24 Commits (cdd3e50e986fd85375bd3cb89eabfa60ecc31896)

Author SHA1 Message Date
  Zhang Qinghua a137fa1d0b Optimize the Executors routines. 4 years ago
  chenhaozhe 086a871975 Change Loss to LossBase 4 years ago
  yangzhenzhang 7a40741048 add parallel operator for conv2d 4 years ago
  chenhaozhe 9da8534396 change _Loss to Loss 4 years ago
  yao_yf e967f1939b parallel envs variable check 4 years ago
  l00591931 9ec100d069 Change TensorAdd to Add, from r1.1 to master 5 years ago
  ms_yan deb1e6e965 use from_numpy and add do_copy option 5 years ago
  Xiaoda Zhang aa84484049 enabling approximation in DP algorithms 5 years ago
  hesham 9cee0d2143 Add num_epochs to non-sink training 5 years ago
  Yi Huaijie eb83ea9607 change internal API _get_strategy() to _get_shard_strategy() 5 years ago
  yao_yf 07117e4dd4 mv ParallelMode to context 5 years ago
  Xiaoda Zhang 9097b36950 add resnet50 testcases for gpu 5 years ago
  anzhengqi 008b91b2a1 inject epoch ctrl op in the execution tree and send eos at the end of epoch 5 years ago
  Yi Huaijie 8cfc05e4cf clean pylint warnings of parallel test cases 5 years ago
  jinyaohui 5a914994ba clean pylint 5 years ago
  jinyaohui 2907cf4488 remove some context param 5 years ago
  jinyaohui 391a060f21 remove two context param 5 years ago
  zjun c538b83712 remove enbale hccl 5 years ago
  Xiaoda Zhang 3ff9e54734 add the resnet50 32k-8p testcase 5 years ago
  lirongzhen1 56f785f7e6 add context configration 5 years ago
  Ziyan 0d208e00bd Model ALLTOALL as a single operator in cost model; scale the ALLTOALL, 6 years ago
  c00425699 ab917a734d fix bug for allreduce fusion and add resnet unit test 6 years ago
  Xiaoda Zhang fb6eed23ae refining strategy-checking for resnet50 6 years ago
  zhunaipan 930a1fb0a8 initial version 6 years ago