PaddlePaddle / Paddle
大约 1 年前同步成功

代码
- 文件
- 提交
- 分支
- Tags
- 贡献者
- 分支图
- Diff
Issue 1423
- 列表
- 看板
- 标记
- 里程碑
合并请求 543
Wiki 0
- Wiki
分析
- 仓库
- DevOps
项目成员
Pages

Please add GELU in-place execution support

#24328 · opened 5月 07, 2020 by saxon_zh Intel InterestingIdea
- 1
更新于5月 20, 2020
[DNNL] Check for distributing training example and try running on MKL-DNN engine

#23135 · opened 3月 22, 2020 by saxon_zh Intel InterestingIdea
- 0
更新于3月 22, 2020
[DNNL] Evaluate trained paddle models using mkl-dnn kernels

#23134 · opened 3月 22, 2020 by saxon_zh Intel InterestingIdea
- 0
更新于3月 22, 2020
[DNNL] Speedup conv grad DNNL kernel

#23133 · opened 3月 22, 2020 by saxon_zh Intel InterestingIdea
- 0
更新于3月 22, 2020
[DNNL] Implement operator softmax with cross entropy

#23073 · opened 3月 18, 2020 by saxon_zh Intel InterestingIdea
- 0
更新于3月 18, 2020
[DNNL] Doc on Int8 architecture

#22712 · opened 2月 21, 2020 by saxon_zh Intel InterestingIdea design_doc int8
- 0
更新于2月 21, 2020
[DNNL] Add DNNL reshape/reshape2 kernel

#22543 · opened 2月 11, 2020 by saxon_zh Intel InterestingIdea
- 0
更新于2月 11, 2020
[DNNL] speedup convolution grad ops

#22539 · opened 2月 11, 2020 by saxon_zh Intel InterestingIdea
- 0
更新于2月 11, 2020
[DNNL] Gather shapes of all FC and work with DNNL to have it speeded up

#22477 · opened 2月 06, 2020 by saxon_zh Intel InterestingIdea
- 0
更新于2月 06, 2020
[DNNL] Add fuse of FC + GELU

#22476 · opened 2月 06, 2020 by saxon_zh Intel InterestingIdea
- 0
更新于2月 06, 2020
[DNNL] Enable mkldnn_placement pass to work on inner ops of while op

#22407 · opened 1月 30, 2020 by saxon_zh Intel InterestingIdea
- 0
更新于1月 30, 2020
[DNNL] Inspect if Passes works fine for NHWC Paddle layout

#22396 · opened 1月 23, 2020 by saxon_zh Intel InterestingIdea inference
- 0
更新于1月 23, 2020
[MKL-DNN] Implement pass fusing FC and LayerNorm

#21737 · opened 12月 13, 2019 by saxon_zh Intel InterestingIdea
- 1
更新于12月 13, 2019
[MKL-DNN] Demo_CI should not link to mkl-dnn directly

#21297 · opened 11月 21, 2019 by saxon_zh Intel InterestingIdea
- 0
更新于12月 12, 2019
[MKL-DNN] Explore if MKLDNN::memory::reshape & subview can be useful

#20898 · opened 10月 30, 2019 by saxon_zh Intel InterestingIdea
- 0
更新于12月 12, 2019
[MKL-DNN] Reduce memory footprint by using custom scratchpad of MKL-DNN 1.0

#20615 · opened 10月 14, 2019 by saxon_zh Intel InterestingIdea
- 0
更新于12月 12, 2019
[MKL-DNN] FC does not support multiple input Tensors use case

#20145 · opened 10月 01, 2019 by saxon_zh Intel InterestingIdea
- 0
更新于12月 12, 2019
[MKL-DNN] Enable demo training for MKL-DNN

#19977 · opened 9月 24, 2019 by saxon_zh Intel InterestingIdea enhancement
- 0
更新于12月 12, 2019
[MKL-DNN] Replace manually implemented reorder iohw->oihw with mkl-dnn one

#18842 · opened 7月 26, 2019 by saxon_zh Code Cleanup Intel InterestingIdea
- 0
更新于12月 12, 2019
[MKL-DNN] Add acquire API reusing for quantize , dequantize & requantize ops

#18729 · opened 7月 22, 2019 by saxon_zh Intel InterestingIdea enhancement inference int8
- 1
更新于2月 19, 2020