1. 13 10月, 2017 1 次提交
  2. 12 10月, 2017 2 次提交
    • Cudnn conv op (#4195) · a3ccbdb3
      武毅 提交于
      * add cudnn_conv_op
      
      * WIP
      
      * update
      
      * update
      
      * fix grad check
      
      * use platform::memory
      
      * add support group for cudnn
      
      * update
      
      * follow comments
      
      * fix onlycpu build
      
      * update cuda define
      
      * follow comments
      
      * follow comments
      
      * merge with updates
      
      * fix compile error
      
      * follow comments
      
      * follow comments
      a3ccbdb3
    • Q
      Unify CUDA stream in Tensor CopyFrom interface (#4692) · 2603cb7e
      QI JUN 提交于
      * init
      
      * unify CopyFrom interface
      
      * fix gpu build error
      
      * fix bug in tensor_py.h
      
      * refine code comments and add TODO list
      
      * fix conflicts in FeedOp and FetchOp
      2603cb7e
  3. 11 10月, 2017 4 次提交
  4. 10 10月, 2017 3 次提交
    • Y
      Polish CompileTime InferShape · d350cdbc
      Yu Yang 提交于
      1. InferShape should be a method for `OpDescBind` not `Operator`, since
        `OpDescBind` is a compile-time concept.
      2. Pre-create operators and store them into a map. Make InferShape
         faster
      d350cdbc
    • Y
      Use PROTO_LITE when refactoring Paddle · dcb09e93
      Yu Yang 提交于
      It will significantly reduce binary size. It is useful for mobile
      deployment.
      dcb09e93
    • Y
      Stash · 49ca0b48
      Yu Yang 提交于
      49ca0b48
  5. 07 10月, 2017 1 次提交
  6. 06 10月, 2017 1 次提交
  7. 05 10月, 2017 3 次提交
  8. 03 10月, 2017 1 次提交
  9. 29 9月, 2017 2 次提交
  10. 28 9月, 2017 7 次提交
  11. 27 9月, 2017 5 次提交
  12. 26 9月, 2017 5 次提交
  13. 23 9月, 2017 5 次提交