wangwei
/
MegEngine
Not watched
1
0
Fork 0
Code
Releases 31 Wiki Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain

Commit Graph

  • *
  • *
  • *
  • *
  • |\
  • | *
  • * |
  • |\ \
  • | * |
  • |/ /
  • * |
  • * |
  • * |
  • * |
  • |\ \
  • * | |
  • * | |
  • * | |
  • * | |
  • |\ \ \
  • | | | | *
  • | | | | |\
  • | | | | | *
  • | | | | |/
  • | | | | | *
  • | | | | |/
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • |\ \ \ \ \
  • * \ \ \ \ \
  • |\ \ \ \ \ \
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • | | | * | | |
  • | | | | |_|/
  • | | | |/| |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • |\ \ \ \ \ \
  • * | | | | | |
  • | | | | | | | *
  • | | | | | | | *
  • | | | | | | | *
  • * | | | | | | |
  • * | | | | | | |
  • * | | | | | | |
  • | |_|_|/ / / /
  • |/| | | | | |
  • | | | | | | *
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • * | | | | | |
  • | | | | | | *
  • | |_|_|_|_|/
  • |/| | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • * | | | | |
  • | |_|_|_|/
  • |/| | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • * | | | |
  • 411253f8 feat(mge): implement warp_affine backward by Megvii Engine Team 2022-08-11 18:46:43 +0800
  • 496070cf fix(mge/quant): call is_distributed func correctly by Megvii Engine Team 2022-09-14 15:42:20 +0800
  • f0291883 fix(mgb): make error infomation of group conv input channel mismatch more readable by Megvii Engine Team 2022-09-09 16:21:30 +0800
  • 31218a18 Merge pull request #476 from MegEngine/HuaHua404-patch-4 by kagome1007 2022-09-26 18:29:49 +0800
  • bebaab3a (HuaHua404-patch-4) docs(readme): update key features description by HuaHua404 2022-08-12 16:55:06 +0800
  • 7cd9b168 Merge pull request #487 from kagome1007/master by megvii-mge 2022-09-26 18:11:15 +0800
  • e73f4c73 fix(mge): update readme by wenjuan 2022-09-26 15:48:16 +0800
  • 6a0d797a feat(ci): update submodules by Megvii Engine Team 2022-09-17 10:39:43 +0800
  • d9770792 feat(third_party): update cpuinfo by Megvii Engine Team 2022-09-13 15:22:51 +0800
  • d27a4456 fix(mge): update format by Megvii Engine Team 2022-09-13 16:10:55 +0800
  • f50db0b8 Merge pull request #464 from fanhqme2:patch-1 by Megvii Engine Team 2022-09-14 16:43:29 +0800
  • d205e7be docs(mge/functional): update copy example by Megvii Engine Team 2022-09-11 11:08:02 +0800
  • 6e48e593 docs(mge/functional): update copy example by Megvii Engine Team 2022-09-09 10:32:26 +0800
  • e89ee631 docs(mge/functional): update copy example by Megvii Engine Team 2022-09-08 17:50:18 +0800
  • 7655d99a Merge pull request #478 from thunderstudying:docstring-copy by Megvii Engine Team 2022-09-14 16:33:30 +0800
  • f045d3fb (dev-support-lite-fork-debug-mode) Merge pull request #486 from lyq998/support-lite-fork-debug-mode by haolongzhangm 2022-09-14 14:05:30 +0800
  • 37d81eae feat(lite): support base lite ipc fork debug by liuyuqiao 2022-09-09 20:43:18 +0800
  • fb2329e9 feat(dnn) add nchw44 deconv by wangxiang 2022-08-29 09:59:38 +0800
  • a5dea703 build(mge): add boost (part 1) by Megvii Engine Team 2022-09-09 14:01:06 +0800
