103 Commits (39f2582e987871c198f2f2526cd4435cbd569741)

Author SHA1 Message Date
littsk 83b52c56cd
[feature] Add clip_grad_norm for hybrid_parallel_plugin (#4837) 1 year ago
littsk ffd9a3cbc9
[hotfix] fix bug in sequence parallel test (#4887) 1 year ago
Hongxin Liu 079bf3cb26
[misc] update pre-commit and run all files (#4752) 1 year ago
flybird11111 eedaa3e1ef
[shardformer]fix gpt2 double head (#4663) 1 year ago
flybird11111 7486ed7d3a
[shardformer] update llama2/opt finetune example and fix llama2 policy (#4645) 1 year ago
Hongxin Liu bd18678478
[test] fix gemini checkpoint and gpt test (#4620) 1 year ago
Hongxin Liu e71d245293
[test] ignore gpt2 shardformer test (#4619) 1 year ago
Jianghai 24c0768795
[shardformer] Pytree fix (#4533) 1 year ago
Baizhou Zhang 2c787d7f47
[shardformer] fix submodule replacement bug when enabling pp (#4544) 1 year ago
flybird11111 ec18fc7340
[shardformer] support pp+tp+zero1 tests (#4531) 1 year ago
flybird11111 d367b88785
[shardformer] fix opt test hanging (#4521) 1 year ago
Bin Jia e241b74f24
[shardformer] Add overlap support for gpt2 (#4535) 1 year ago
Baizhou Zhang 0387a47e63
[shardformer] fix emerged bugs after updating transformers (#4526) 1 year ago
Bin Jia c554b7f559
[shardformer/fix overlap bug] fix overlap bug, add overlap as an option in shardco… (#4516) 1 year ago
Jianghai 376533a564
[shardformer] zero1+pp and the corresponding tests (#4517) 1 year ago
flybird11111 de8a65babc
[shardformer] opt fix. (#4514) 1 year ago
flybird11111 3353e55c80
[shardformer] vit/llama/t5 ignore the sequence parallelism flag and some fix. (#4498) 1 year ago
Hongxin Liu 27061426f7
[gemini] improve compatibility and add static placement policy (#4479) 1 year ago
Jianghai e04436a82a
[shardformer] tests for 3d parallel (#4493) 1 year ago
Jianghai 5545114fd8
rename chatglm to chatglm2 (#4484) 1 year ago
Baizhou Zhang 1c7df566e2
[shardformer] support tp+zero for shardformer (#4472) 1 year ago
Jianghai 8739aa7fa0
[shardformer] Pipeline/whisper (#4456) 1 year ago
Bin Jia 7c8be77081
[shardformer/sequence parallel] support gpt2 seq parallel with pp/dp/tp (#4460) 1 year ago
Baizhou Zhang 6ef33f75aa
[shardformer] support DDP in HybridPlugin/add tp+dp tests (#4446) 1 year ago
Bin Jia 424629fea0
[shardformer/sequence parallel] Cherry pick commit to new branch (#4450) 1 year ago
Hongxin Liu 172f7fa3cf [misc] resolve code factor issues (#4433) 1 year ago
flybird11111 328a791d10 [shardformer] update bloom/llama/vit/chatglm tests (#4420) 1 year ago
flybird11111 108e54a0b4 [shardformer]update t5 tests for using all optimizations. (#4407) 1 year ago
flybird11111 1edc9b5fb3 [shardformer] update tests for all optimization (#4413) 1 year ago
Baizhou Zhang 7711bd524a [shardformer] rewrite tests for opt/bloom/llama/vit/chatglm (#4395) 1 year ago
flybird11111 21e0a42fd1 [shardformer]fix, test gpt2 for AMP+TP (#4403) 1 year ago
Jianghai 7596e9ae08 [pipeline] rewrite bert tests and fix some bugs (#4409) 1 year ago
flybird1111 d2cd48e0be [shardformer] test all optimizations (#4399) 1 year ago
Baizhou Zhang ed4c448488 [pipeline] rewrite t5 tests & support multi-tensor transmitting in pipeline (#4388) 1 year ago
flybird1111 906426cb44 [Shardformer] Merge flash attention branch to pipeline branch (#4362) 1 year ago
Jianghai a88e92251d [pipeline] add chatglm (#4363) 1 year ago
Baizhou Zhang b1feeced8e [shardformer] add util functions for shardformer tests/fix sync_shared_param (#4366) 1 year ago
Bin Jia 5c6f183192 [test] Hotfix/fix some model test and refactor check util api (#4369) 1 year ago
FoolPlayer 726541afe2 update some module with new api version 1 year ago
FoolPlayer 879301d0da [shardformer] support Blip2 (#4243) 1 year ago
klhhhhh 8120eca0c0 [shardformer] support ChatGLMForConditionalGeneration & add fusedlayernorm for vit 1 year ago
klhhhhh 1a29e8fc29 [shardformer] polish chatglm code 1 year ago
klhhhhh 8620009dd7 [sharformer] add first version of policy of chatglm 1 year ago
klhhhhh c49286985d [shardformer] vit test finish and support 1 year ago
klhhhhh f60162b265 [shardformer] added tests 1 year ago
Kun Lin ed34bb1310 Feature/chatglm (#4240) 1 year ago
FoolPlayer 9ee4ebea83 [shardformer] support whisper (#4212) 1 year ago
FoolPlayer dd2bf02679 [shardformer] support SAM (#4231) 1 year ago
Baizhou Zhang 0ceec8f9a9 [pipeline] support fp32 for HybridPlugin/merge shardformer test and pipeline test into one file (#4354) 1 year ago
Jianghai f13954cd58 [pipeline] refactor test pipeline and remove useless utils in pipeline (#4324) 1 year ago