1. 27 10月, 2021 14 次提交
  2. 26 10月, 2021 13 次提交
  3. 25 10月, 2021 7 次提交
    • A
      [NPU] modifications for model ernie-1.0 (#36642) · 19b02d95
      Aganlengzi 提交于
      * [NPU] modifications for model ernie-1.0
      
      * rollback 503003 and change cast to dtype
      19b02d95
    • Z
      add op: fused_feedforward(backward) (#35611) · 2dd0a46a
      zhangkaihuo 提交于
      这个PR是fused_feedforward反向的代码
      
      相关kernel实现:fused_dropout_act_bias, fused_residual_dropout_bias, fused_layernorm_residual_dropout_bias
      
      fused_feedforward是一个融合算子,该算子对transformer模型的feed forward层的算子进行融合和封装,使得前端只呈现一个接口,通过融合减少部分访存和kernel launch的时间,以此提升性能。
      2dd0a46a
    • S
      Add bincount op (#36317) · 39f19127
      smallv0221 提交于
      * Add bincount op
      
      * upload cpu version
      
      * fix unitest
      
      * fix unittest
      
      * fix unittest
      
      * fix en doc
      
      * add more test
      
      * fix en doc
      
      * add more test case
      
      * fix test
      
      * fix input vailidation
      
      * fix input check
      
      * fix unittest
      
      * fix test
      
      * fix en doc
      39f19127
    • Z
      Create CinnCompiler class for compiling subgraphs found by build_cinn_pass. (#36562) · 4c460378
      Zhen Wang 提交于
      * Init the functions of CinnCompiler.
      
      * Add the unit test for CinnCompiler.
      
      * Fix some compilation errors.
      
      * Update the UT of cinn_compiler.
      
      * Use Decomposer&OpFusion passes in CinnCompiler::CompileGraph.
      
      * Update some comments.
      
      * Uncomment some includes in build_cinn_pass.cc.
      
      * Use refs instead of ptrs as returned types of FindGraph & Compile in
      CinnCompiler.
      
      * Use the merged CinnGraphSymbolization functions in CinnCompiler.
      4c460378
    • T
      add some ops to train ssd on kunlun (#36407) · 50778ad6
      TTerror 提交于
      * add some ops to train ssd on kunlun
      
      * add some ops to train ssd on kunlun
      
      * add some ops to train ssd on kunlun
      
      * update cast op unittest
      
      * update cast op unittest
      
      * update cast op unittest
      
      * update xpu cmake
      
      * update cast unittest
      50778ad6
    • Z
      add op: fused_feedforward(forward) (#35843) · b18cbfb2
      zhangkaihuo 提交于
      这个PR只包含fused_feedforward前向的代码。
      
      相关kernel实现:fused_dropout_act_bias, fused_residual_dropout_bias, fused_layernorm_residual_dropout_bias
      
      fused_feedforward是一个融合算子,该算子对transformer模型的feed forward层的算子进行融合和封装,使得前端只呈现一个接口,通过融合减少部分访存和kernel launch的时间,以此提升性能。
      b18cbfb2
    • J
      fix pool2d convert case (#36667) · 087c3abe
      JingZhuangzhuang 提交于
      087c3abe
  4. 24 10月, 2021 1 次提交
  5. 22 10月, 2021 4 次提交
    • W
      correct slice serialize data (#36588) · 5e880840
      wenbin 提交于
      * slice
      
      * add UT
      5e880840
    • Z
      add fp16 kernel for clip_op (#36577) · 1962d3af
      zhangbo9674 提交于
      1962d3af
    • L
      Fused attention op forward (#35905) · d4906214
      Li Min 提交于
      功能:本PR的目标是提高attention模块的计算性能。
      为了减少框架层对op的调度开销,本PR通过在C++层手动实现attention模块,对外提供attention 大op;
      为了减少防存开销,本PR采取了两种优化方法:
      (1)在q,k,v计算时通过共享输入X,将该处的gemm,transpose和bias add从三次调用减少为一次;
      (2)使用kernel融合优化技术,在不同cuda kernel之间通过寄存器传输数据;
      d4906214
    • L
      [hapi] support dygraph amp O2 (#36441) · 08248db0
      Leo Chen 提交于
      * [hapi] support dygrapg amp O2
      
      * fix problem of static pure fp16 in hapi
      
      * fix bug
      
      * fix format
      
      * fix ut
      
      * follow comments
      
      * update ut
      
      * update amp save/load
      
      * fix ut
      
      * refine code format
      08248db0
  6. 21 10月, 2021 1 次提交