  • 1dd27b39 docs(mge): update functional.nn.max_pool2d docstring by Megvii Engine Team 2022-09-02 11:53:17 +0800
  • 6e54a3bf docs(mge): update MaxPool2d & functional.nn.max_pool2d docstring by Megvii Engine Team 2022-08-30 15:49:29 +0800
  • 923dc2d5 Merge pull request #441 from Seeker98:maxpool_doc by Megvii Engine Team 2022-09-13 10:23:26 +0800
  • a8b146d7 Merge pull request #397 from kagome112:master by Megvii Engine Team 2022-09-11 11:03:32 +0800
  • 5655636c chore(mge/metric): funtional.topk_acc (add nn) use functional.metric by Megvii Engine Team 2022-09-02 15:59:42 +0800
  • d5688c3d chore(mge/functional): deprecate debug_param and prefer config by Megvii Engine Team 2022-09-01 13:21:43 +0800
  • 9bb2e395 chore(mge/tensorcache): rename the module making it not public by Megvii Engine Team 2022-09-01 14:06:55 +0800
  • 6e56d1bd chore(mge/deprecated): add __deprecated__ attr for condition case by Megvii Engine Team 2022-09-01 13:45:26 +0800
  • 98460f58 chore(mge/misc): rename APIs that should not be public by Megvii Engine Team 2022-09-01 17:30:01 +0800
  • 217999b1 feat(arm): add winograd F43 NCHW44 algo and winograd F43 44 algo by Megvii Engine Team 2022-08-15 19:06:44 +0800
  • f0f6f5fe build(mge/whl): package libnvrtc-builtins.so.11.x by Megvii Engine Team 2022-08-11 05:59:47 +0000
  • 3f5e8e9c docs(mge/functional): update copy example by thunder 2022-08-28 14:27:05 +0800
  • fb9d73b2 fix(src/tensorrt): trt7 manage all workspace by Megvii Engine Team 2022-09-01 20:51:45 +0800
  • dee02289 fix(serialization): when the dump fbsv2 model is used, the middle_tensor becomes optional by Megvii Engine Team 2022-08-31 18:51:28 +0800
  • b2959589 docs(mge): update AvgPool2d example by Megvii Engine Team 2022-08-30 11:40:52 +0800
  • ada54ba5 docs(mge): update AvgPool2d & functional.nn.avg_pool2d example by Megvii Engine Team 2022-08-24 17:23:08 +0800
  • d8ea0168 docs(mge): update AvgPool2d & functional.nn.max_pool2d docstring by Megvii Engine Team 2022-08-24 16:59:05 +0800
  • 01353b2d Merge pull request #442 from Seeker98:avgpool_doc by Megvii Engine Team 2022-09-01 18:03:27 +0800
  • 1529bce5 perf(opencl): add opencl weight transpose kernel by Megvii Engine Team 2022-08-22 19:23:39 +0800
  • 5f08b82f (tag: v1.11.0, release-1.11) fix(dnn/cuda): fix ptx mma algo compute bugs by Megvii Engine Team 2022-08-17 16:11:49 +0800
  • d3e786ef feat(imperative): load_nerwork_and_run enable weight preprocess by Megvii Engine Team 2022-08-24 13:31:39 +0800
  • c6ff878d feat(mgb): add cu114 wheel by Megvii Engine Team 2022-08-19 16:58:03 +0800
  • 76cd4a52 feat(mgb): add cu114 wheel by Megvii Engine Team 2022-08-19 16:58:03 +0800
  • 25f97b76 feat(imperative): load_nerwork_and_run enable weight preprocess by Megvii Engine Team 2022-08-24 13:31:39 +0800
  • 5ee00943 fix(dnn/cuda): fix ptx mma algo compute bugs by Megvii Engine Team 2022-08-17 16:11:49 +0800
  • d1d8ddee fix(lite): fix lar multithread options invalid by Megvii Engine Team 2022-08-18 15:02:31 +0800
  • d4bf57d6 fix(lite): fix lar multithread options invalid by Megvii Engine Team 2022-08-18 15:02:31 +0800
