308 Commits (4e0c9ad3c669489a95c53dc63ef7efafd959b65a)
 

Author SHA1 Message Date
  Megvii Engine Team 4e0c9ad3c6 feat(mgb/external): extern-c-opr dumper and loader for MACE 5 years ago
  Megvii Engine Team ca52a93e9f fix(mge/quant): fix init value of histogram observer 5 years ago
  Megvii Engine Team 9f352b1c45 feat(megbrain/dnn): add indexing remap int32 for naive and cuda 5 years ago
  Megvii Engine Team 5dbf218d19 feat(dnn/x86): add sse 8816 matmul 5 years ago
  Megvii Engine Team 25b6a13148 feat(dnn/x86): add x86 avx2 8x8x16 matmul 5 years ago
  Megvii Engine Team 273f891b55 fix(mgb/gopt): fix run-time winograd-transform and nchwxx error 5 years ago
  Megvii Engine Team 02abc36ea6 fix(mbg/arm_common): fix nchw44-dot misc issue 5 years ago
  Megvii Engine Team 9ed3882a94 fix(opr/dnn): fix winograd fast run mismatch 5 years ago
  Megvii Engine Team 18be23f328 fix(mbg/gopt): fix nchwxx gopt with no fuse conv_bias and winograd 5 years ago
  Megvii Engine Team 38f7cbd9aa fix(mge/module): fix redundant recursion in `train()` 5 years ago
  Megvii Engine Team 5c2323529d test(mge/quantization): add `quantize_disabled` related test 5 years ago
  Megvii Engine Team ab91302515 feat(mge/quantization): add `quantize_disabled` attribute in Module 5 years ago
  Megvii Engine Team f4ead78852 feat(mgb): static allocation with given padding 5 years ago
  Megvii Engine Team 575a6dca9f ci(docker): change object files permission 5 years ago
  Megvii Engine Team 08ac685e3a feat(mge/functional): add logsumexp 5 years ago
  Megvii Engine Team 65ec4f7c26 fix(ci): fix test timeout 5 years ago
  Megvii Engine Team ea6bfe6cd9 fix(dnn/cuda-stub): simplify and use proper search paths 5 years ago
  Megvii Engine Team 01092feb9b feat(mgb): add PackAllReducePass 5 years ago
  Megvii Engine Team c7e6c658fd refactor(mge/distribute): use is_root (and rank) in stead of rank and root at collective comm 5 years ago
  Megvii Engine Team ff308e3b62 feat(mgb/comp_node): generate uid for cuda comp node 5 years ago
  Megvii Engine Team 32c86211ee fix(dnn/cuda): enable cuda algos for nchw quantized 5 years ago
  Megvii Engine Team b8d8886e35 refactor(mge/tensor): combine Dict and TensorDict 5 years ago
  Megvii Engine Team 7751a0676e docs(mge/tensor): add advanced index related docs 5 years ago
  Megvii Engine Team 7b0dbe6af8 fix(dnn/arm): fix stride 1 support for int8 nchw_nchw44 5 years ago
  Megvii Engine Team 198f3eb5f6 fix(dnn/arm): fix fp32 nchw44 direct workspace bug 5 years ago
  Megvii Engine Team 49fdddef8d fix(gopt): fix reorder arith chain pass 5 years ago
  Megvii Engine Team 6742a58b7e fix(quant): observer do not use cond_take 5 years ago
  Megvii Engine Team 9e876203b5 feat(dnn): add int8 direct conv dot nchw44 5 years ago
  Megvii Engine Team 09ceaaaecf fix(dnn/arm): stride1 support for nchw_nchw44 fp32 conv 5 years ago
  Megvii Engine Team 50db9b84c2 fix(gopt): fix paramfuse if the endpoint is const 5 years ago
  Megvii Engine Team 35bc0e1f60 fix(mge/function): do not deeply copy saved tensor in Function 5 years ago
  Megvii Engine Team 47377c7be5 fix(core): fix memory defragmenter 5 years ago
  Megvii Engine Team f56f187f6e fix(mbg/gopt): fix nchw44-dot channel wise trans to nchw44 5 years ago
  Megvii Engine Team af29fcb2e3 feat(mgb/plugin): add param json func for indexing oprs 5 years ago
  Megvii Engine Team 62753c4d30 fix(mge/sdk): fix comp_node bug in dump_with_testcast_mge 5 years ago
  Megvii Engine Team f1c86606cb fix(dnn/cuda): fix FuseConvBiasWithZ pass for HSwish activation 5 years ago
  Megvii Engine Team adfa468899 fix(mge/functional): fix scatter doctest failed for GPU platform issue 5 years ago
  Megvii Engine Team 4f8e60801c feat(dnn): fix Werror by adding macro 5 years ago
  Megvii Engine Team d7bb62cfa1 refactor(mgb): move mm_handler from python module into opr-mm 5 years ago
  Megvii Engine Team 84068a6bb1 fix(mge/data): fix typos in voc and objects365 5 years ago
  Megvii Engine Team 3966bb08b3 feat(dnn/test): split cpu.convolution 5 years ago
  Megvii Engine Team 8f87a3e988 feat(dnn/arm_common): add int8 nchw44 winograd f23_4x4 f23_8x8 compute float32/int16 output int8 5 years ago
  Megvii Engine Team 8ffed043be fix(dnn/x86): fix matrix_mul quantized performance on vnni 5 years ago
  Megvii Engine Team 1d860f4d6f fix(dnn/x86): fix dnnl int8 algo on vnni 5 years ago
  Megvii Engine Team 871e6a516f feat(dnn/x86): opt x86 quantized heuristic 5 years ago
  Megvii Engine Team 6c29548d20 fix(dnn/arm): fix nchw_nchw44 dot stride1 support 5 years ago
  Megvii Engine Team 02cbb13bbc fix(dnn/arm): fix nchw44 fp32 direct algo oh block and unused stride2 algo 5 years ago
  Megvii Engine Team d2f5874a52 fix(mge/module): fix non-str key error of dict in module 5 years ago
  Megvii Engine Team 30b3d3aa3e fix(dnn/gopt): add convolution nchw44-dot format gopt 5 years ago
  Megvii Engine Team 48d1ac1433 fix(dnn/arm): fix consistence between create_conv1x1_strategy and can_create_conv1x1_strategy 5 years ago