OpenI
/
MegEngine
Not watched
1
0
Fork 0
Code
Releases 31 Wiki Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain
1481 Commits
28 Branches
25 MB
3466 Download
Tree: c218d4b029
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from 'c218d4b029'
${ noResults }
Commit Graph

1481 Commits (c218d4b029adb9fd14c4d061da75370ae3768e0e)
 

Author SHA1 Message Date
  Megvii Engine Team c218d4b029 feat(dnn/cuda): fallback conv qs4 support channel not aligend to 64 4 years ago
  Megvii Engine Team 4fe68ac9ed feat(dnn/cuda): support transforming layout between nchw and nchw64 when channel not aligned to 64 4 years ago
  Megvii Engine Team ae6ff2c5a6 feat(mgb/gopt): add opt pass for nchw64 layout transform 4 years ago
  Megvii Engine Team 63a9bd30a8 feat(mgb/gopt): add an opt pass for padding channels to enable fast int8/int4 support on GPU 4 years ago
  Megvii Engine Team 56e863b7d4 fix(dnn/cuda): fix int4 epilogue stg bug 4 years ago
  Megvii Engine Team cff61a53d4 perf(dnn/cuda): optimize int4 sass conv main loop and epilogue without fuse_z 4 years ago
  Megvii Engine Team 12a0e61542 feat(dnn/cuda): add cuda elemwise int4 4 years ago
  Megvii Engine Team df1af59b5c feat(dnn): warp perspective support int4 4 years ago
  Megvii Engine Team 2398df079c feat(dnn/cuda): add cuda int4 pooling 4 years ago
  Megvii Engine Team 2a2a7f4552 test(mgb/opr): add testcase for conv bias int4 4 years ago
  Megvii Engine Team 858261af1f fix(python_module): fix conversion between numpy-ndarray and mgb tensor for qint4 and quint4 4 years ago
  Megvii Engine Team e250afb08f feat(dnn/cuda): support conv_bias for nchw64 and qint4 4 years ago
  Megvii Engine Team 3b9b87809d refactor(dnn): refactor lowbit tensor format 4 years ago
  Megvii Engine Team c74660ea88 fix(dnn/cuda): fix invalid local read for relayout format kernel 4 years ago
  Megvii Engine Team 8fef78d06d feat(dnn/cuda): add relayout format when width is an odd number 4 years ago
  Megvii Engine Team 91d6160769 feat(dnn/common): add tensor format for low-bits tensor layout 4 years ago
  Megvii Engine Team 19a554d674 test(dnn/cuda): add testcase for transforming tensor layout between nchw and nchw64 4 years ago
  Megvii Engine Team 71c2f61254 feat(dnn/cuda): add relayout format to support layout transform between NCHW and NCHW64 4 years ago
  Megvii Engine Team df009e89e1 feat(dnn/cuda): add cuda conv bias impls for NCHW format tensors with qint4 data type 4 years ago
  Megvii Engine Team ed92207585 feat(dnn/cuda): add conv bias impl for int4 data type using sass language 4 years ago
  Megvii Engine Team 52b55564d7 refactor(dnn/cuda): refactor reorder filter and bias kernel to support conv imma with data type s4 4 years ago
  Megvii Engine Team d2673c5abf fix(ci/windows): add windows cuda test 4 years ago
  Megvii Engine Team 2d6827c168 fix(mgb/windows): temporary workround on cuda-windows python exit 4 years ago
  Megvii Engine Team 517cc6846a ci(gitlab-ci): add inline lineno checking in copybara linter 4 years ago
  Megvii Engine Team 23032f50f2 feat(dnn/cuda): support float16 for index_incr_multi_axis_vec 4 years ago
  Megvii Engine Team 938944027d fix(mgb/dnn): fix cudnn8 convbias 4 years ago
  Megvii Engine Team 5427a67c96 fix(cmake/subdirectory): fix project import by other sdk by add-subdirectory 4 years ago
  Megvii Engine Team 241b35a697 refactor(ops): remove BackwardGraph op 4 years ago
  Megvii Engine Team d2e33af52f fix(mgb): fix wrong set of strategy in lar 4 years ago
  Megvii Engine Team 8b7d8d290b fix(core): fix json dump when weight preprocess 4 years ago
  Megvii Engine Team ec65e1f9ba fix(build/windows): fix windows build: 4 years ago
  Megvii Engine Team 8a918717c0 feat(mgb): add megengine inference 4 years ago
  Megvii Engine Team 379a28f9b0 feat(mgb): add load_and_run option 4 years ago
  Megvii Engine Team b06b589960 feat(mgb): get static graph memory info 4 years ago
  Megvii Engine Team 0cf4ff702e perf(dist): add fastpath for bcast params 4 years ago
  Megvii Engine Team 5e54821845 docs(syncbn): complete syncbn document 4 years ago
  Megvii Engine Team 3591ef1f6a fix(mgb): fix conv cudnnconvbackwarddata algo witch is not shake 4 years ago
  Megvii Engine Team e417798f9e fix(mge): correct pytype when calling apply from python 4 years ago
  Megvii Engine Team c4048519c6 chore(mge/version): change dev version to 1.5 4 years ago
  Megvii Engine Team 3352fe7003 refactor(dtr): import dtr as submodule 4 years ago
  Megvii Engine Team a7581986bc perf(dist): speed up bcast_val 4 years ago
  Megvii Engine Team 1525a02530 feat(mge/module): add python wrapper for unfold 4 years ago
  Megvii Engine Team d07ef83aa8 feat(mge/tools): add bash_completion for user to hint script 4 years ago
  Megvii Engine Team efc7f04fd3 refactor(mge/jit): using static global_enable for apply ctx insted of global variable 4 years ago
  Megvii Engine Team 750bde9513 refactor(mge/jit): remove is_compiled flag in cpp tensor 4 years ago
  Megvii Engine Team 4f240ec2d3 refactor(mge/jit): make trace return any kind of output 4 years ago
  Megvii Engine Team c6b552cf84 refactor(mge/functional): remove dependence to trace in functional implementations 4 years ago
  Megvii Engine Team 46d96478a1 feat(mge/tools): optimize statistical tools 4 years ago
  Megvii Engine Team 7f9d441cbb docs(mge/module): add note about weight shape in conv 4 years ago
  Megvii Engine Team 8be70d170c docs(misc): correct docstring format broadly 4 years ago