1. 11 10月, 2022 1 次提交
  2. 12 8月, 2022 1 次提交
  3. 10 8月, 2022 4 次提交
  4. 09 8月, 2022 2 次提交
    • C
      [Cherry-pick] Several bugs fix (#44991) · e00aa903
      Chen Weihang 提交于
      * fix device context init error (#43910)
      
      * Fix core so name mismatch error (#43977)
      
      * fix core avx soname error
      
      * remove print info
      
      * add clip_extra (#44008)
      
      * fix tensor stream error in custom op (#44500)
      
      * fix custom op attr names size error (#44938)
      e00aa903
    • C
      add post layer norm (#44931) · c5f4a9cc
      carryyu 提交于
      c5f4a9cc
  5. 08 8月, 2022 2 次提交
  6. 05 8月, 2022 2 次提交
  7. 04 8月, 2022 3 次提交
  8. 03 8月, 2022 1 次提交
  9. 02 8月, 2022 4 次提交
  10. 01 8月, 2022 1 次提交
    • zhouweiwei2014's avatar
      [UT]fix test_poisson op random fail (#44763) · b71833ea
      zhouweiwei2014 提交于
      修复poisson op单测随机挂
      
      原因:由于随机OP的无法直接验证数值正确性,该单测随机采样100万个样本,统计落到直方图各区间的数量,计算出粗略的概率密度函数,与标准概率密度函数对比,这种测试方式会有一定误差。
      当采样数量越小,误差越大,因此该PR增大采样样本数量(100万->200万),误差进一步减小在rtol范围内。
      b71833ea
  11. 25 7月, 2022 1 次提交
  12. 19 7月, 2022 1 次提交
    • C
      Record op shape data for profiler [cherry-pick PR43405 43578 43822] (#44384) · a2240190
      chenjian 提交于
      * add serialization for new field in event node (#43405)
      
      * add serialization for new field in event node
      
      * fix a bug
      
      * add more field to memory record (#43578)
      
      * Add infer shape in dygraph (#43822)
      
      * record memory and op supplement info
      
      * update
      
      * update
      
      * fix a bug
      
      * fix memory recording
      
      * fix a bug
      
      * update
      
      * update
      
      * fix a bug
      
      * update
      
      * fix a bug
      
      * fix a bug
      
      * fix a bug
      
      * update dygraph record
      
      * add infer shape record
      
      * fix
      
      * fix
      
      * fix
      
      * add comments
      
      * fix a bug
      
      * fix
      
      * fix
      
      * add record op info
      
      * fix file mode
      
      * add op input shape info
      
      * fix dependency
      a2240190
  13. 12 7月, 2022 1 次提交
  14. 01 7月, 2022 1 次提交
  15. 30 6月, 2022 4 次提交
  16. 29 6月, 2022 2 次提交
    • Q
      Fix elementwise_div UT by providing user defined gradients (#43536) (#43909) · 26187c27
      Qi Li 提交于
      Cherry-pick of #43536
      
      Backgroud in #43262
      
      In elementwise_div UT, the numeric gradient (validation) has large relative error in comparison to analytic gradient (Paddle OP).
      
          The default rtol for UTs is 0.005
          The rtol for float32 and float64 elementwise_div OP is set to be 0.05
          The rtol for float16 and bfloat16 elementwise_div OP is set to be 1.0
      
      The relative error is too large, so this PR provides user defined gradients to test elementwise_div followed by the analytic method.
      26187c27
    • R
      cherry pick 43890 (#43892) · 69e82d83
      ronnywang 提交于
      * cherry pick 43890
      69e82d83
  17. 28 6月, 2022 4 次提交
  18. 27 6月, 2022 2 次提交
  19. 25 6月, 2022 2 次提交
  20. 24 6月, 2022 1 次提交
    • A
      [cherry-pick] NVIDIA fixes (#43780) · 9edbe4aa
      Aganlengzi 提交于
      * Use all sitepackages path as the library/include path (#42940)
      
      * Fix several unit tests and increase the unit tests stability (#43670)
      
      * Reduce gather op unit tests size and increase the timeout
      
      * Add NVIDIA_TF32_OVERRIDE for multi-processes environment
      
      * Remove record test for device event ut
      
      * Fix 3 unittest errors (#43532)
      
      * Fix test_fuse_resnet_unit failure
      
      * Fix test_imperative_auto_mixed_precision failure
      
      * Fix sparse_attention_op error
      
      * Fix sparse_attention_op error
      
      * Use fixed random seed (#43659)
      
      * for CI test_collective_sendrecv_api
      Co-authored-by: Nzlsh80826 <rewang@nvidia.com>
      Co-authored-by: NShijie <505749828@qq.com>
      9edbe4aa