101 Commits (bc738d4ec3e73f4e25a3631f71b9833fe6830e47)

Author SHA1 Message Date
  hangq 1efc6d4394 support scalar 5 years ago
  lyvette ef86629298 rewrite aware train converter 5 years ago
  mindspore-ci-bot fbab2ffbdf !7541 [lite] add identity pass and adjust log 5 years ago
  xuanyue 920cbb1e22 add identity pass and adjust log 5 years ago
  mindspore-ci-bot 934bb07656 !7655 【MSLITE】Weight quant add condition stop in kmeans method 5 years ago
  guohongzilong b12c8564b5 add condition stop for kmeans 5 years ago
  xutianchun bc3e37c8d5 fix post-training quantization mulit-output 5 years ago
  mindspore-ci-bot 4df56b6c1e !7513 [MSLITE] Add KMeans method for quant 5 years ago
  guohongzilong 16bfc0bd8b add kmeans method for quant 5 years ago
  hangq a3c6ad09a0 rewrite schduler & fix subGraph segmentation bug 5 years ago
  mindspore-ci-bot 15f57029a5 !7499 post training quantization support multi-input model 5 years ago
  xutianchun d401034645 support multi-input model 5 years ago
  guohongzilong 9790727216 support 16bit weight quant 5 years ago
  mindspore-ci-bot e14a6fc015 !7342 [MS][LITE][DEVELOP] add DeviceContext to Context 5 years ago
  mengyuanli 43633e28a6 add device_ctx to context 5 years ago
  guohongzilong 7ad32a773d add percent method 5 years ago
  mindspore-ci-bot eeda385f61 !7188 Change FC Quantization Method 5 years ago
  liuxiao78 470069c099 change FC quant to perchannel 5 years ago
  hangq 544023583b optimize include interface & optimize converter/benchmark/time_profiler interface 5 years ago
  xutianchun c5b5cb03c9 fix bias correction bug 5 years ago
  cjh9368 9f5ed13887 support scale int8 op 5 years ago
  mindspore-ci-bot d087c703eb !6829 bias correction 5 years ago
  xutianchun 46d6c6f197 bias correction 5 years ago
  cjh9368 ffc004395a support deconv quant op 5 years ago
  mindspore-ci-bot 6300fa822f !6841 [MSLITE][Develop] remove mul op weight quant 5 years ago
  mindspore-ci-bot faf871c971 !6555 generate quant param use warning log 5 years ago
  cjh9368 0ac7c7884d fuse benchmark and timeprofiler 5 years ago
  yangruoqi713 cd40cbbfb2 [MSLITE][Develop] remove mul op weight quant 5 years ago
  yeyunpeng 738b21c970 change log_adapter 5 years ago
  mindspore-ci-bot 1f07ffb19d !6754 promote weight quantization precision 5 years ago
  mindspore-ci-bot a3260757c3 !6717 fix quant relative 5 years ago
  xutianchun 561ebcca7e promote weight quantization precision 5 years ago
  xutianchun 5cf7874ac4 post training quantizaton: scale restrict 5 years ago
  yankai a1fea80b07 fix quant 5 years ago
  kai00 aa13c14944 log level change 5 years ago
  mindspore-ci-bot a845d36c26 !6494 support landmark models 5 years ago
  cjh9368 c5db8e0a32 int32 tensors don't insert dtype_cast op 5 years ago
  kai00 3b4c36223b add weight quant input check 5 years ago
  mindspore-ci-bot 81a5794f2d !6314 make error log clear 5 years ago
  cjh9368 fb217f3c81 remove inputInferenceType Param 5 years ago
  kai00 2628263077 deconv weight quant 5 years ago
  cjh9368 1cd9445087 Aware training support patial quant 5 years ago
  kai00 fb129986e5 fix weight quant 5 years ago
  mindspore-ci-bot 4bb05529d5 !5956 fix quant relative 5 years ago
  yankai 184e935c88 fix quant relative 5 years ago
  kai00 7dfac75de2 weight quant fix 5 years ago
  hangq 18c6ec2c8a optimize structure and interface 5 years ago
  hangq a6d9a4c3e0 clean redundant code 5 years ago
  hangq f1637f1157 1. clean review bot 5 years ago
  mindspore-ci-bot ed71f1134d !5697 mat mul weight quant 5 years ago