Commit Graph

5860 Commits

Author SHA1 Message Date
wsq3 5a268a5345 add argminwithvalue oparetor arithmetric 2021-07-20 20:01:18 +08:00
i-robot a7d40fc220 !20520 [AutoParallel]Add op AllToAllv
Merge pull request !20520 from lichen/add_op_AllToAllv
2021-07-20 07:57:03 +00:00
i-robot ed4c9682b5 !20505 limit the scope of lift free variable
Merge pull request !20505 from xychow/limit-lift-scope
2021-07-20 06:10:17 +00:00
lichenever 8c1998fd6b add_op_AllToAllv 2021-07-20 10:49:28 +08:00
i-robot c9d3c1d346 !20411 enable optimizer parallel for inference
Merge pull request !20411 from gziyan/enable_opt_shard_predict
2021-07-20 01:37:06 +00:00
i-robot 6061194083 !20490 update check strategy for conv2d
Merge pull request !20490 from yangzhenzhang/update-check-strategy-for-conv2d
2021-07-20 01:23:08 +00:00
i-robot 71e8af3800 !18540 Gpu kernel for TensorScatterMax/Min/Mul/Div and their grad operators
Merge pull request !18540 from Peilin/tensor-scatter-mul/div/max/min-gpu-kernel
2021-07-20 01:17:09 +00:00
i-robot a1d952b249 !19300 [GraphKernel] add ops sin, cos, asin, acos, sign, squared_difference
Merge pull request !19300 from chenlei_autodiff/add_sin
2021-07-20 01:07:06 +00:00
zhousiyi 597f29ea7d limit the lambda lift scope to the func_graph passed 2021-07-20 00:37:21 +00:00
chenlei_autodiff 7d55cef106 [GraphKernel] add sponge ops. 2021-07-19 19:34:23 +08:00
zhousiyi b590f6d929 don't replace u with fprop_u in primal_graph and bprop_fg 2021-07-19 08:14:40 +00:00
i-robot fd72de08d9 !20174 Send compilation attrs to akg
Merge pull request !20174 from DeshiChen/0708_compilewithjson
2021-07-19 08:13:36 +00:00
yangzhenzhang b31cd27a08 update check strategy for conv2d 2021-07-19 15:03:26 +08:00
i-robot 2b0de5af92 !20339 The certificate can be configured through the configuration file
Merge pull request !20339 from anancds/cert
2021-07-17 08:38:32 +00:00
i-robot 3ccf3a9305 !20386 Add a st for a kind of loss-grad usage
Merge pull request !20386 from jxlang910/master
2021-07-17 07:16:31 +00:00
chendongsheng dc3d6dc915 added cert 2021-07-17 10:14:28 +08:00
Peilin Wang 0824f13ea6 initial commit: change of requirements, no more div/mul, need sub 2021-07-16 20:04:00 -04:00
Ziyan 1c9166e0a6 remove restriction for opt shard in inference 2021-07-16 17:16:12 +08:00
dayschan 3ab53dd26d Send compilation attrs to akg
1. Add a new message type "AKG/ATTR" in AkgKernelBuilder.
   the attrs was sent before the kernel infos.
2. Send "online_tuning" attribute when the flag is not zero,
   but error occurs in the latest akg submodule.
3. Send "repository_path" attribute when the flag is not empty.
4. Add a new value "compute_capability" into kernel info when the processor is GPU.
2021-07-16 16:52:18 +08:00
i-robot 3d022c9db8 !20246 Print ascend warning&error message using ErrorManager
Merge pull request !20246 from tanghuikang/error_manager
2021-07-16 06:49:31 +00:00
jin-xiulang bedb4956e5 Add a st for inversion attack 2021-07-16 10:09:57 +08:00
i-robot 51cd4215be !20120 InTopK gpu kernel bugfix
Merge pull request !20120 from Peilin/topk-bugfix-index
2021-07-15 11:16:56 +00:00
i-robot eaf9588ac9 !20111 add adam offload for pangu & fix AdamWeightDecay nnacl
Merge pull request !20111 from zhaosida/zsd_adam_simd
2021-07-15 08:53:37 +00:00
i-robot 047765a91f !20305 New TBE compile server
Merge pull request !20305 from laiyongqiang/new_python
2021-07-15 08:21:55 +00:00
tanghuikang 21c4bf1f27 Print ascend warning and error message using ErrorManager 2021-07-15 11:52:25 +08:00
LaiYongqiang ba8fdcfeae new TBE compile server 2021-07-15 10:56:02 +08:00
i-robot e67b74e8e3 !20137 Optimize round kernel's return code.
Merge pull request !20137 from ZPaC/optimize-return-code
2021-07-14 07:50:16 +00:00
zhaosida f323903af3 fix AdamWeightDecay nnacl & add adam offload in pangu scripts 2021-07-14 14:46:24 +08:00
i-robot fac767598e !16220 [Auto parallel] Add the function of sharding propagation
Merge pull request !16220 from Xiaoda/64-make-strategy-follow-in-auto-parallel
2021-07-14 06:28:06 +00:00
i-robot fd8eda24c4 !19904 fix fusion type
Merge pull request !19904 from laiyongqiang/fusionType
2021-07-14 06:08:20 +00:00
Xiaoda Zhang 04381273b3 Add the sharding propagation function:
1) users configure sharding strategies for operators;
2) framework will propagate the strategies from configured-ops to
non-configured ops using BFS;
3) the propagation goal is to minimize redistribution communication
cost;
2021-07-14 10:54:18 +08:00
Peilin Wang b0c8734064 initial commit: fix topk index
k <= 0 is now valid behavior, k > input size is optimized

