Commit Graph

27157 Commits

Author SHA1 Message Date
i-robot 7703c631b3 !20025 codex & support windows
Merge pull request !20025 from lanzhineng/r1.3
2021-07-12 13:15:54 +00:00
i-robot e56fb99d67 !19890 MD CI code warning fixes [r1.3]
Merge pull request !19890 from cathwong/ckw_r1.3_ci_q3_codedex1
2021-07-12 12:56:25 +00:00
i-robot c3cb98fca6 !20102 optimize Resnet training speed
Merge pull request !20102 from zhouneng/code_docs_fix_issue_I3W1H1_r1.3
2021-07-12 12:54:06 +00:00
zhouneng2 8d7e3b216c optimize training speed 2021-07-12 20:28:16 +08:00
i-robot 93ae1bffcd !20048 [MS][RDR] add trigger function of RDR in CPU backend
Merge pull request !20048 from louie5/r1.3
2021-07-12 11:50:10 +00:00
i-robot be411bed0d !20079 update the example of batchnorm and dropout operators.
Merge pull request !20079 from wangshuide/code_docs_wsd_r1.3
2021-07-12 11:48:11 +00:00
i-robot 03c744371a !20033 fix numpy docstrings
Merge pull request !20033 from wangrao124/code_docs_numpy_0712_r1.3
2021-07-12 11:32:25 +00:00
i-robot e271af104b !20089 Fix resnet multi server accuracy.
Merge pull request !20089 from linqingke/code_docs_r1.3
2021-07-12 11:30:08 +00:00
i-robot 36c7d685a1 !20054 Fix the output format problem in the example of Flatten and OneHot operators.
Merge pull request !20054 from dinglinhe/code_docs_dlh_r1.3_I4023Z
2021-07-12 11:24:10 +00:00
i-robot 8f3a3222e0 !20006 Fix an issue of input type check
Merge pull request !20006 from jxlang910/push-tor1.3
2021-07-12 11:22:10 +00:00
i-robot a0f24b6f96 !20082 delete note info in comment of lr_callback
Merge pull request !20082 from wangnan39/code_docs_delete_note_at_lr_callback
2021-07-12 11:21:27 +00:00
i-robot be32cf94c8 !20028 [MS][LITE] Add build function for model path
Merge pull request !20028 from cjh9368/r1.3
2021-07-12 11:02:47 +00:00
i-robot f799fa80ea !20017 fix type bug in fp16 PReLU grad
Merge pull request !20017 from zhangbuxue/fix_type_bug_in_fp16_PReLU_grad_r1.3
2021-07-12 09:54:53 +00:00
i-robot 300e23e573 !19799 modify hccl dlopen
Merge pull request !19799 from changzherui/mod_dlopen_1.3
2021-07-12 09:52:47 +00:00
王南 3b721532df delete note in lr_callback 2021-07-12 17:37:07 +08:00
wangshuide2020 12428e1957 update the example of batchnorm and dropout operators. 2021-07-12 17:27:42 +08:00
zhangzhenghai fe9809b03f !20060 run0710
Merge pull request !20060 from run_0710
2021-07-12 09:13:10 +00:00
i-robot b9fc89f032 !20068 Add Device Matrix from 1024 to 4096
Merge pull request !20068 from huangxinjing/code_docs_fix_device_matrix
2021-07-12 09:00:53 +00:00
i-robot 672883b785 !20065 Fix the file format problem of the operator
Merge pull request !20065 from dinglinhe/code_docs_dlh_r1.3_I408LD
2021-07-12 08:36:47 +00:00
linqingke dc5e10c02a fix resnet multi-rank accuracy. 2021-07-12 16:31:27 +08:00
huangxinjing e9a50de9fe Change Device Number to 4096 2021-07-12 16:19:29 +08:00
i-robot 27b0debd23 !20039 Add Ascend Memory For PanGuModel
Merge pull request !20039 from huangxinjing/code_docs_add_mem_for_pangu
2021-07-12 08:15:10 +00:00
