1. 21 7月, 2017 3 次提交
  2. 17 7月, 2017 1 次提交
  3. 11 7月, 2017 1 次提交
    • Y
      Fix slow parsing a recursive depends topology · 313e9f55
      Yu Yang 提交于
      * Fix  #2797
      * It because trainer_config_helpers' __dfs_travel__ did not record the
        node which travelled, and if the topology has a recursive dependency,
        there are some nodes will be travelled multiple times.
      * Add a `travelled` set to record which node is travelled.
      * Also add a unittest for this situation.
      313e9f55
  4. 10 7月, 2017 1 次提交
  5. 03 7月, 2017 1 次提交
  6. 29 6月, 2017 3 次提交
  7. 27 6月, 2017 1 次提交
  8. 24 6月, 2017 1 次提交
  9. 22 6月, 2017 2 次提交
  10. 07 6月, 2017 1 次提交
  11. 26 4月, 2017 1 次提交
    • Y
      Add error clipping to MT demo. · 3f1151a5
      Yu Yang 提交于
      * Compose GRU step naive layer in trainer config helpers.
        * It is uses mixed_layer for gate.
        * It supports ERROR_CLIPPING, DROPOUT
      * Add error clipping in MT demo.
      * Fix #1143
      * Fix #1891
      3f1151a5
  12. 02 2月, 2017 1 次提交
  13. 31 1月, 2017 2 次提交
  14. 09 12月, 2016 1 次提交
  15. 12 11月, 2016 1 次提交
  16. 02 11月, 2016 1 次提交
    • Q
      Add job=time in trainer, refine cudnn_conv to reduce gpu memory and speed up training. (#218) · 45c81a41
      qingqing01 提交于
      * Add benchmark for PaddlePaddle, tensorflow and caffe
      
      * ConvProjection to reduce memory for goolenet
      
      * Add unit test for ConvProjection.
      1. unit test in test_LayerGrad.
      2. compare the ConvPorjection and CudnnConvLayer, also compare the concat_layer+img_conv_layer and concat_layer_conv_projection.
      
      * Reduce cudnn_conv memory and add benchmark document.
      1. Use TmpMatrix as the workspace in cudnn_conv to reduce gpu memory. It reduce lots of memory.
      2. Add benchmark document.
      3. fix smallnet_mnist_cifar.py in paddle.
      
      * Add job=time and refine cudnn_conv to reduce gpu memroy and speed up
      
      * Refine cudnn_conv and shared biases operation in concat_layer and mixed_layer.
      
      * follow comments
      
      * follow comments
      
      * Use unique_ptr to prevent memory leaks in CudnnConvLayer.
      45c81a41
  17. 24 10月, 2016 1 次提交
  18. 17 10月, 2016 2 次提交
  19. 09 10月, 2016 1 次提交
  20. 28 9月, 2016 1 次提交
  21. 03 9月, 2016 1 次提交
  22. 01 9月, 2016 1 次提交
  23. 31 8月, 2016 1 次提交
  24. 30 8月, 2016 2 次提交
  25. 29 8月, 2016 1 次提交