Commit Graph

  • eef96e0877 polish code for gptq (#4793) littsk 2023-09-25 14:05:48 +08:00
  • cb3a25a062 [checkpointio] hotfix torch 2.0 compatibility (#4824) Hongxin Liu 2023-10-07 10:45:52 +08:00
  • ad23460cf8 Merge pull request #4856 from KKZ20/test/model_support_for_low_level_zero ppt0011 2023-10-06 09:32:33 +08:00
  • 81ee91f2ca Merge pull request #4858 from Shawlleyw/main ppt0011 2023-10-06 09:27:54 +08:00
  • c97a3523db fix: typo in comment of low_level_zero plugin shaoyuw 2023-10-05 16:30:34 +00:00
  • db40e086c8 [test] modify model supporting part of low_level_zero plugin (including correspoding docs) Zhongkai Zhao 2023-10-05 15:10:31 +08:00
  • d1fcc0fa4d [infer] fix test bug (#4838) Xu Kai 2023-10-04 10:01:03 +08:00
  • 013a4bedf0 [inference]fix import bug and delete down useless init (#4830) Jianghai 2023-10-04 09:18:45 +08:00
  • 573f270537 [Infer] Serving example w/ ray-serve (multiple GPU case) (#4841) Yuanheng Zhao 2023-10-02 17:48:38 +08:00
  • 3a74eb4b3a [Infer] Colossal-Inference serving example w/ TorchServe (single GPU case) (#4771) Yuanheng Zhao 2023-10-02 17:42:37 +08:00
  • ed888183ec [fix]downgrade colossalai (#4840) cloud/coati flybird11111 2023-09-28 17:56:16 +08:00
  • ed06731e00 update Colossal (#4832) Tong Li 2023-09-28 16:05:05 +08:00
  • c3bef20478 add autotune (#4822) Xu Kai 2023-09-28 13:47:35 +08:00
  • 822051d888 [doc] update slack link (#4823) binmakeswell 2023-09-27 17:37:39 +08:00
  • 1fa8c5e09f Update Qwen-7B results (#4821) Yuanchen 2023-09-27 17:33:54 +08:00
  • be400a0936 [chat] fix gemini strategy (#4698) flybird11111 2023-09-27 13:15:32 +08:00
  • bbbcac26e8 fix format (#4815) Tong Li 2023-09-27 12:50:22 +08:00
  • fb46d05cdf [format] applied code formatting on changed files in pull request 4595 (#4602) github-actions[bot] 2023-09-27 10:45:03 +08:00
  • 11f1e426fe [hotfix] Correct several erroneous code comments (#4794) littsk 2023-09-27 10:43:03 +08:00
  • 54b3ad8924 [hotfix] fix norm type error in zero optimizer (#4795) littsk 2023-09-27 10:35:24 +08:00
  • da15fdb9ca [doc] add lazy init docs (#4808) Hongxin Liu 2023-09-27 10:24:04 +08:00
  • a22706337a [misc] add last_epoch in CosineAnnealingWarmupLR (#4778) Yan haixu 2023-09-26 14:43:46 +08:00
  • b6cf0aca55 [hotfix] change llama2 Colossal-LLaMA-2 script filename (#4800) Chandler-Bing 2023-09-26 11:44:27 +08:00
  • 62b6af1025 Merge pull request #4805 from TongLi3701/docs/fix Desperado-Jia 2023-09-26 11:39:35 +08:00
  • 8cbce6184d update Tong Li 2023-09-26 11:36:53 +08:00
  • 4965c0dabd [lazy] support from_pretrained (#4801) Hongxin Liu 2023-09-26 11:04:11 +08:00
  • bd014673b0 update readme Tong Li 2023-09-26 10:58:05 +08:00
  • 64a08b2dc3 [checkpointio] support unsharded checkpointIO for hybrid parallel (#4774) Baizhou Zhang 2023-09-26 10:58:03 +08:00
  • a2db75546d [doc] polish shardformer doc (#4779) Baizhou Zhang 2023-09-26 10:57:47 +08:00
  • 26cd6d850c [fix] fix weekly runing example (#4787) flybird11111 2023-09-25 16:19:33 +08:00
  • d512a4d38d [doc] add llama2 domain-specific solution news (#4789) binmakeswell 2023-09-25 10:44:15 +08:00
  • ce777853ae [feature] ColossalEval: Evaluation Pipeline for LLMs (#4786) Yuanchen 2023-09-24 23:14:11 +08:00
  • 74aa7d964a initial commit: add colossal llama 2 (#4784) Tong Li 2023-09-24 23:12:26 +08:00
  • 4146f1c0ce [release] update version (#4775) v0.3.3 Hongxin Liu 2023-09-22 18:29:17 +08:00
  • ce7ade3882 [inference] chatglm2 infer demo (#4724) Jianghai 2023-09-22 11:12:50 +08:00
  • 946ab56c48 [feature] add gptq for inference (#4754) Xu Kai 2023-09-22 11:02:50 +08:00
