Megvii Engine Team
|
207a346351
|
chore(mge): run get_device_count("gpu") in subprocess
GitOrigin-RevId: 0f0dc001cf
|
5 years ago |
Megvii Engine Team
|
869a03271b
|
perf(mgb): disable FoldingConvBiasDimshufflePass in cuda10 for performance
GitOrigin-RevId: d1b95a6f01
|
4 years ago |
liuke
|
0baf6b0d63
|
Merge pull request #175 from tpoisonooo:fix-spell-error
GitOrigin-RevId: fe649676ac
|
4 years ago |
Megvii Engine Team
|
239916a997
|
fix(mgb/gopt): fix testcase for enable nchw64 pass
GitOrigin-RevId: 2ae8d1608d
|
4 years ago |
Megvii Engine Team
|
2ab5c53f1d
|
feat(mgb/gopt): support nhwc conv in tensor reformat pass
GitOrigin-RevId: 43e78d758a
|
4 years ago |
Megvii Engine Team
|
009c90a2fe
|
feat(mgb/gopt): modify padding policy for 4bit conv bias oprs
GitOrigin-RevId: 188a2c3728
|
4 years ago |
Megvii Engine Team
|
b4687ce8da
|
feat(dnn/cuda): add convolution with i8 input and u4 output
GitOrigin-RevId: 8be439abf1
|
4 years ago |
Megvii Engine Team
|
bba04f02e5
|
feat(mgb/gopt): add fusion support for conv, astype(s4) and reformat
GitOrigin-RevId: 6329ca2c5f
|
4 years ago |
Megvii Engine Team
|
6d686ff26f
|
feat(gopt/inference): allow Float32 output dtype in EnableNCHW64Pass
GitOrigin-RevId: 1891efb76f
|
4 years ago |
Megvii Engine Team
|
7d3df995cb
|
feat(gopt/inference): allow Float32 output dtype in EnableNCHW4Pass
GitOrigin-RevId: 81100dbaf7
|
4 years ago |
Megvii Engine Team
|
e6caa9ff89
|
feat(opr): add bn backward for inference mode
GitOrigin-RevId: bb643cb62f
|
4 years ago |
Megvii Engine Team
|
77ead9377b
|
fix(src/serialization): fix compatibility error of oss model
GitOrigin-RevId: 43e0fa4fe1
|
4 years ago |
Megvii Engine Team
|
07de15713c
|
fix(mgb): remove static mem record from tee
GitOrigin-RevId: ac61b2a5eb
|
4 years ago |
Megvii Engine Team
|
4e4497b903
|
refactor(mgb/dnn): x86 pooling rebase algochooser
GitOrigin-RevId: 96cdc57180
|
4 years ago |
Megvii Engine Team
|
8a73193c2d
|
feat(dtr): remove eviction threshold
GitOrigin-RevId: 35c2014bf3
|
4 years ago |
Megvii Engine Team
|
43098fb8f1
|
feat(mge): add SlidingWindowTranspose opr
BREAKING CHANGE:
GitOrigin-RevId: 54d726d2fe
|
4 years ago |
Megvii Engine Team
|
1eaf32cd78
|
fix(mgb): fix typo in message
GitOrigin-RevId: 92d778b5df
|
4 years ago |
Megvii Engine Team
|
2cd9823210
|
fix(mgb/tensorrt): fix trt runtime, padding channel to a multiple of 4 when using kCHW4 IOFormat
GitOrigin-RevId: c5f1ed70da
|
4 years ago |
Megvii Engine Team
|
b078dda90b
|
feat(mge/random): add some random op and remove random/distrbution.py
GitOrigin-RevId: 4c05ebc266
|
4 years ago |
Megvii Engine Team
|
f30c0e06a6
|
feat(mgb/opr): add lsq opr
GitOrigin-RevId: 45494a2b57
|
4 years ago |
Megvii Engine Team
|
6cd01d5a74
|
feat(imperative/functional): let elemwise support empty IO & add some tests
GitOrigin-RevId: a5dc3b997c
|
4 years ago |
Megvii Engine Team
|
dea5278172
|
feat(mgb/opr): let PowC & TypeCvt support empty IO
GitOrigin-RevId: f97b3005fd
|
4 years ago |
Megvii Engine Team
|
2f68aeb9b6
|
feat(imperative/jit): let trace support empty IO
GitOrigin-RevId: 97a55242bf
|
4 years ago |
Megvii Engine Team
|
809d5056cd
|
feat(mge/distributed): enable pt shm allreduce
GitOrigin-RevId: 1dd5a02a51
|
4 years ago |
Megvii Engine Team
|
88898e63a5
|
fix(mgb): replace if_constexpr with runtime function to avoid potential
bug
GitOrigin-RevId: 27fe093d50
