Commit Graph

  • eb2f8b1f6b
    add how to build tfrecord dataset (#48) ver217 2021-12-02 16:31:23 +0800
  • 32cb13281b add how to build tfrecord dataset ver217 2021-12-02 00:35:47 +0800
  • 81220f2374
    Merge branch 'hotfix/vit-example' into hotfix/vit-example ver217 2021-12-02 16:19:13 +0800
  • f26618d3a8 add how to build tfrecord dataset in readme ver217 2021-12-02 16:17:53 +0800
  • 4da256a584
    add some details in vit-b16 example (#46) ver217 2021-12-02 09:29:27 +0800
  • 48bce611fb add some details in vit-b16 example ver217 2021-12-02 00:35:47 +0800
  • 622709788f add some details in vit-b16 example ver217 2021-12-02 00:35:47 +0800
  • e67dab92a9
    add some details in vit-b16 example (#43) (#44) ver217 2021-12-02 08:55:11 +0800
  • 6273d4d190
    add some details in vit-b16 example (#43) ver217 2021-12-02 00:37:51 +0800
  • 63421f8c86 add some details in vit-b16 example ver217 2021-12-02 00:35:47 +0800
  • 610e51d8ea change seed ver217 2021-12-01 18:53:28 +0800
  • e907f1bf05 add log for ddp ver217 2021-12-01 18:27:05 +0800
  • a97564d167 fix loss accumulation ver217 2021-12-01 14:27:23 +0800
  • 216a9b4c74 add example of self-supervised SimCLR training xinzhang 2021-12-01 01:54:29 +0000
  • 4981f8f686 support torch ddp ver217 2021-11-30 23:48:54 +0800
  • 2c45cf09fa
    Merge branch 'develop/experiments' into feature/pipeline ver217 2021-11-30 17:27:05 +0800
  • 536ce12772 optimize communication of pipeline parallel ver217 2021-11-26 19:06:05 +0800
  • 94a7b41d0c
    Merge branch 'hpcaitech:develop/experiments' into develop/experiments Frank Lee 2021-11-30 11:19:38 +0800
  • dd2b1586dc Update 2.5d code BoxiangW 2021-11-29 16:33:07 +0800
  • d0cb7bb4aa
    Integrate 1d tensor parallel in Colossal-AI (#39) puck_WCR 2021-11-29 14:00:14 +0800
  • 1251d1b57b Integrate 1d tensor parallel in Colossal-AI WANG-CR 2021-11-29 13:30:49 +0800
  • 2528adc62f
    add explanation for ViT example (#35) (#36) binmakeswell 2021-11-29 10:25:38 +0800
  • 8fb88b8b06 optimized 2D operations FrankLeeeee 2021-11-29 03:23:36 +0100
  • 55a85ed924 Integrated 1d tensor parallel in Colossal-AI: Updated 1d basic layers Added layers for transformers/ViT Added 1D model zoo of ViT model WANG-CR 2021-11-08 19:36:42 +0800
  • dba3a8e0c5 added Chinese documents and fixed some typos in English documents Fan Cui 2021-11-02 23:01:13 +0800
  • c4db755df2
    add explanation for ViT example (#35) binmakeswell 2021-11-26 12:06:14 +0800
  • 04ca4484af add explanation for ViT example binmakeswell 2021-11-26 12:03:26 +0800
  • 6d770a98e1 added ignore steps argument for timer MeluXina user 2021-11-25 07:07:27 +0100
  • 3eeb13e9c3 added ignore steps argument for timer MeluXina user 2021-11-25 07:07:27 +0100
  • 5e8c9b6362
    Merge pull request #30 from 1SAA/develop/experiments アマデウス 2021-11-24 17:52:52 +0800
  • e0ae532c07 Split conv2d, class token, positional embedding in 2d, Fix random number in ddp Fix convergence in cifar10, Imagenet1000 1SAA 2021-11-18 17:54:19 +0800
  • e5d284a04e
    Merge pull request #2 from hpcaitech/main binmakeswell 2021-11-23 14:23:25 +0800
  • dbe62c67b8
    add an example of ViT-B/16 and remove w_norm clipping in LAMB (#29) ver217 2021-11-18 23:45:09 +0800
  • 6baa21d0d6 add an example of ViT-B/16 and remove w_norm clipping in LAMB ver217 2021-11-18 20:10:50 +0800
  • a0f9dde182 add an example of ViT-B/16 and remove w_norm clipping in LAMB ver217 2021-11-18 20:10:50 +0800
  • 3defa32aee
    Support TP-compatible Torch AMP and Update trainer API (#27) Frank Lee 2021-11-18 19:45:06 +0800
  • 4ec1ca0170
    improved consistency between trainer, engine and schedule (#23) Frank Lee 2021-11-15 17:19:41 +0800
  • 2b05de4c64
    use env to control the language of doc (#24) (#25) ver217 2021-11-15 16:53:56 +0800
  • eb20ef4542 improved consistency between trainer, engine and schedule 1SAA 2021-11-11 17:09:46 +0800
  • fa11657b2c