dinglinhe 46bf65408f fix the docs problems of some operators 2021-07-12 16:10:52 +08:00
i-robot e21781ce55 !19915 fix_pipeline_opt_shard
Merge pull request !19915 from yao_yf/fix_pipeline_opt_shard
2021-07-12 08:05:14 +00:00
dingpeifei e772f5afcd upgrade_run 2021-07-12 16:01:42 +08:00
i-robot d6e3de5f93 !20037 r1.3 Export header files of lite_cv when MSLITE_ENABLE_TRAIN=on
Merge pull request !20037 from luoyang/son_r1.3
2021-07-12 07:45:12 +00:00
dinglinhe 69575e8ea5 fix the wrong output in the Flatten and OneHot 2021-07-12 15:38:19 +08:00
i-robot ec052373d1 !20018 [MSLITE][Develop] fix bug of tensor_rt
Merge pull request !20018 from yangruoqi713/r1.3
2021-07-12 07:17:27 +00:00
i-robot 810f424bfd !20012 [MS][LITE][CPU]fill fp16算子优化
Merge pull request !20012 from liuzhongkai/fill2
2021-07-12 07:15:10 +00:00
i-robot 54c93a227f !20003 [lite]1.3 add header file for converter
Merge pull request !20003 from 徐安越/r1.3
2021-07-12 07:13:12 +00:00
i-robot 5a1dbac549 !20009 move format.h to api
Merge pull request !20009 from zhoufeng/move-format-h-file-1.3
2021-07-12 07:11:15 +00:00
louie5 2081fbacb5 addding record kernel graph & graph execute orders in RDR
adding trigger function of RDR in CPU backend pipeline
2021-07-12 14:27:21 +08:00
lanzhineng f14cdf5a16 infer_opt support windows & codex 2021-07-12 14:22:03 +08:00
zhangzhenghai 30f19265ee !20036 upgrade_ascend_0712_mindspore
Merge pull request !20036 from upgrade_ascend_0712
2021-07-12 06:07:06 +00:00
i-robot 3bf1381e4b !19926 [MS][LITE]Fix matmul and convolution delegate
Merge pull request !19926 from gongdaguo/r1.3_fix_matmul
2021-07-12 05:05:08 +00:00
cjh9368 88b670611e [MS][LITE] Add build function for model_path 2021-07-12 12:53:00 +08:00
jin-xiulang 03328ee86d Fix an issue of input type check 2021-07-12 12:07:40 +08:00
xuanyue c7543a437e fix nnie compile 2021-07-12 11:47:13 +08:00
i-robot 1300fbd697 !20023 [lite]1.3 add header files for converter
Merge pull request !20023 from 徐安越/code_docs_r1.3
2021-07-12 03:37:55 +00:00
huangxinjing 3211e527ad Add ascend memory for pangu model 2021-07-12 11:35:27 +08:00
dingpeifei 470c16af1d upgrade_ascend_0712_mindspore 2021-07-12 11:35:26 +08:00
YangLuo d0491237b4 Export header files of lite_cv when MSLITE_ENABLE_TRAIN=on 2021-07-12 11:35:11 +08:00
lzk 1d076fd311 fp16 op 2021-07-11 20:32:39 -07:00
huangmengxi f7618f422d fix numpy docstrings 2021-07-12 11:06:37 +08:00
i-robot 5491b2c982 !19921 [MS][LITE][Develop] lite nnie release package
Merge pull request !19921 from sunsuodong/lite_release_1.3
2021-07-12 02:56:00 +00:00
yao_yf bb94a9e811 fix_pipeline_opt_shard 2021-07-12 10:47:08 +08:00
i-robot 7a49005969 !19838 tensor setitem bug fix
Merge pull request !19838 from huangmengxi/setitem_pr
2021-07-12 02:02:49 +00:00
i-robot 88a84a6fbb !19843 fix naming collision for gpu op cumsum and cumprod [r1.3]
Merge pull request !19843 from 杨林枫/fix_cumsum_cumprod_naming_collision_r1.3
2021-07-12 02:00:50 +00:00
buxue 156afadb28 fix type bug in fp16 PReLU grad 2021-07-12 09:58:17 +08:00
yangruoqi713 492940d6e2 [MSLITE][Develop] fix bug of tensor_rt 2021-07-12 09:58:15 +08:00