Commit Graph

45876 Commits

Author SHA1 Message Date
i-robot 1089d55dd9
!35216 julia testcase only support x86_64 now
Merge pull request !35216 from r1chardf1d0/julia4
2022-05-31 18:05:59 +00:00
i-robot 25a85b7f4e
!35139 Add embedding cache scheduler
Merge pull request !35139 from zyli2020/embedding_cache_unify_runtime
2022-05-31 16:49:24 +00:00
i-robot d59bcd363b
!35246 gpu prioirty replay buffer cuda kernel
Merge pull request !35246 from chenweifeng/gpu-prb-commit
2022-05-31 15:04:25 +00:00
Lai Yongqiang eed0fb92ed using FuncGraphPtr instead of KernelGraphPtr in device context 2022-05-31 22:26:42 +08:00
i-robot 6ae9129677
!35023 support ScatterUpdate in functional, tbe register, bprop, vmap
Merge pull request !35023 from zhengzuohe/scatterupdate
2022-05-31 12:47:43 +00:00
hezhenhao1 d5d0930a03 Replace thrust with cub for NonZero operator. 2022-05-31 20:40:56 +08:00
albert-yan f4122956ab fix op batch_to_space int8 2022-05-31 20:22:00 +08:00
i-robot 70f75f7e05
!35077 [MS][LITE][parallel] support big model: read tensor data for build by buf
Merge pull request !35077 from yefeng/328-copy_311_big_model
2022-05-31 11:35:07 +00:00
zhangyinxia e21c5c06b9 fix bug 2022-05-31 19:23:05 +08:00
i-robot 21bcf23bbd
!35215 Update svd api docs
Merge pull request !35215 from huanghui/code_docs-update-svd-api
2022-05-31 11:17:54 +00:00
liangzelang f350e1dd9f fix ci cost time bug 2022-05-31 19:17:23 +08:00
Yang Jiao 7bd719390e auto tensor inplace 2022-05-31 19:09:00 +08:00
gaoyong10 ee1e15eaca eliminate nopnode. 2022-05-31 18:59:41 +08:00
yefeng 830b9d6dc0 parallel predict bug fix: core list && queue 2022-05-31 18:51:42 +08:00
i-robot fb4272b910
!35186 support debug cuda code
Merge pull request !35186 from chenweifeng/support-cuda-gdb
2022-05-31 10:47:12 +00:00
i-robot 47b02897bf
!34955 Duplicate Tensor id
Merge pull request !34955 from caifubi/master-tensor-id-duplicate
2022-05-31 10:43:26 +00:00
yangzhenzhang 3df7174d25 modify gen strategy for parallel op 2022-05-31 18:41:42 +08:00
i-robot 753bf25217
!34847 change GRU with seq_length to CudnnGRU ops
Merge pull request !34847 from 吕昱峰(Nate.River)/rnn
2022-05-31 10:21:12 +00:00
liutongtong d5460cc34e fix fit and callbacks 2022-05-31 18:08:23 +08:00
qkeys e1f26b0bfd [feat][assistant][I4XJG9] add new Ascend operator Rad2Deg 2022-05-31 17:57:25 +08:00
wilfChen 0110523efc priority replay buffer cuda kernel 2022-05-31 17:48:14 +08:00
i-robot 941a58df6e
!33504 Fix the summary datagraph show at the end of training
Merge pull request !33504 from maning202007/master
2022-05-31 09:37:11 +00:00
ling a71de92483 unsorted segment arithmetic check bug 2022-05-31 17:37:03 +08:00
i-robot 4aaf6f8df6
!34981 Add dynamic and vmap support for `GatherD`
Merge pull request !34981 from TronZhang/gatherd_implement_more
2022-05-31 09:32:30 +00:00
i-robot 29f521a7bb
!35160 [MSLITE] Support PreLUFusion TensorRT Op
Merge pull request !35160 from zhangyongxian/dev_zhangyongxian_prelu
2022-05-31 09:26:34 +00:00