    use env to control the language of doc (#24) ver217 2021-11-15 16:49:06 +0800
  • 9a75814e6e use env to control the language of doc ver217 2021-11-15 16:44:53 +0800
  • 9942fd5bfa
    remove redundancy func in setup (#19) (#20) ver217 2021-11-15 16:43:28 +0800
  • 2457bc9747 Revert "fixed trainer" Frank Lee 2021-11-11 18:07:46 +0800
  • 2e0b0b7699 fixed trainer Frank Lee 2021-11-11 17:41:45 +0800
  • f58b744ab7
    Merge branch 'hotfix/trainer' into hotfix/trainer HELSON 2021-11-11 17:25:57 +0800
  • c84902c980 Changed API in Schedule, Engine 1SAA 2021-11-11 17:09:46 +0800
  • 008a1b4c9e
    remove redundancy func in setup (#19) ver217 2021-11-11 13:54:30 +0800
  • 10a7a17a10 remove redundancy func in setup ver217 2021-11-11 13:51:52 +0800
  • af88570f4b fixed bugs in compatibility between torch amp and tensor parallel and performed some minor fixes Frank Lee 2021-11-10 10:47:58 +0800
  • c8cb9f9e34
    fix FP16 optimizer and adapted torch amp with tensor parallel (#18) ver217 2021-11-08 16:47:32 +0800
  • 08f25d385b
    Merge branch 'develop/experiments' into hotfix/fp16 ver217 2021-11-08 16:46:51 +0800
  • fa1e21b697 fix FP16 optimizer and adapted torch amp with tensor parallel ver217 2021-11-03 13:27:29 +0800
  • 26dce0a898
    Merge pull request #17 from 1SAA/main Frank Lee 2021-11-08 16:39:27 +0800
  • 8aa21d6bc5 Add gradient accumulation, fix lr scheduler 1SAA 2021-11-08 15:48:27 +0800
  • 0aa07e600c
    Merge pull request #15 from hpcaitech/feature/zhdoc ver217 2021-11-04 14:26:28 +0800
  • 05e7069a5b fixed some typos in the documents, added blog link and paper author information in README binmakeswell 2021-11-03 16:07:28 +0800
  • 5476081a06
    Merge pull request #1 from hpcaitech/feature/zhdoc Neo 2021-11-03 16:28:32 +0800
  • 440ecc61d3
    Merge branch 'feature/zhdoc' into feature/zhdoc Neo 2021-11-03 16:27:55 +0800
  • 01b4c8d83e
    Merge pull request #13 from binmakeswell/feature/zhdoc binmakeswell 2021-11-03 16:09:13 +0800
  • 0c402a9127 fixed typos in documents, add blog in README binmakeswell 2021-11-03 16:07:28 +0800
  • 937f9c3b90 fixed some typos in English documents Fan Cui 2021-11-03 16:03:02 +0800
  • c24becbdd8 fix: gradient clipping of FP16 optimizer ver217 2021-11-03 13:27:29 +0800
  • 5f506ab0e5 fix: gradient clipping of FP16 optimizer ver217 2021-11-03 13:27:29 +0800
  • e07165ead5
    Merge pull request #1 from hpcaitech/feature/zhdoc binmakeswell 2021-11-03 13:21:30 +0800
  • ccb44882e1
    Merge pull request #10 from hpcaitech/feature/zhdoc Frank Lee 2021-11-03 11:38:06 +0800
  • 18ba66e012 added Chinese documents and fixed some typos in English documents Fan Cui 2021-11-02 23:01:13 +0800
  • c1b10e26ae added Chinese documents and fixed some typos in English documents Fan Cui 2021-11-02 23:01:13 +0800
  • 4a3767216a added Chinese documents and fixed some typos and inconsistencies in English documents Fan Cui 2021-11-02 23:01:13 +0800
  • 906640246f added Chinese documents and fixed some typos in English documents Fan Cui 2021-11-01 15:22:11 +0800
  • b667f0b64f added Chinese documents and fixed some typos in English documents Fan Cui 2021-11-02 16:34:03 +0800
  • 115e5b5a50
    Merge branch 'hpcaitech:hotfix/doc' into hotfix/doc Neo 2021-11-02 18:01:47 +0800
  • ccbc918c11
    Merge pull request #4 from hpcaitech/hotfix/doc Frank Lee 2021-11-02 14:35:06 +0800
  • b7e3812601 fixed the inconsistency between the document and the current version of code Fan Cui 2021-11-01 15:22:11 +0800
  • 50982c0b7d reoder parallelization methods in parallelization documentation ver217 2021-11-01 14:31:55 +0800
  • 3c7604ba30 update documentation ver217 2021-10-29 09:29:20 +0800
  • 3245a69fc2
    cleaned test scripts v0.0.1-beta アマデウス 2021-10-29 00:48:14 +0800
  • da2042f5c1
    updated readme アマデウス 2021-10-29 00:39:21 +0800
  • 404ecbdcc6 Migrated project zbian 2021-10-28 18:21:23 +0200
  • 2ebaefc542
    Initial commit アマデウス 2021-10-29 00:19:45 +0800