  • 1e0e080837 [bug] Fix the version check bug in colossalai run when generating the cmd. (#4713) littsk 2023-09-22 10:50:47 +08:00
  • 3e05c07bb8 [lazy] support torch 2.0 (#4763) Hongxin Liu 2023-09-21 16:30:23 +08:00
  • 901ab1eedd [chat]: add lora merge weights config (#4766) Wenhao Chen 2023-09-21 16:23:59 +08:00
  • 493a5efeab [doc] add shardformer doc to sidebar (#4768) Baizhou Zhang 2023-09-21 14:53:16 +08:00
  • 66f3926019 [doc] clean up outdated docs (#4765) Hongxin Liu 2023-09-21 11:36:20 +08:00
  • df66741f77 [bug] fix get_default_parser in examples (#4764) Baizhou Zhang 2023-09-21 10:42:25 +08:00
  • c0a033700c [shardformer] fix master param sync for hybrid plugin/rewrite unwrapping logic (#4758) Baizhou Zhang 2023-09-20 18:29:37 +08:00
  • 7b9b86441f [chat]: update rm, add wandb and fix bugs (#4471) Wenhao Chen 2023-09-20 15:53:58 +08:00
  • 07c2e3d09c Merge pull request #4757 from ppt0011/main ppt0011 2023-09-20 11:57:43 +08:00
  • 4d7537ba25 [doc] put native colossalai plugins first in description section Pengtai Xu 2023-09-20 09:24:10 +08:00
  • e10d9f087e [doc] add model examples for each plugin Pengtai Xu 2023-09-19 18:01:23 +08:00
  • a04337bfc3 [doc] put individual plugin explanation in front Pengtai Xu 2023-09-19 16:27:37 +08:00
  • 10513f203c [doc] explain suitable use case for each plugin Pengtai Xu 2023-09-19 15:28:01 +08:00
  • 079bf3cb26 [misc] update pre-commit and run all files (#4752) Hongxin Liu 2023-09-19 14:20:26 +08:00
  • 3c6b831c26 [format] applied code formatting on changed files in pull request 4743 (#4750) github-actions[bot] 2023-09-18 16:52:42 +08:00
  • b5f9e37c70 [legacy] clean up legacy code (#4743) Hongxin Liu 2023-09-18 16:31:06 +08:00
  • 32e7f99416 [kernel] update triton init #4740 (#4740) Xuanlei Zhao 2023-09-18 09:44:27 +08:00
  • d151dcab74 [doc] explaination of loading large pretrained models (#4741) Baizhou Zhang 2023-09-15 21:04:07 +08:00
  • 4c4482f3ad [example] llama2 add fine-tune example (#4673) flybird11111 2023-09-15 18:45:44 +08:00
  • ac2797996b [shardformer] add custom policy in hybrid parallel plugin (#4718) Xuanlei Zhao 2023-09-15 17:53:13 +08:00
  • 451c3465fb [doc] polish shardformer doc (#4735) Baizhou Zhang 2023-09-15 17:39:10 +08:00
  • 73eb3e8862 Merge pull request #4738 from ppt0011/main ppt0011 2023-09-15 17:34:42 +08:00
  • 608cffaed3 [example] add gpt2 HybridParallelPlugin example (#4653) Bin Jia 2023-09-15 17:12:46 +08:00
  • 6a03c933a0 [shardformer] update seq parallel document (#4730) Bin Jia 2023-09-15 16:09:32 +08:00
  • cd4e61d149 [legacy] remove deterministic data loader test Pengtai Xu 2023-09-15 15:52:18 +08:00
  • 46162632e5 [shardformer] update pipeline parallel document (#4725) flybird11111 2023-09-15 14:32:04 +08:00
  • e4fc57c3de Optimized some syntax errors in the documentation and code under applications/ (#4127) digger yu 2023-09-15 14:18:22 +08:00
  • 50e5602c2d [doc] add shardformer support matrix/update tensor parallel documents (#4728) Baizhou Zhang 2023-09-15 13:52:30 +08:00
  • 8c2dda7410 [format] applied code formatting on changed files in pull request 4726 (#4727) github-actions[bot] 2023-09-15 13:17:32 +08:00
  • f911d5b09d [doc] Add user document for Shardformer (#4702) Baizhou Zhang 2023-09-15 10:56:39 +08:00
  • ce97790ed7 [doc] fix llama2 code link (#4726) binmakeswell 2023-09-14 23:19:25 +08:00
  • 20190b49a5 [shardformer] to fix whisper test failed due to significant accuracy differences. (#4710) flybird11111 2023-09-14 21:34:20 +08:00
