2516 Commits (824af20bd895d4908c3f130eed597cb52ee69e14)
 

Author SHA1 Message Date
  Megvii Engine Team 0447574446 feat(opencl): add OpenCL cache compat level api 3 years ago
  Megvii Engine Team 6fb5a34360 build(flatbuffer/cx2): fix cx2 build and fix uclibc build flatbuffer 3 years ago
  Megvii Engine Team 87de704a46 feat(gopt): fuse conv h_swish 3 years ago
  Megvii Engine Team 4adba37867 feat(lite): add example script and some small change for lar 3 years ago
  Megvii Engine Team 87f00232f2 fix(mge/gm): fix missing dtype checking while attach tensors 3 years ago
  Megvii Engine Team 3726f5cc92 feat(gopt): merger consecutive relayout and dimshuffle to one relayout to optimize CD4 performarce 3 years ago
  Megvii Engine Team 1fead9b6b0 feat(gopt): merge consecutive dimshuffle and relayout to one relayout to optimize CD4 performace 3 years ago
  Megvii Engine Team 26d1e4f7ed feat(gopt): optimize cd4 pass rule for elemwise and typecvt to let cd4 start as soon as possible 3 years ago
  Megvii Engine Team ac26bdcef5 fix(cuda): fix direct conv speed and memory problem 3 years ago
  Megvii Engine Team f7994683bd feat(cuda): add large kernel direct conv to heuristic algo chooser 3 years ago
  Megvii Engine Team 6dc0c0b9cc fix(dnn): fix the sync problem in some kernels 3 years ago
  Megvii Engine Team 04193e3bd1 feat(dnn): add nearest mode for remap and resize 3 years ago
  Megvii Engine Team 69b89388e8 docs(mge/functional): fix debug_param set_execution_strategy docstring 3 years ago
  Megvii Engine Team 93c7e45188 feat(arm): delete the reduant implement 3 years ago
  Megvii Engine Team e34a642b31 feat(fallback): reduce support general intrinsic 3 years ago
  Megvii Engine Team 10f23778a8 feat(fallback): add simd general intrinsic 3 years ago
  Megvii Engine Team 286051ede1 feat(dnn): differentiate sass kernel with cuda version 3 years ago
  Megvii Engine Team f78b60ec10 feat(bazel): make bazel gensass depend on cuda toolchain version automatically 3 years ago
  Megvii Engine Team f48227c07d feat(mgb): show more details for cuda driver api call 3 years ago
  Megvii Engine Team bb5af9b475 feat(lite): hidden lar gflags symbols for static link 3 years ago
  Megvii Engine Team d8bb3ff5b4 fix(cuda): fix fp16 tensorcore gemm split k workspace 3 years ago
  Megvii Engine Team 597efed40b feat(lite): add get last error code interface in lite c 3 years ago
  Megvii Engine Team 90c8a58cca docs(docstring): add pad docstring 3 years ago
  Megvii Engine Team 5f4501e0f3 fix(gopt): fix conv bias fuse 2 noline 3 years ago
  Megvii Engine Team ac2f548c9a docs(imperative/dataloader): update preload description 3 years ago
  Megvii Engine Team 73b518b718 feat(lite): add get physic addr interface in lite 3 years ago
  Megvii Engine Team f67086adde fix(lite): fix lite global layout transform symvar replace error 3 years ago
  megvii-mge 4462953fba feat(mge/third_party): update cutlass version 3 years ago
  Megvii Engine Team d7b0994a3e feat(cuda): add fp16 compute 16 kernel 3 years ago
  Megvii Engine Team 8a2e92bd6c refactor(cuda): depthwish large kernel 3 years ago
  Megvii Engine Team 6b8a69d5b6 feat(cuda): float16 depthwise large kernel conv compute fp32 3 years ago
  Megvii Engine Team bc385b5374 feat(cuda): support float16 depthwise large kernel conv 3 years ago
  Megvii Engine Team 7d2063e35a perf(cuda): speedup conv backward data with small feature map and large filter size 3 years ago
  Megvii Engine Team 72403e8929 perf(cuda): speedup chanwise conv with small feature map and large filter size 3 years ago
  Megvii Engine Team 28d48f2f7a fix(mgb/src): fix megbrain cmake unsupport android_nn 3 years ago
  Megvii Engine Team ab6d12caff feat(mge): add conv padding mode 3 years ago
  Megvii Engine Team 177001d5e5 refactor(dispatch): allow dynamic type creation 3 years ago
  Megvii Engine Team 150a6a6151 perf(dispatch/trace): remove unnecessary h2d for constant 3 years ago
  Megvii Engine Team 81d8c73a41 perf(dispatch/trace): serval tricks to speed up trace 3 years ago
  Megvii Engine Team 4fa6162027 perf(dispatch): improve performance of dispatch system 3 years ago
  Megvii Engine Team ca00177719 perf(dispatch): speed up dispatch system 3 years ago
  Megvii Engine Team 187c1dc081 fix(jit): copy aux var when shallow copying JITExecutor 3 years ago
  Megvii Engine Team 7bd848ce04 fix(subgraph): fix hand-written backward for serval jit-elemwise ops 3 years ago
  Megvii Engine Team 7be7656c9f fix(imperative): explicitly manage global structures 3 years ago
  Megvii Engine Team 62034fb262 fix(imperative): make CompNode finalize happens before global object destructor 3 years ago
  Megvii Engine Team 59cbf9583d fix(subgraph): use CompiledOp in cpu to avoid workspace error 3 years ago
  Megvii Engine Team b6ce02a152 fix(subgraph): fallback back to cg if jit unsupported 3 years ago
  Megvii Engine Team 21f5a7fcc0 fix(subgraph): fix device recognition and scalar propagate 3 years ago
  Megvii Engine Team 27346b0b65 test(opr): add scalar check for opr_test 3 years ago
  Megvii Engine Team 225045236b perf(imperative): improve shape inference 3 years ago