Commit Graph

176 Commits

Author SHA1 Message Date
Liyan Zheng 37f3e9cf22 Add: NMutator::memboundToJson to export memboundOp 2023-04-18 09:56:14 +08:00
Liyan Zheng 885a978016 Add: nnet::Serializer supports FuncNode 2023-04-17 20:15:40 +08:00
Liyan Zheng a65dbff5f9 Chore: disable nnet_unimplemented_continue output 2023-04-17 13:24:56 +08:00
Liyan Zheng da3f5605e5 Add: exclude compile-time computable operator time 2023-04-17 13:24:14 +08:00
Liyan Zheng 45f7841d83 Add: C++ callback to export ONNX 2023-04-17 11:48:26 +08:00
Liyan Zheng 7a1d271c79 Add: RangeOpNode::getFullExpression() 2023-04-17 11:41:13 +08:00
Liyan Zheng b2c53458d9 Add: fuse membound operators 2023-04-17 11:38:48 +08:00
Liyan Zheng e76f2d0f10 Chore: simplify type names 2023-04-15 13:58:18 +08:00
Liyan Zheng f881db7ceb Add: export to ONNX with custom operators 2023-04-15 00:47:11 +08:00
Liyan Zheng b72662bb9b Add: infogan python interface 2023-04-14 16:36:55 +08:00
Liyan Zheng 582de83629 Fix: avoid reload library 2023-04-14 15:10:47 +08:00
Liyan Zheng b6b37ccf33 Add: invoke TVM through pipe 2023-04-13 21:43:56 +08:00
Liyan Zheng 9a009be337 Add: search engine uses estimated time 2023-04-13 19:46:54 +08:00
Liyan Zheng f0785225d7 Chore: format and comments 2023-04-12 11:48:11 +08:00
Liyan Zheng f939ebf8bb Add: comments for Matmul 2023-04-12 11:24:31 +08:00
Liyan Zheng 31b03ef91a Add: detailed output of Matmul 2023-04-12 11:24:31 +08:00
Liyan Zheng bd76a5c8d6 Add: CMake support of TVM 2023-04-12 11:24:31 +08:00
Liyan Zheng 83d5842f56 Fix: specify runtime 2023-04-12 11:24:31 +08:00
Liyan Zheng 005c113948 Add: support TVM packed function 2023-04-12 11:24:31 +08:00
Liyan Zheng 0d70643867 Add: Random generator 2023-04-12 11:24:31 +08:00
Liyan Zheng f29f54687c Add: PRelu Expr and AsTVMVisitor 2023-04-12 11:24:31 +08:00
huangshuhong 67e9fb9632 Fix: test_memboundOp 2023-04-12 11:24:29 +08:00
huangshuhong af80830d56 Fix: membound_tvm 2023-04-12 11:24:25 +08:00
huangshuhong f19bd4d1d6 Fix: remove cuCtxCreate which makes TVM fails 2023-04-12 11:24:25 +08:00
huangshuhong 2e501de8db Add: PackedFunc (broken) 2023-04-12 11:24:13 +08:00
Liyan Zheng 31b1edc600 Add: CMake config 2023-04-12 11:24:13 +08:00
Liyan Zheng 69d894e003 Add: TVM headers and CMake include paths 2023-04-12 11:24:13 +08:00
Liyan Zheng e8b4e3f03f Add: expr simplication for TVM 2023-04-12 11:24:13 +08:00
Liyan Zheng 0d8685b1e0 Fix: graph OpList ctor 2023-04-12 11:24:13 +08:00
Liyan Zheng 04858012bf Add: Exception for checkCuError 2023-04-12 11:24:13 +08:00
Liyan Zheng 6aaf963f50 Fix: cuBLAS failure when codegen is enabled 2023-04-12 11:24:13 +08:00
Liyan Zheng 6441abb88f Fix: GraphObj ctor should use cloneTensor 2023-04-12 11:24:13 +08:00
huangshuhong 220b76f355 Add: matmul broadcast 2023-04-12 11:24:09 +08:00
Liyan Zheng aac4bd4650 Fix: [Intermediate] NMutator::expressionToGraph
To be fix: matmul with implicit broadcast
2023-04-12 11:24:09 +08:00
Liyan Zheng 7478b698dc Add: expression reader as a cmake target 2023-04-12 11:24:09 +08:00
huangshuhong 8e6fc4136e Add: cache and padding (bugs!!) 2023-04-12 11:24:01 +08:00
Liyan Zheng bd12b67ea3 Add: mutator InfoGAN minimum test 2023-04-11 20:44:08 +08:00
Hardy fe1afe38fa
fix code of bang conv (#76)
* fix code of bang conv