|
4 years ago |
Megvii Engine Team
|
1cfdbc565c
|
feat(dnn): add deterministic max pooling
GitOrigin-RevId: 9ab4c7a748
|
4 years ago |
Megvii Engine Team
|
933dd9a497
|
feat(mge/distributed): add cuda env check before forked thread
style(core/comp_node): reformat code
GitOrigin-RevId: 372452a8eb
|
4 years ago |
Megvii Engine Team
|
2a54196117
|
fix(tee): fix tee link
GitOrigin-RevId: db7b98524d
|
4 years ago |
Megvii Engine Team
|
a5060a2bfe
|
feat(mgb/opr): add check_has_inf kernel and opr
GitOrigin-RevId: 0d042dbfce
|
4 years ago |
Megvii Engine Team
|
3597a6dbd7
|
feat(dnn/arm): nchw_nchw44 conv support 1x1s1
GitOrigin-RevId: 8c8f7d7c76
|
4 years ago |
Megvii Engine Team
|
40085acbae
|
fix(mgb): remove unnecessary cudnn8 warning
GitOrigin-RevId: 04cf1bfca9
|
4 years ago |
Megvii Engine Team
|
54a4d70eb5
|
feat(src/serialization): add support of serializing metadata
GitOrigin-RevId: b563c94451
|
4 years ago |
Megvii Engine Team
|
721091faf0
|
fix(core): fix thread local is not supported in ios
GitOrigin-RevId: b7a6928f0b
|
4 years ago |
Megvii Engine Team
|
62bd6c823b
|
feat(cmake/debug): misc for build
* add asan build option
* fix cpuinfo build opt level
* fix host release build with out debug info
* opt "fix lite bazel/cmake symbols MR"
* other misc build opt
GitOrigin-RevId: 6ca286e195
|
4 years ago |
Megvii Engine Team
|
3e4e4c4604
|
feat(mgb/jit): add graph_opt_config and jit_config interfaces
GitOrigin-RevId: 170d9eeab2
|
4 years ago |
Megvii Engine Team
|
1c7d0802ab
|
fix(cuda): remove cuda driver version check and runtime minor version
GitOrigin-RevId: 4463beccf1
|
4 years ago |
tpoisonooo
|
7038a7f5d0
|
fix(quant): fix spell error
|
4 years ago |
Megvii Engine Team
|
355153e158
|
feat(mge/dtr): add DTR in computing graph
GitOrigin-RevId: 8941810319
|
4 years ago |
Megvii Engine Team
|
76f4f97536
|
refactor(sublinear): add SeqModifierBase
GitOrigin-RevId: 2d0393be6b
|
4 years ago |
Megvii Engine Team
|
f584416aa2
|
fix(dnn/bn): revise the conditions for inplace flag
GitOrigin-RevId: 59a104bf6a
|
4 years ago |
Megvii Engine Team
|
2eea00097c
|
feat(mgb): add fast run batch size graph option
GitOrigin-RevId: 94e333ec80
|
4 years ago |
Megvii Engine Team
|
47dcdf3e17
|
fix(mgb/core): fix dtype and resize modifiers for tensor
GitOrigin-RevId: a9d95a4cd8
|
4 years ago |
Megvii Engine Team
|
29f7cdb84a
|
fix(mgb/opr): correct nvof out shape computation
GitOrigin-RevId: 16bf086e92
|
4 years ago |
Megvii Engine Team
|
03ab8136e7
|
fix(core): fix asan error cause by wild thread_pool ptr
GitOrigin-RevId: b1c1b452cd
|
4 years ago |
Megvii Engine Team
|
0fb9cc41e4
|
fix(gopt): fix nchw64 opt pass
GitOrigin-RevId: dec18d1ab1
|
4 years ago |
Megvii Engine Team
|
86b69cacd0
|
fix(dnn): fixes for int4
GitOrigin-RevId: 845e164fd3
|
4 years ago |
Megvii Engine Team
|
adf75a291d
|
perf(dnn/cuda): add sass int4 128x128
GitOrigin-RevId: 1bc5482102
|
4 years ago |
Megvii Engine Team
|
8da2f698a3
|
feat(dnn/cuda): support warp perspective/pooling op when channel not aligned to 64
GitOrigin-RevId: 39f29ec990
|
4 years ago |
Megvii Engine Team
|
c218d4b029
|
feat(dnn/cuda): fallback conv qs4 support channel not aligend to 64
GitOrigin-RevId: f0d080f35c
|
4 years ago |
Megvii Engine Team
|
ae6ff2c5a6
|
feat(mgb/gopt): add opt pass for nchw64 layout transform
GitOrigin-RevId: adede7cef6
|
4 years ago |