Commit Graph

Select branches

Hide Pull Requests

Conv_NHWC

NNET_231111

NNET_231111_from_master

NNET_GAN

NNET_OpSearch

NNET_anyOp

NNET_bias

NNET_bias_0630

NNET_e2e

NNET_e2e_fix

NNET_e2e_for_merge

NNET_eliminateOP

NNET_gcn

NNET_gcn_fuse

NNET_op_test

NNET_transpose

TC_revision

activation

add_paddle_model

allocator_memPool

ascend

bang-softmax

benchmark

benchmark_conv

benchmark_softmax

broadcast_wanghailu_0916

case-fsrcnn

change_path

constroy/doc_on_ares

conv_half

cpu_backend2

cuda-attention

cuda-transpose

dcj/for_multiple_datatype

ddp

dev-dynamic-graph

dev-dynamic-graph-allocator

dev-leakyrelu

dev-memory

dist/graph

dist_bench

dropout

dump/init

fsrcnn-conv-bias-act-fuse

gencode

gpt

graph-onnx

graphFactory

instance_norm

kunlun_temp

kvcache_attention_fp16

kvcache_backup

master

model_test

nnet_e2e_for_merge

op_timer

optimization-pass

point2point

power-fusion

pure_engine

search_engine

support_fp16

testAccuracy

test_codegen

test_onnx

train_wanghailu_1010

update_doc

update_pybind11

v0630

xpu_allreduce

test-models

Mono Color

• 6aaf963f50 Fix: cuBLAS failure when codegen is enabled Liyan Zheng 2023-03-29 12:52:04 +0800
• 6441abb88f Fix: GraphObj ctor should use cloneTensor Liyan Zheng 2023-03-14 21:32:15 +0800
• 220b76f355 Add: matmul broadcast huangshuhong 2023-03-14 20:17:16 +0800
• aac4bd4650 Fix: [Intermediate] NMutator::expressionToGraph Liyan Zheng 2023-03-12 21:51:39 +0800
• 7478b698dc Add: expression reader as a cmake target Liyan Zheng 2023-03-06 20:33:39 +0800
• 8e6fc4136e Add: cache and padding (bugs!!) huangshuhong 2023-03-05 22:50:31 +0800
• bd12b67ea3 Add: mutator InfoGAN minimum test Liyan Zheng 2023-03-03 20:58:29 +0800
• f2be122efc change Device::MKL to Device::INTELCPU wangw 2022-11-25 17:55:23 +0800
• fe1afe38fa
  fix code of bang conv (#76) Hardy 2023-03-29 15:47:32 +0800
• 823e66a9ff
  Support perf bang 1115 (#57) Hardy 2023-03-29 13:52:56 +0800
• 86ec4036ce
  ADD: add mkl runtime for intel cpu , and add mkl kernel for matmul/conv/convtransposed. (#61) wendy12022 2023-03-27 21:28:49 +0800
• 65a3abf5dc
  feat: inference (#71) Haojie Wang 2023-03-25 12:09:22 +0800
• d9886e9de3 fix: remove inline keyword in class; rename getter and setter for inputOf and outputOf whjthu 2023-03-25 12:04:24 +0800
• fb2a6a8fb2 add transpose operation activation wanghailu 2023-03-24 16:42:45 +0800
• aebd7440aa add lrn operation wanghailu 2023-03-23 07:59:19 +0000
• 8b58b08240 add lrn operation wanghailu 2023-03-23 15:31:42 +0800
• aff2b538ce fix: 删除单独的拷贝函数 YdrMaster 2023-03-22 10:13:06 +0800
• 64a5de51f3 fix wanghailu 2023-02-27 10:22:03 +0800
• 5aeacedab3 fix: 从模板导出每个类型的 python 接口 YdrMaster 2023-03-22 09:46:40 +0800
• 73e895b8ce feat: 导出拷出张量值方法 YdrMaster 2023-03-21 14:40:13 +0800
• 9db97eb212 refactor: 整合操作张量数据的方法 YdrMaster 2023-03-21 14:00:04 +0800