  • 1404437a fix(mgb): fix the compatibility issue of cuda stub with older version drivers by Megvii Engine Team 2022-08-10 17:23:04 +0800
  • a6a2646c feat(arm): add AlgoFP32Winograd F43, and add filter size into name of winograd-related algorithms by Megvii Engine Team 2022-08-15 16:59:43 +0800
  • b8821edb perf(dnn/aarch64): optimize aarch64 sigmoid with asm by Megvii Engine Team 2022-08-12 14:03:41 +0800
  • 2b99bfec feat(arm): supports weight pre-processing for winograd benchmark tests by Megvii Engine Team 2022-08-15 16:54:32 +0800
  • ab34bac4 fix(opencl/extern_c_opr): fix cl_mem UAF issue 2/2 by Megvii Engine Team 2022-08-16 15:21:46 +0800
  • 421bcfd3 style(mgb/tools): add format for tools, dnn and ci by Megvii Engine Team 2022-05-10 17:36:27 +0800
  • 99309fa3 feat(mge/functional): add param output_padding for deconv ops by Megvii Engine Team 2022-08-01 15:18:41 +0800
  • 116781ba fix(mgb): fix megtee build errors by Megvii Engine Team 2022-08-15 16:37:25 +0800
  • 0467778f chore(release): bump version by 温娟 2022-08-15 06:17:27 +0000
  • 54b5db17 feat(x86/rvv): add AGENT_NCHW_NCHW44 algo by Megvii Engine Team 2022-08-12 17:55:25 +0800
  • eaa18018 feat(x86/rvv): opt gi intrinsic helper for rvv, detail: https://github.com/riscv-collab/riscv-gnu-toolchain/issues/1106 by Megvii Engine Team 2022-08-10 00:27:34 +0800
  • 399db31a fix(dnn): fix build by Megvii Engine Team 2022-08-11 12:14:40 +0800
  • b8c7557b fix(mm): fix mm error when use sync by Megvii Engine Team 2022-08-10 16:57:24 +0800
  • 73ad06ba fix(mgb): fix fbsv2 model format no dump tensor format by Megvii Engine Team 2022-08-10 20:46:34 +0800
  • 399200b3 perf(serialization): optimized the memory usage when load new format model by Megvii Engine Team 2022-08-02 19:49:14 +0800
  • f31e52d5 feat(mgb): warpperspective support multi src input by Megvii Engine Team 2022-08-01 17:10:11 +0800
  • 669816e2 feat(dnn): warpperspective support multi src input by Megvii Engine Team 2022-07-31 00:46:23 +0800
  • 33b27be8 fix(mgb/build): fix trt8 build error by Megvii Engine Team 2022-08-05 19:42:04 +0800
  • bccda5c4 fix(mgb/imperative): fix repeat bug in trace mode by Megvii Engine Team 2022-08-04 15:50:20 +0800
  • fca6c76a fix(lite): fix input invalid bug in lar for fitting mode by Megvii Engine Team 2022-08-05 15:58:18 +0800
  • 1b943807 fix(dnn): fix reduce sum/mean error when b is large by Megvii Engine Team 2022-08-01 15:52:38 +0800
  • c7a99098 feat(cuda): add int4 ptx 256x64 mma kernel by Megvii Engine Team 2022-07-21 14:04:11 +0800
  • cf3ca1e9 feat(cuda): add int4 ptx 128x256 mma kernel by Megvii Engine Team 2022-07-21 14:03:51 +0800
  • 1f8e930e feat(cuda): add int4 ptx 128x128 mma kernel by Megvii Engine Team 2022-07-21 14:03:19 +0800
  • 1a2ed8c4 feat(cuda): add convbias ptx algo testcase by Megvii Engine Team 2022-07-21 14:00:15 +0800
  • 64551105 feat(cuda): add convbias ptx algo by Megvii Engine Team 2022-07-21 13:59:58 +0800
  • 8395a459 fix(dnn/fallback): fix naive shift multidefination error and optimize GiCvtFromInt32V4ToUint8 by Megvii Engine Team 2022-08-09 17:40:35 +0800