cjh9368 4b02866eef extract ema param from quant.py to golden_stick layer_policy 2022-05-31 17:25:52 +08:00
i-robot dcecfa5a1e
!35258 modify white list
Merge pull request !35258 from 冯一航/code_docs_modify_checkparam
2022-05-31 09:24:54 +00:00
i-robot d8547cc27e
!35232 modify format
Merge pull request !35232 from 俞涵/code_docs_0531
2022-05-31 09:15:11 +00:00
i-robot bc52caaa1d
!35170 [MS][LITE] fix python api codex
Merge pull request !35170 from luoyuan/fix-codex-0530
2022-05-31 09:12:57 +00:00
i-robot e46f47b5e8
!35088 [MS][LITE] fix server inference run testcase no glog so
Merge pull request !35088 from luoyuan/fix-server-infer-st-no-glog
2022-05-31 09:12:05 +00:00
chujinjin eecbb9060a add st for adam weight decay op 2022-05-31 17:06:55 +08:00
fengyihang b13dcc0dd7 modify_checkparam 2022-05-31 16:58:16 +08:00
i-robot 588c8fd928
!35219 [AutoParallel]Parallel support subgraph redistribution
Merge pull request !35219 from lichen/parallel_support_subgraph_redistribution
2022-05-31 08:57:39 +00:00
liqiliang d6ebf39ab2 Refactor the Activation and ActivationGrad for GPU. 2022-05-31 16:53:03 +08:00
yefeng 368aa3a3d0 support big model: read tensor data for build by buf 2022-05-31 16:38:52 +08:00
i-robot df9b900c04
!35148 Delete the definition of ControlDepend
Merge pull request !35148 from huangbingjian/rm_definition
2022-05-31 08:35:54 +00:00
i-robot 3c07413969
!35156 Return true when copy size is 0
Merge pull request !35156 from liangzelang/dev_master
2022-05-31 08:25:44 +00:00
i-robot 2eefb34afd
!35180 Add vmap and it's tests.
Merge pull request !35180 from liqiliang/vmap_and_dynamic_and_tests
2022-05-31 08:14:15 +00:00
zhujingxuan b92c902a2f Add Linspace functional API && enable batch support && float64 support 2022-05-31 16:08:03 +08:00
zhujingxuan ef2006780e Remove sstream for logging 2022-05-31 15:59:40 +08:00
twc 8e1e4a93cb fix codecheck 2022-05-31 15:58:50 +08:00
zjun dcf58cd750 Add auto dynamic shape check
Signed-off-by: zjun <zhangjun0@huawei.com>
2022-05-31 15:52:46 +08:00
i-robot 8268d5ef88
!35242 add owners
Merge pull request !35242 from caifubi/code_docs_owners
2022-05-31 07:43:03 +00:00
xiongkun a57c4f4013 add clip by value dynamic shape
add clip by value dynamic shape

reconstruct bitwise

reconstruct bitwise

reconstruct bitwise

add fastgelu grad shape

add clip by value c primitive

add clip by value c primitive

clip by value shape

tick bitwise

tick bitwise

tick bitwise

tick bitwise

tick bitwise

tick bitwise

tick bitwise

tick bitwise
2022-05-31 15:41:58 +08:00
caifubi c2ce8377c6 add owners 2022-05-31 15:37:06 +08:00
huodagu 797cf1ef1b modify format 2022-05-31 15:35:09 +08:00
xulei 9b0cff1bd4 op tiling for KLDivLoss and adapt vmap st 2022-05-31 15:23:31 +08:00
yao_yf 4741af1639 fix independent stream error and opt pipeline parallel performance 2022-05-31 15:21:07 +08:00
i-robot ad04b46aa9
!34954 add steal thread pool
Merge pull request !34954 from fangzehua/steal_thread
2022-05-31 07:09:29 +00:00
wangpingan2 1ce54b9b7f kernel_executor update called funtions and fix bug. 2022-05-31 15:02:53 +08:00