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

Commit Graph

  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • *
  • |\
  • * \
  • |\ \
  • | | | *
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • * | | |
  • | * | |
  • | | * |
  • | |/ /
  • | | | *
  • | | |/
  • | |/|
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • |/ /
  • * |
  • * |
  • * |
  • |\ \
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • | | *
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • | | *
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • * | |
  • | |/
  • |/|
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • * |
  • 10a0349e feat(lite): add assert log for set_data_by_share and set_data_by_copy. pylite network input is not correct when input np is not continuous by Megvii Engine Team 2022-07-19 17:31:55 +0800
  • f5597d9a fix(mgb): make error infomation of input channel mismatch more readable by Megvii Engine Team 2022-07-19 11:24:36 +0800
  • 38bd5999 fix(mgb): make error infomation of invalid MatMul more readable by Megvii Engine Team 2022-07-19 11:23:33 +0800
  • e0d505e6 fix(mgb/dnn): fix bug that some cutlass file compile very slowly on SM86 by Megvii Engine Team 2022-07-12 16:53:12 +0800
  • cc31b9db docs(mge/functional): fix vision docs by Megvii Engine Team 2022-07-05 15:58:56 +0800
  • 6f7649e9 docs(docstring): fix pad docstring by Megvii Engine Team 2022-07-07 15:00:24 +0800
  • bf32d1e0 docs(dataloader): update dataloader docstring by Megvii Engine Team 2022-07-12 15:58:16 +0800
  • 4a32cc49 docs(mge/data): update Dataset class docstring by Megvii Engine Team 2022-07-12 14:00:47 +0800
  • 70fc5682 docs(mge/data): update MNIST dataset docstring by Megvii Engine Team 2022-07-11 19:41:12 +0800
  • 8fb062df Merge pull request #468 from MegEngine/HuaHua404-patch-2 by kagome1007 2022-08-01 14:58:53 +0800
  • e5ad3ea5 Merge pull request #469 from MegEngine/HuaHua404-patch-3 by kagome1007 2022-08-01 14:54:10 +0800
  • b6f247d3 (tag: v1.10.0, release-1.10) fix(build): disable cupti temporarily by Megvii Engine Team 2022-07-11 19:07:58 +0800
  • 58ba080d feat(x86/rvv): make gi conv algo adapt to vv and vf model by Megvii Engine Team 2022-07-14 18:37:02 +0800
  • bd50e457 feat(x86/rvv): make MATRIX_MUL_GI_F32_4x12 and FP32_GEMV_MK4_GI adapt to vv and vf model by Megvii Engine Team 2022-07-14 18:34:18 +0800
  • 5c3b4e95 feat(x86/rvv): opt AlgoFP32WinogradF63_4x4_NCHW44 by Megvii Engine Team 2022-07-13 19:45:19 +0800
  • fa59a7b0 feat(x86/rvv): opt AlgoF32DirectNCHWNCHW44 and opt GiMaximumFloat32/GiMinimumFloat32 on x86 by Megvii Engine Team 2022-07-13 18:43:55 +0800
  • 0d82e9b7 feat(x86/rvv): opt FB_GI_F32_MK4_4x8 by Megvii Engine Team 2022-07-12 17:23:05 +0800
  • 3fbceb3a fix(mgb/version): fix nvinfer.h not found by Megvii Engine Team 2022-07-13 13:18:55 +0800
  • a54d9cb9 feat(x86/rvv): opt FB_GI_F32_MK4_PACK_4x12 algo by Megvii Engine Team 2022-07-11 13:53:21 +0800
  • d60d028a feat(mge/device): enable to get cuda/cudnn/tensorrt version by Megvii Engine Team 2022-06-27 14:42:38 +0800
  • 8fed114a fix(build): disable cupti temporarily by Megvii Engine Team 2022-07-11 19:07:58 +0800
  • 19af2688 fix(mge/tools): fix module_stats for duplicated module by Megvii Engine Team 2022-07-05 18:18:07 +0800
  • 4cd4a38a fix(mge/tools): fix network_visualize for op without out shapes by Megvii Engine Team 2022-07-05 18:17:18 +0800
  • 7badcb72 build(wheel): copy libcupti.so by Megvii Engine Team 2022-07-09 04:51:49 +0000
  • 83fb1622 build(wheel): copy libcupti.so by Megvii Engine Team 2022-07-08 18:53:06 +0800
  • 6dfd5a4c fix(win7): workaround for hang when progress exit on win7+32bit by Megvii Engine Team 2022-07-04 20:08:58 +0800
  • ba1508e3 fix(lite): fix exception bug for load and run by Megvii Engine Team 2022-07-01 12:19:09 +0800
  • 603d0941 fix(lite): fix layout transform bug in lar for testcase model by Megvii Engine Team 2022-07-05 19:00:19 +0800
