1. 22 11月, 2016 2 次提交
  2. 20 11月, 2016 1 次提交
    • Y
      Fix several cpp issues · 20600e70
      Yu Yang 提交于
      * Different Type compare.
      * ostream << should pass a const object.
      * remove always true checks.
      20600e70
  3. 17 10月, 2016 1 次提交
    • E
      Fix sparse training for trainer_count=1 (#204) · 28bc05b1
      emailweixu 提交于
      * Fix sparse training for trainer_count=1
      
      For trainer_count=1, the gradient machine is NeuralNetwork, which does not create parameter buf for PARAMETER_GRADIENT for sparse update in Parameter::enableType. But gradient parameter buf is still used in SgdThreadUpdater.
      
      * Minor update to comment
      28bc05b1
  4. 16 9月, 2016 1 次提交
  5. 14 9月, 2016 2 次提交
  6. 31 8月, 2016 1 次提交
  7. 30 8月, 2016 1 次提交
  8. 29 8月, 2016 1 次提交