  • cc218550 feat(lite): load_and_run support optimize for inference by Megvii Engine Team 2022-08-04 13:56:59 +0800
  • 9bbe5500 fix(opencl/extern_c_opr): fix cl_mem UAF issue when run model OpenCL + ExternCOprRunner, for example by Megvii Engine Team 2022-08-05 15:31:53 +0800
  • 23a3d133 fix(dnn/softmax): create redcue and elemwise opr when get workspace size by Megvii Engine Team 2022-08-06 10:38:19 +0800
  • 2797fcfa fix(mge/device): add missed API to __all__ scope by Megvii Engine Team 2022-08-04 15:28:53 +0800
  • d7c546c9 fix(mge/interpreter): regenerates tensor when its dev value is needed by Megvii Engine Team 2022-07-28 17:59:13 +0800
  • 1f7bf1ad fix(opr): fix the compatilibity of elemwise multitype new mode by Megvii Engine Team 2022-07-29 21:07:42 +0800
  • b3a7d149 feat(dnn/fallback): add some new gi api by Megvii Engine Team 2022-08-01 17:13:50 +0800
  • 198ee068 feat(mgb/trt): update tensorRT toolchain to 8 by Megvii Engine Team 2022-07-06 14:26:08 +0800
  • 626222c6 fix(test): fix test for brainpp docker env by Megvii Engine Team 2022-08-03 13:50:31 +0800
  • fac67e7c feat(gopt): support nchw44 global pooling with fuse_grain by Megvii Engine Team 2022-07-22 12:07:12 +0800
  • 8461c8d8 fix(lite): fix ldr use lite interface error when open both fast-run and nchw44 by Megvii Engine Team 2022-08-01 18:40:59 +0800
  • 43bd949a fix(dnn): fix cudnn include by Megvii Engine Team 2022-08-01 16:43:07 +0800
  • 8abc3ab8 fix(imperative): fix convolution in rocm by Megvii Engine Team 2022-07-27 15:10:33 +0800
  • 3b1101b5 feat(ci): update image by huangxinda 2022-07-29 14:43:25 +0800
  • 32b31fd5 fix(mgb): change the check method of cuda sm code by Megvii Engine Team 2022-07-24 15:31:41 +0800
  • 5f863682 Revert "feat(dnn): add elemwise modes" by Megvii Engine Team 2022-07-28 19:49:44 +0800
  • d2a1905a Revert "feat(mgb): add cumprod opr" by Megvii Engine Team 2022-07-28 19:49:24 +0800
  • 49e14f87 feat(mgb): add cumprod opr by Megvii Engine Team 2022-04-25 18:46:17 +0800
  • 87aedc29 feat(dnn): add elemwise modes by Megvii Engine Team 2022-06-14 12:38:17 +0800
  • 25e89d68 feat(gi/rvv): remove winograd rvv do not use FIXLEN workaround by Megvii Engine Team 2022-07-27 20:17:43 +0800
  • fe5b1834 fix(mgb/imperative): fix the problem of occasional failure during testing of redis by Megvii Engine Team 2022-07-18 11:51:13 +0800
  • b3f46734 feat(megdnn/softmax): add softmax operator in fallback by Megvii Engine Team 2022-06-01 10:23:14 +0800
  • 6c78e684 fix(lite): fix lite memory leak by Megvii Engine Team 2022-07-12 15:08:21 +0800
  • ff239c63 feat(lite): add unit test for lar by Megvii Engine Team 2022-07-08 13:43:17 +0800
  • 7bf1c38c fix(mgb/imperative): fix imperative code gen by Megvii Engine Team 2022-07-19 11:58:16 +0800
  • c49d3070 refactor(imperative/ops): extends DnnOprCaller with template by Megvii Engine Team 2022-07-01 16:58:22 +0800
  • 2d6476a4 feat(lite): add auto decide model inference format option by Megvii Engine Team 2022-06-30 17:11:16 +0800