* test: 向 master push 时也执行 ci

Signed-off-by: YdrMaster <ydrml@hotmail.com>

---------

Signed-off-by: YdrMaster <ydrml@hotmail.com>
Co-authored-by: wanghailu <wanghailu@qiyuanlab.com>
Co-authored-by: YdrMaster <ydrml@hotmail.com>
2023-03-29 15:47:32 +08:00
Hardy 823e66a9ff
Support perf bang 1115 (#57)
* support matmul

* add matmul

* add matmul

* add code for cnnl matmul operation and test

* add conv

* add code for conv test on mlu

* add code for test cnnl conv on mlu

* add code for perf conv and matmul on mlu

* clang format

* fix convolution operation

* fxi cmaklist

* code format

* fix code

* code format

---------

Co-authored-by: wanghailu <wanghailu@qiyuanlab.com>
Co-authored-by: wanghailu <wanghailu0717@163.com>
2023-03-29 13:52:56 +08:00
wendy12022 86ec4036ce
ADD: add mkl runtime for intel cpu , and add mkl kernel for matmul/conv/convtransposed. (#61)
* move memory format transformation to TensorObj

clang format

add MemoryFormat for tensorObj.

use post_ops for fused conv/deconv

Distinguish mkl  op_timer from cuda op timer.

add act optype to conv and deconv

add operator timer

add mkl kernel for convTransposed

minor fix for group conv

do not use cblas_sgemm_batch

CpuRuntimeObj->NativeCpuRuntimeObj

add  matmul op for mkl

* fix: fix bugs when rebasing from master

fix: fix bugs when rebasing from master

* fix: update api after rebasing

* fix: fix format; fix onnx import

* fix: fix clang-format

* [fix] fix conv_transpose test

* [fix] use stronger test case for transposed conv

* [fix] remove tensor memory format; fix mkl transpose conv

* [fix] add FIXME tag for op_timer python api

---------

Co-authored-by: whjthu <haojie0429@gmail.com>
2023-03-27 21:28:49 +08:00
Haojie Wang 65a3abf5dc
feat: inference (#71)
导出推理接口,支持通过 python 调用框架推理
2023-03-25 12:09:22 +08:00
whjthu d9886e9de3 fix: remove inline keyword in class; rename getter and setter for inputOf and outputOf 2023-03-25 12:04:24 +08:00
YdrMaster aff2b538ce fix: 删除单独的拷贝函数
Signed-off-by: YdrMaster <ydrml@hotmail.com>
2023-03-22 10:13:06 +08:00
wanghailu 64a5de51f3 fix 2023-03-22 10:08:31 +08:00
YdrMaster 5aeacedab3 fix: 从模板导出每个类型的 python 接口
Signed-off-by: YdrMaster <ydrml@hotmail.com>
2023-03-22 09:46:40 +08:00
YdrMaster 73e895b8ce feat: 导出拷出张量值方法
Signed-off-by: YdrMaster <ydrml@hotmail.com>
2023-03-21 14:40:13 +08:00
YdrMaster 9db97eb212 refactor: 整合操作张量数据的方法
Signed-off-by: YdrMaster <ydrml@hotmail.com>
2023-03-21 14:00:04 +08:00
YdrMaster e1c976568d fix: 增加推理接口
Signed-off-by: YdrMaster <ydrml@hotmail.com>
2023-03-15 17:23:32 +08:00
YdrMaster c18845a2fd feat: 增加推理接口
Signed-off-by: YdrMaster <ydrml@hotmail.com>
2023-03-15 17:23:32 +08:00
YdrMaster 6e1af09dd0 fix: remove print
Signed-off-by: YdrMaster <ydrml@hotmail.com>
2023-03-15 17:23:32 +08:00