  • 8e48b17a (HuaHua404-patch-3) docs(README_CN): add key features description by HuaHua404 2022-07-06 14:27:19 +0800
  • 0004291a (HuaHua404-patch-2) docs(Readme): add key features description by HuaHua404 2022-07-06 14:21:58 +0800
  • 78f90469 (HuaHua404-patch-1) Update README_CN.md by HuaHua404 2022-07-06 13:59:51 +0800
  • 51397bfc feat(mgb): supports value infer and empty input tensor in ElemwiseMultiType by Megvii Engine Team 2022-06-24 13:31:33 +0800
  • 247e2f59 feat(mgb/dnn): add modes that the output type is bool in elemwise by Megvii Engine Team 2022-05-09 22:43:38 +0800
  • f3863810 fix(imperative): fix inplace operation of optim by Megvii Engine Team 2022-07-01 11:36:01 +0800
  • 9330929f feat(lite): add c_opr_init_interface for lar by Megvii Engine Team 2022-06-30 18:41:42 +0800
  • 16ba05a8 fix(dnn): fix dnn nchwxx elemwise performance by Megvii Engine Team 2022-07-01 15:08:52 +0800
  • 9de8c122 feat(mge/third_party): add cudnn-frontend by huangxinda 2022-07-05 14:23:37 +0800
  • 2b85948c feat(mge/third_party): update flatbuffers version by huangxinda 2022-07-04 18:11:41 +0800
  • 6c4c4ca6 Merge pull request #459 from Qsingle:fix_overflow_of_flops_calculate by Megvii Engine Team 2022-07-01 18:33:31 +0800
  • 7b17c118 refactor(dnn): make cudnn_frontend work by Megvii Engine Team 2022-06-28 15:37:49 +0800
  • 35e9cc98 feat(dnn/cuda): add cudnn frontend api by Megvii Engine Team 2021-12-06 16:30:09 +0800
  • 9be6d7b6 fix(lite): fix lite header by Megvii Engine Team 2022-06-29 17:21:53 +0800
  • ab8f6398 fix(test): make test install by Megvii Engine Team 2022-06-28 17:43:48 +0800
  • 99cfefbf fix(test): fix test copybara by Megvii Engine Team 2022-06-27 19:26:56 +0800
  • 0d7ace15 fix(mgb/dnn): suport fp16 for resize nhwc by Megvii Engine Team 2022-06-20 17:09:51 +0800
  • cfed86f9 feat(persistentcache): change file persistent cache with append model by Megvii Engine Team 2022-06-27 19:05:39 +0800
  • d19fc2c1 fix(imperative): add alloc TensorPtr in imperative by Megvii Engine Team 2022-06-06 20:44:35 +0800
  • 5d9af3ec fix(scripts): fix error of building whl with py310 by Megvii Engine Team 2022-06-27 10:41:50 +0800
  • d1b6c040 feat(mge/third_party): update MegRay version by huangxinda 2022-06-30 10:20:54 +0800
  • ac51f780 feat(mge/distributed): add support for batch send recv op by Megvii Engine Team 2022-06-01 21:10:13 +0800
  • 013bb14f build(tablegen): pregen opdef tablegen targets by Megvii Engine Team 2022-05-13 15:38:02 +0800
  • f12b75c0 perf(dnn/fallback): optimize some corner case in reduce by Megvii Engine Team 2022-06-22 19:59:46 +0800
  • 35cf0422 fix(ci): relax async timeout by Megvii Engine Team 2022-06-24 14:59:57 +0800
  • 7f024072 perf(dnn): speed up pad kernel by Megvii Engine Team 2022-05-30 12:35:52 +0800
  • 2886245b perf(imperative/src): improve pad host performance by Megvii Engine Team 2022-05-26 20:22:03 +0800
  • c6a350b1 chore(release): bump version by 温娟 2022-06-28 02:29:02 +0000
  • b55942a9 feat(dnn/naive/norm,-dnn/cuda/norm,-dnn/test/norm): add norm dnn opr, fwd only by Megvii Engine Team 2022-06-06 20:56:30 +0800
  • 7a7af8d7 fix(scripts): fix error of building whl with py310 by Megvii Engine Team 2022-06-27 10:41:50 +0800
  • 9d397727 fix(traced_module): fix bug of renaming tesnor by Megvii Engine Team 2022-06-23 14:28:22 +0800
  • 31f31cef fix(lite): fix record invalid for load and run by Megvii Engine Team 2022-06-23 15:08:52 +0800
  • 5a355138 fix(mgb): fix profile skip condition by Megvii Engine Team 2022-06-14 13:19:07 +0800
  • 5bdc430e fix(mgb/fastrun): fix megbrain fastrun memory overflow bug by Megvii Engine Team 2022-05-24 10:44:23 +0800
  • d7ddd43f feat(imperative): support python3.9 and fix some tests by Megvii Engine Team 2022-06-15 15:04:13 +0800
  • a4f019e7 fix(mgb/atlas): fix issue with profiling on Atlas by Megvii Engine Team 2022-06-15 11:28:34 +0800
  • 5ba76637 fix(traced_module): fix trace_module function may raise error in finally scope by Megvii Engine Team 2022-06-22 15:31:25 +0800