• e1c976568d fix: 增加推理接口 YdrMaster 2023-02-27 10:13:55 +0800
• c18845a2fd feat: 增加推理接口 YdrMaster 2023-02-27 10:03:06 +0800
• 6e1af09dd0 fix: remove print YdrMaster 2023-02-27 09:35:17 +0800
• e294e46436 feat: 导出 pool 到 onnx YdrMaster 2023-02-27 09:32:08 +0800
• 8a871c3773 feat: 导出 conv 到 onnx YdrMaster 2023-02-27 09:15:45 +0800
• afed749b74 feat: 支持导出权重 YdrMaster 2023-02-24 16:57:37 +0800
• 40fb8390b1 feat: 导入时保存权重 YdrMaster 2023-02-24 16:49:53 +0800
• a5e692baea feat: 导出 batchnorm 到 onnx YdrMaster 2023-02-24 15:42:39 +0800
• 71ca4459d9 fmt YdrMaster 2023-02-24 15:04:31 +0800
• 5b6698bac7 feat: 导出全图的输出张量到 onnx YdrMaster 2023-02-24 15:02:52 +0800
• 59bf59c10b docs: update README.md YdrMaster 2023-02-24 14:28:34 +0800
• fb3478bf3e build: update Makefile YdrMaster 2023-02-24 14:14:52 +0800
• 514666591e add batch_norm wanghailu 2023-02-24 13:55:53 +0800
• 3d122aebfe feat: 支持导出浮点向量 YdrMaster 2023-02-24 11:19:37 +0800
• cf9bdb0562 feat: 支持打印结果 YdrMaster 2023-02-24 10:08:04 +0800
• f44a4daf70 feat: 导出未初始化的张量 YdrMaster 2023-02-24 09:39:30 +0800
• 6dce129cb3 fix: TensorObj::dataMalloc YdrMaster 2023-02-24 09:22:03 +0800
• dc79b72655 fix: 重新导出 cuda_runtime() YdrMaster 2023-02-23 17:20:05 +0800
• 9ab78f13f7 feat: 导出 cuda_runtime YdrMaster 2023-02-23 16:29:20 +0800
• 60c5d6b5b8 fix: 先不在 cpu 上测试 YdrMaster 2023-02-23 16:22:03 +0800
• ed81861375 temp: 实现初始值导入，但 resnet 报错 YdrMaster 2023-02-23 15:29:16 +0800
• 4ffaa44c1e fix: Matmul 支持 2 维或以上的输入 YdrMaster 2023-02-23 11:51:46 +0800
• a27391fcdc fix: 修正 batchNorm 实现 YdrMaster 2023-02-23 11:48:28 +0800
• dd5d091dbc
  wip: onnx 导出 (#65) Haojie Wang 2023-03-15 15:22:09 +0800
• 62fd619987 fix: 移除中文注释 YdrMaster 2023-03-15 15:16:16 +0800
• 71a87c27d1 feat: 导出 ReduceMean 到 onnx YdrMaster 2023-02-22 11:26:48 +0800
• bb9b62b169 fix: 改正类型转换 YdrMaster 2023-02-22 11:05:54 +0800
• 2a23669394 feat: 导出 Reshape 到 onnx YdrMaster 2023-02-21 14:30:06 +0800
• ffd0473bd2 feat: check everything YdrMaster 2023-02-21 14:06:32 +0800
• 9e0f8f21bf feat: 生成模型对象 YdrMaster 2023-02-21 14:01:57 +0800
• 6b7af7077b feat: 导出 Gather Concat 到 onnx YdrMaster 2023-02-21 13:10:54 +0800
• 9d9fbd44af feat: 导出 MatMul Concat 到 onnx YdrMaster 2023-02-21 12:44:45 +0800
• 32f6f02c81 feat: 导出 5 个单目算子到 onnx YdrMaster 2023-02-20 14:38:13 +0800
• 0517089dca feat: 导出输入张量到 onnx YdrMaster 2023-02-20 14:15:54 +0800
• eff4c14a85 feat: 封装上下文对象以复用建图代码 YdrMaster 2023-02-20 12:12:58 +0800
• 0833a2f779 feat: 导出加减乘除幂到 onnx YdrMaster 2023-02-17 17:15:15 +0800
• f2591edbb4 feat: 导出 OpType，为节点命名 YdrMaster 2023-02-17 15:58:26 +0800
• fe81fccf76 feat: 导出 OperatorObj YdrMaster 2023-02-17 15:00:34 +0800
• 45a3cdfa30 feat: GraphObj 增加一个拓扑排序方法及其测试 YdrMaster 2023-02-17 13:21:10 +0800
• f20e791cf5 style: 修改 graph.h/graph.cc YdrMaster 2023-02-16 14:58:04 +0800
• 0f52d04882
  Merge branch 'master' into dev-onnx Haojie Wang 2023-03-15 14:52:03 +0800