  • e2c0e7f92a [hotfix] Fix import error: colossal.kernel without triton installed (#4722) Yuanheng Zhao 2023-09-14 18:03:55 +08:00
  • c7d6975d29 [shardformer] fix GPT2DoubleHeadsModel (#4703) flybird11111 2023-09-13 15:57:16 +08:00
  • 068372a738 [doc] add potential solution for OOM in llama2 example (#4699) Baizhou Zhang 2023-09-13 10:43:30 +08:00
  • 9c2feb2f0b fix some typo with colossalai/device colossalai/tensor/ etc. (#4171) digger yu 2023-09-12 17:41:52 +08:00
  • d8ceeac14e [hotfix] fix typo in hybrid parallel io (#4697) Baizhou Zhang 2023-09-12 17:32:19 +08:00
  • 8844691f4b [shardformer] update shardformer readme (#4689) flybird11111 2023-09-12 15:14:24 +08:00
  • 1d454733c4 [doc] Update booster user documents. (#4669) Baizhou Zhang 2023-09-12 10:47:23 +08:00
  • bce0f16702 [Feature] The first PR to Add TP inference engine, kv-cache manager and related kernels for our inference system (#4577) Cuiqing Li 2023-09-12 01:22:56 +08:00
  • eedaa3e1ef [shardformer]fix gpt2 double head (#4663) flybird11111 2023-09-11 18:35:03 +08:00
  • 554aa9592e [legacy] move communication and nn to legacy and refactor logger (#4671) Hongxin Liu 2023-09-11 16:24:28 +08:00
  • 536397cc95 [devops] fix concurrency group (#4667) Hongxin Liu 2023-09-11 15:32:50 +08:00
  • 7486ed7d3a [shardformer] update llama2/opt finetune example and fix llama2 policy (#4645) flybird11111 2023-09-09 22:45:36 +08:00
  • a686f9ddc8 [devops] fix concurrency group and compatibility test (#4665) Hongxin Liu 2023-09-08 13:49:40 +08:00
  • 295b38fecf [example] update vit example for hybrid parallel plugin (#4641) Baizhou Zhang 2023-09-07 17:38:45 +08:00
  • 660eed9124 [pipeline] set optimizer to optional in execute_pipeline (#4630) Baizhou Zhang 2023-09-07 10:42:59 +08:00
  • c3d5fa3bac [shardformer] Support customized policy for llamav2 based model with HybridParallelPlugin (#4624) eric8607242 2023-09-07 10:15:13 +08:00
  • 9709b8f502 [release] update version (#4623) v0.3.2 Hongxin Liu 2023-09-06 23:41:04 +08:00
  • efba0f44b9 Merge pull request #4612 from hpcaitech/feature/shardformer Hongxin Liu 2023-09-05 23:20:00 +08:00
  • fae6c92ead Merge branch 'main' into feature/shardformer Hongxin Liu 2023-09-05 21:54:08 +08:00
  • ac178ca5c1 [legacy] move builder and registry to legacy (#4603) Hongxin Liu 2023-09-04 19:56:42 +08:00
  • 8accecd55b [legacy] move engine to legacy (#4560) Hongxin Liu 2023-09-04 11:33:40 +08:00
  • 89fe027787 [legacy] move trainer to legacy (#4545) Hongxin Liu 2023-08-31 13:51:28 +08:00
  • bd18678478 [test] fix gemini checkpoint and gpt test (#4620) Hongxin Liu 2023-09-05 16:02:23 +08:00
  • 807e01a4ba [zero] hotfix master param sync (#4618) Hongxin Liu 2023-09-05 15:04:02 +08:00
  • e71d245293 [test] ignore gpt2 shardformer test (#4619) Hongxin Liu 2023-09-05 14:21:31 +08:00
  • ec0866804c [shardformer] update shardformer readme (#4617) flybird11111 2023-09-05 13:14:41 +08:00
  • 86d22581e4 [shardformer] Add overlap optional for HybridParallelPlugin (#4615) Bin Jia 2023-09-05 11:52:23 +08:00
  • a39a5c66fe Merge branch 'main' into feature/shardformer Hongxin Liu 2023-09-04 23:43:13 +08:00
  • e79b1e80e2 [checkpointio] support huggingface from_pretrained for all plugins (#4606) Baizhou Zhang 2023-09-04 23:25:01 +08:00
  • 0a94fcd351 [shardformer] update bert finetune example with HybridParallelPlugin (#4584) flybird11111 2023-09-04 21:46:29 +08:00
  • 24c0768795 [shardformer] Pytree fix (#4533) Jianghai 2023-09-04 17:52:23 +08:00
  • aaeb520ce3 Merge pull request #4542 from hpcaitech/chatglm yingliu-hpc 2023-09-04 16:09:45 +08:00