.xu00@gmail.com>
2023-04-04 15:30:01 +08:00
ver217 26b7aac0be
[zero] reorganize zero/gemini folder structure (#3424)
* [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
2023-04-04 13:48:16 +08:00
Yuanchen b09adff724
[chat]fix sft training for bloom, gpt and opt (#3418)
fix sft training for bloom, gpt and opt
2023-04-04 09:46:23 +08:00
Camille Zhong 30412866e0
[chatgpt] add pre-trained model RoBERTa for RLHF stage 2 & 3 (#3223)
* Add RoBERTa for RLHF Stage 2 & 3 (test)

RoBERTa for RLHF Stage 2 & 3 (still in testing)

* Revert "Add RoBERTa for RLHF Stage 2 & 3 (test)"

This reverts commit 06741d894d.

* Add RoBERTa for RLHF stage 2 & 3

1. add roberta folder under model folder
2. add  roberta option in train_reward_model.py
3. add some test in testci

* add test for reward model training

* Update test_ci.sh

* Revert "Update test_ci.sh"

This reverts commit 9c7352b81766f3177d31eeec0ec178a301df966a.

* Add RoBERTa for RLHF Stage 2 & 3 (test)

RoBERTa for RLHF Stage 2 & 3 (still in testing)

* Revert "Add RoBERTa for RLHF Stage 2 & 3 (test)"

This reverts commit 06741d894d.

* Add RoBERTa for RLHF stage 2 & 3

1. add roberta folder under model folder
2. add  roberta option in train_reward_model.py
3. add some test in testci

* Update test_ci.sh

* Revert "Update test_ci.sh"

This reverts commit 9c7352b81766f3177d31eeec0ec178a301df966a.

* update roberta with coati
2023-04-03 10:11:03 +08:00
Fazzie-Maqianli b0ce5a1032
[Coati] first commit (#3283) 2023-03-28 20:25:36 +08:00