fix ci
2021-07-13 15:19:26 -04:00
i-robot 0f6d55289f !20110 fix the test case of CPU dump
Merge pull request !20110 from limingqi107/bug_fix
2021-07-13 15:57:44 +00:00
ZPaC c46f4ac8d1 Optimize round kernel's return code. 2021-07-13 20:53:04 +08:00
i-robot 77cc850c1a !19882 change kMaliGPU to kGPU
Merge pull request !19882 from yeyunpeng2020/new_api
2021-07-13 07:08:06 +00:00
LaiYongqiang 898a48d6ea fix fusion type 2021-07-13 10:44:50 +08:00
limingqi107 06a6e8d186 fix the test case of CPU dump 2021-07-12 22:58:34 +08:00
i-robot 344de3f14e !20015 add yolov3_darknet53 test case
Merge pull request !20015 from 吴书全/PR0316
2021-07-12 13:39:36 +00:00
i-robot 884f5ae1de !19764 [MD] Remove Builder class phase 3
Merge pull request !19764 from harshvardhangupta/rem-build-2
2021-07-12 13:25:34 +00:00
i-robot e81aaf5043 !19936 Optimize signal util for ctrl+c
Merge pull request !19936 from VectorSL/signal
2021-07-12 11:58:09 +00:00
i-robot 21f36f1610 !20030 Fix issues of mindspore federated
Merge pull request !20030 from jxlang910/master
2021-07-12 08:09:10 +00:00
i-robot 4902090b65 !19260 optimize resnet50+imagenet2012 by prefetch_size and num_parallel_workers.
Merge pull request !19260 from anzhengqi/optimize_resnet_imagenet_network
2021-07-12 06:27:11 +00:00
vector ee057e57ec update signal 2021-07-12 11:24:49 +08:00
jin-xiulang 33de9e66d3 Fix issues of mindspore federated 2021-07-12 11:14:31 +08:00
yeyunpeng2020 f545bcd9b0 new api use gpu 2021-07-12 11:00:55 +08:00
wsq3 631a20d74a add test_yolov3_darknet53 test case 2021-07-12 09:51:08 +08:00
i-robot cb555f4b6e !19494 add ResizeBilinear ops for aicpu
Merge pull request !19494 from yanzhenxiang2020/add_resize_bilinear_aicpu
2021-07-10 08:46:27 +00:00
i-robot 53633ebfd8 !19748 Fix param check and default value of center in Rotate and RandomRotation Ops
Merge pull request !19748 from xiaotianci/fix_rotate_center
2021-07-10 08:07:00 +00:00
i-robot f10786f063 !19913 Update loss value in bert precision test case
Merge pull request !19913 from chenhaozhe/code_docs_update_bert_loss_value
2021-07-10 07:24:55 +00:00
i-robot 6dd5965e17 !19892 topk and intopk float16 bug workaround
Merge pull request !19892 from Peilin/topk-float16-cast
2021-07-10 07:04:55 +00:00