  • e6dcfbe8 fix(traced_module): fix traced module compatible issues by Megvii Engine Team 2022-06-10 16:57:39 +0800
  • 18f83a25 docs(mge/functional): fix F.svd docstring by Megvii Engine Team 2022-05-10 13:53:14 +0800
  • 0ba9326e docs(api/lite): add doc for lite common_enum_c_h by Megvii Engine Team 2022-06-22 16:23:52 +0800
  • 92ded721 docs(api/lite): add global setting docstring for megenginelite by Megvii Engine Team 2022-06-21 16:07:36 +0800
  • 171c683d docs(api/lite): add struct python api desc by Megvii Engine Team 2022-06-20 20:03:44 +0800
  • 3cd54dd6 docs(api/lite): add doc for lite tensor by Megvii Engine Team 2022-05-24 18:50:18 +0800
  • a891f9b3 docs(api/lite): add megenginelite.network api doc by Megvii Engine Team 2022-05-31 16:07:16 +0800
  • 5ef1ac75 docs(api/lite): add lite network api doc by Megvii Engine Team 2022-05-26 17:09:18 +0800
  • c47f48ef docs(api/lite): add lite global.h and pylite utils.py doc by Megvii Engine Team 2022-05-24 18:50:18 +0800
  • 5821c0b6 docs(lite): initial part of lite::Tensor comments as template by Megvii Engine Team 2022-04-19 18:55:58 +0800
  • 26c2563b feat(ci): update flatbuffers by huangxinda 2022-06-24 12:16:26 +0800
  • 4cdb7454 feat(rvv/fallback): make nchw44 happly on rvv by Megvii Engine Team 2022-06-22 16:52:21 +0800
  • 5e306b75 feat(x86): make conv1x1 and im2col available on with x86-NCHW44 add AlgoF32GiMK4Pack4x12 matrix_mul algo by Megvii Engine Team 2022-06-21 19:26:11 +0800
  • 481a6cbb feat(x86): make nchw44 happly on x86 by Megvii Engine Team 2022-06-21 19:23:02 +0800
  • 5873d5f5 feat(gi): add more gi api by Megvii Engine Team 2022-06-21 19:21:59 +0800
  • cfc41648 fix(mge): fix grad of maximum(x, x) by Megvii Engine Team 2022-06-20 16:21:09 +0800
  • bbafe699 feat(dnn): add elemwise COND_LT_MOV by Megvii Engine Team 2022-06-20 15:23:30 +0800
  • ed92b9c1 fix(cmake): fix compilation options for benchmark in cmake by Megvii Engine Team 2022-06-10 16:00:19 +0800
  • a49d5bf9 fix(autodiff): fix inplace operation on autodiff.Function by Megvii Engine Team 2022-06-20 16:46:03 +0800
  • 7252825c fix(functional): broadcast_to supports mutable target shape by Megvii Engine Team 2022-06-20 16:45:12 +0800
  • 2484cd27 fix(tensor): check args when construct tensor with existing tensor by Megvii Engine Team 2022-06-20 16:44:43 +0800
  • e7587617 fix(lite): fix packed model compatibility by Megvii Engine Team 2022-06-14 21:57:50 +0800
  • a0a5fcf1 feat(dnn): support tf32 by Megvii Engine Team 2022-04-19 17:49:03 +0800
  • f0088335 feat(mgb): upgrade flatbuffer by Megvii Engine Team 2022-06-15 15:22:25 +0800
  • 657db8dc chore(tools): remove dump_with_testcase_mge, user should use jit.dump instead by Megvii Engine Team 2022-06-08 16:36:04 +0800
  • 4d22e85b feat(ci): add completeness compatibility check by Megvii Engine Team 2022-06-10 10:54:39 +0800
  • f7b03959 perf(mgb/compile): improve compile time according the file map of compile time by Megvii Engine Team 2022-05-13 15:12:05 +0800
  • 124f38c4 perf(mgb/compile): improve compile time for megbrain by Megvii Engine Team 2022-05-12 17:17:47 +0800
  • d3247bee fix(dtr): always write shape when tensor produced by Megvii Engine Team 2022-06-13 14:03:02 +0800
  • 0a266d7a feat(riscv): speed up bazel build and fix rv64gc without rvv build by Megvii Engine Team 2022-06-15 20:12:12 +0800
  • 36ba1d6d fix(riscv): fix ci fp16 build and move test GI_TEST_NAIVE by megdnn_gi_api_test by Megvii Engine Team 2022-06-16 11:16:57 +0800
  • dcce4610 feat(cmake/riscv): make riscv happy by Megvii Engine Team 2022-06-14 18:16:58 +0800
  • 698dcef4 feat(gi/x86): fix _mm_slli_si128 build at clang by Megvii Engine Team 2022-06-13 17:59:43 +0800
  • 2d806f9c feat(gi): make conv_bias apply gi class type by Megvii Engine Team 2022-06-14 18:15:52 +0800
  • 19d36fa0 feat(gi): make pooling apply gi class type by Megvii Engine Team 2022-06-14 18:15:07 +0800