149 Commits (48f4fd2f5213b3e27af8ae8df4e2baff1eafebfb)

Author SHA1 Message Date
  zhoufeng 8e97ce5cd2 build core.so 4 years ago
  huanghui 082fe01780 Add flag for more tuple inputs 4 years ago
  i-robot e7e7875088
!28792 Clean code 4 years ago
  tanghuikang f7e81a2c63 Clean code 4 years ago
  fangzehua ac11ceeee4 change omp to ms threadpool 4 years ago
  limingqi107 c3888a1e2f fix the pclint 4 years ago
  TronZhang 58f386fe2a Add Custom actor nodes and custom actor support for dynamic shape case 4 years ago
  liangzelang 706ad9360e delete unuseless code which cause nullptr bug. 4 years ago
  i-robot 16ca61644d
!28597 clean code for kernel graph 4 years ago
  kswang 66eb94dc03 clean code for kernelgraph 4 years ago
  liangzelang 26337a3d7d adopt to Ascend kernel by kernel in mindrt 4 years ago
  tanghuikang 87fccab600 MemScheduler handle nop node output and input from other graphs 4 years ago
  i-robot 39b3fc2922 !27975 Change "is_internal_output" to "is_internal_output_nop_node" 4 years ago
  dayschan ada1c80980 Change kAttrIsInternalOutput to kAttrIsInternalOutputNopNode 4 years ago
  i-robot 5645acee6c !27011 adjust comm executor order 4 years ago
  baihuawei e8f641e55b fix heterogeneous with mindrt 4 years ago
  i-robot 5655c9f972 !27390 Optimize swap strategy 4 years ago
  tanghuikang 6eb3cf1aee Optimize swap strategy 4 years ago
  dayschan c6a9547c24 Add "is_internal_output" attribute for related nodes 4 years ago
  liangzelang 20dcd6b148 enable subgraph sink for MindRT 4 years ago
  limingqi107 b84a0ef088 fix the bug of sub graph output tuple include tuple in the control flow sink 4 years ago
  kswang f7334cf57f adjust comm execute order 4 years ago
  He Wei 41dcac9c49 Replace std::unordered_map/set with robin-hood-hashing 4 years ago
  l00591931 9353bd10a4 Enable tuple as variable in optimizer 4 years ago
  dayschan cbb84ff580 Move IsRealKernel and IsRealCNodeKernel from AnfAlgo to AnfUtils 4 years ago
  tanghuikang bd04436fd8 PyNative RunOps support string input 4 years ago
  gaoyong10 83ee99b07d Get output nodes by control node. 4 years ago
  limingqi107 8ef51f22c1 unified runtime optimize the graph output 4 years ago
  limingqi107 19330b6f82 unified runtime support graph sink output update 4 years ago
  He Wei 57c75a9479 Support construct CNode by move inputs in 4 years ago
  He Wei be1ea10ca7 Optimize KernelGraph::ReplaceNode 4 years ago
  zhaosida 85e00ba3ab self code check 4 years ago
  zhaosida 7bbad8a84e self code check 4 years ago
  yelihua 83b3f6fef7 delete Common::GetRealPath 4 years ago
  John Tzanakakis 4b77150f94 make async dump when sink_mode=False use input_ctrl_tensor like sink_mode=True 4 years ago
  zhaosida 10d834f47d code clean master 4 years ago
  ms_yan 36a8886ca2 Revert "[feat] [assistant] [I3T96T] add new Dataset operator CMUARCTICDataset" 4 years ago
  djc b077aa1cab [feat] [assistant] [I3T96T] add new Dataset operator CMUARCTICDataset 4 years ago
  djc 4e6f7dc97d [feat] [assistant] [I3T96X] add new Dataset operator LibriSpeechDataset 4 years ago
  chenfei 4428db97e3 check abstract ref of update parameters 4 years ago
  i-robot 445681830d !21737 fix bug of common opt repeatly match a same node 4 years ago
  i-robot a71c0d19d3 !21696 unified runtime clear the resource of graph 4 years ago
  chenfei 6ea5e289ad add fg in todo list 4 years ago
  limingqi107 473db12d1e unified runtime clear the resource of graph 4 years ago
  hwjiaorui 1fa30a665f tensor device address setting 4 years ago
  i-robot 2f4fc75fcc !21396 modify the limit of axis of reduce ops 4 years ago
  simson 5b34859541 modify the limit of axis of reduce ops 4 years ago
  hwjiaorui e97df3a58f clean code 4 years ago
  i-robot 4c3c3ceee6 !21157 add AdamWeightDecay to optSet 4 years ago
  kswang 8fa85cac34 add adamweightdecay to opset and update lock input tensors 4 years ago