• 978269162a fix: 移除 c++ 中的中文注释，python TODO 改 FIXME YdrMaster 2023-03-15 14:48:39 +0800
• 40d1b1c91b
  Add ConvTransposedNHWC (#67) ddp deathwings602 2023-03-01 14:15:02 +0800
• 6871fff02b feat: 导出分配内存和运行推理的接口 YdrMaster 2023-02-23 11:08:00 +0800
• d7e52054e6 fix: 修正 GlobalAveragePool 和 Reshape 导入 YdrMaster 2023-02-23 08:59:06 +0800
• 4c7fdf44c5 feat: 前端支持 Conv 及单元测试 YdrMaster 2023-02-22 15:05:44 +0800
• ce04177585 style: use __path__ to import YdrMaster 2023-02-21 09:17:34 +0800
• 6a4de807e6 style: remove non-ascii comments from cpp YdrMaster 2023-02-16 14:57:51 +0800
• c9fee3f667 feat: 前端支持 GlobalAveragePool 及单元测试 YdrMaster 2023-02-16 10:33:24 +0800
• 391b9d16c0 cleanup YdrMaster 2023-02-15 14:08:30 +0800
• afa90ec9c9 feat: 前端支持 gemm 及单元测试 YdrMaster 2023-02-15 13:20:34 +0800
• 315763a83a feat: 前端支持 pad 及单元测试 YdrMaster 2023-02-15 11:41:06 +0800
• 7893ae0cca opt: 优化 PadObj 和 SplitObj 构造器实现 YdrMaster 2023-02-15 11:28:49 +0800
• bb0e7540cc fix: revert ci yml YdrMaster 2023-02-15 07:45:21 +0800
• 8fae67b4b4 feat: 前端支持 slice 及单元测试 YdrMaster 2023-02-14 17:35:18 +0800
• f9d0076a86 opt: 优化 SliceObj 构造器实现 YdrMaster 2023-02-14 16:44:08 +0800
• 341cf1f943 feat: 前端支持 pool 及单元测试 YdrMaster 2023-02-14 16:26:47 +0800
• 62ceb78ae3 feat: 前端支持 reduceMean 及单元测试 YdrMaster 2023-02-14 15:35:01 +0800
• fb9d84dbb7 opt: 优化 ReduceMeanObj 构造器实现 YdrMaster 2023-02-14 15:14:28 +0800
• d11fb0ad5f feat: 前端支持 gather 及单元测试 YdrMaster 2023-02-14 14:16:01 +0800
• 45aa0237da feat: 前端支持 concat 及单元测试 YdrMaster 2023-02-14 13:42:35 +0800
• a7e58bd8d0 feat: 补充 DataType 类型 YdrMaster 2023-02-14 11:27:57 +0800
• d9e2953425 fix: 改正 reshap 导入 YdrMaster 2023-02-14 10:14:55 +0800
• 7626efbfa8 feat: 前端支持 reshape YdrMaster 2023-02-14 09:50:32 +0800
• ee0a562006 test: batchNorm 单元测试 YdrMaster 2023-02-14 08:54:58 +0800
• 26be533faa Add documentation for operators. whjthu 2023-02-13 22:48:20 +0800
• cca4d2a491 feat: 前端支持 batchNorm（无单元测试） YdrMaster 2023-02-13 17:15:35 +0800
• e194dd943b feat: 前端支持 flatten 及单元测试 YdrMaster 2023-02-13 13:50:07 +0800
• e4ec9c4230 feat: 前端支持 identity 及单元测试 YdrMaster 2023-02-13 12:13:01 +0800
• 7f0c8ebae3 feat: 前端支持 relu sigmoid tanh softmax abs 及单元测试 YdrMaster 2023-02-13 11:54:54 +0800
• 6e5beceadd feat: 增加 add sub mul div pow 前端 YdrMaster 2023-02-13 11:25:54 +0800
• 296fcc5aa0 feat: 创建 pyinfinitensor 前端 YdrMaster 2023-02-12 08:23:49 +0800
• c7ec9ee6e7
  Add search engine (#64) zhengly123 2023-02-12 18:27:52 +0800
• 14c9c82dab
  test: enhance ci (#62) YdrMaster 2023-02-12 00:01:36 +0800
• 44594f13d3 add net test wanghailu 2023-02-09 05:35:05 +0000
• cc4eb31265 add bitcompute operation wanghailu 2023-02-01 05:49:03 +0000
• 27af0065f7 add arange operation wanghailu 2023-01-31 05:33:40 +0000
• 1fcab531ec add addcdiv and addcmul operation wanghailu 2023-01-30 06:36:36 +0000
• c51b19b198 add logic operation wanghailu 2023-01-29 06:00:12 +0000