313762b8dc
check validation of input shape before running graphes.
4 years ago
7ee1662420
fix doc issues in bijector classes
4 years ago
a51465c78b
add graphkerneloptimize pass
...
align fuse_ops_fusion
align composite_ops_fusion
unify ops table
Init new_code's kernel_info with orig_node's kernel_info in function NewCNodeWithInfo
enable run bert
add pass tensor_promotion
add macro for bias_add and bias_add_grad in expander pass
exclude unused attrs in primitive compare for GraphKernelCSE
exclude fusion_type in kernelinfo cmp for cse in graphkernel
check processor
remove graph kernel pass before select kernel
recover run_standalone_pretrain_ascend.sh
remove is_before_kernel_select
move add_atomic_clean from pass directory to graph_kernel directory
update fuse op list in Ascend back-end
4 years ago
5026ef1d46
load tensors for all graphs
4 years ago
a3066105d5
!7081 Mindspore debugger supporting multigraph
...
From: @lichen_101010
Reviewed-by:
Signed-off-by:
4 years ago
50efdcc2df
small minddata lite for train task
4 years ago
1b6265fa43
Debugger multi-graph support implementation
...
Other Contributor: Adel Shafiei, John Tzanakakis
4 years ago
7e72605fc8
make gaps not lifelong: constraints from neighbor
4 years ago
b8821bb2f3
handle getnext output tensor as normal lifelong tensor
4 years ago
7224880b72
!8571 revert pynative getnext
...
From: @chujinjin
Reviewed-by: @kisnwang,@limingqi107
Signed-off-by: @limingqi107
4 years ago
222159599a
independent node use the somas's memory
4 years ago
48e94a1c5b
fix unique infershape bug with data type int64
4 years ago
60167dae5b
add unique dynamic
4 years ago
98531ebcda
!8513 Add API info about Cases not supported by Windows
...
From: @shenwei41
Reviewed-by: @heleiwang
Signed-off-by:
4 years ago
42c3830938
remove lifelong for communication output
4 years ago
589dd9e782
fix codex
4 years ago
7e67c247b8
Add cell dynamic check
...
Signed-off-by: zjun <zhangjun0@huawei.com>
4 years ago
95ce3eab08
!8591 gatherv2 dynamic shape
...
From: @jonwe
Reviewed-by: @tom__chen,@robingrosman
Signed-off-by: @robingrosman
4 years ago
ecb4a03012
!8593 add all 144 kernels for cast gpu op
...
From: @jonwe
Reviewed-by: @tom__chen,@robingrosman
Signed-off-by: @robingrosman
4 years ago
d697ef58ad
!8539 Fix doc and add check for P.UniformCandidateSampler
...
From: @TFbunny
Reviewed-by: @robingrosman,@tom__chen
Signed-off-by: @tom__chen
4 years ago
009c0e9372
!8590 debugger should not read unused slots for BatchNorm
...
From: @john_tzanakakis
Reviewed-by: @mikef,@nsyca
Signed-off-by: @nsyca
4 years ago
e055522b05
cast 144 kernel
4 years ago
a3893d7d30
gatherv2 dynamic shape
4 years ago
941645c363
fix doc and add check for UniformCandidateSampler
4 years ago
d45ef4cacb
debugger won't read unused slots for BatchNorm
4 years ago
ffeacf13e4
add exoirt air test
4 years ago
1878a1b0f1
fix example error
4 years ago
90aaf8a58c
!8491 use input node as input parameter to allocate communication input address
...
From: @laiyongqiang
Reviewed-by:
Signed-off-by:
4 years ago
32f0b341a4
add conv conv fusion
4 years ago
a57fdc3b2b
add distribute training for gpu benchmark
4 years ago
e511e0d897
check outputFilePath for windows
4 years ago
2b4497a8c1
!8512 move compareOutput functions to /test
...
From: @lyvette
Reviewed-by:
Signed-off-by:
4 years ago
9169d0da6e
!8544 Extend hccl time out and modify lr schedule
...
From: @zhouyaqiang0
Reviewed-by: @oacjiewen,@c_34
Signed-off-by: @c_34
4 years ago
ecba84ae7c
!8447 export script for VGG16 and Bert，fix faster-rcnn and inceptionv3 bug
...
From: @yuzhenhua666
Reviewed-by:
Signed-off-by:
4 years ago
1a152a9663
!8529 fix error when cell construct return list output in pynative mode
...
From: @chujinjin
Reviewed-by: @kisnwang,@jjfeing
Signed-off-by: @jjfeing
4 years ago
189e0ed394
!8531 fix cache error for aicpu op
...
From: @chujinjin
Reviewed-by: @kisnwang,@jjfeing
Signed-off-by: @jjfeing
4 years ago
da11db2fed
!8562 [MSLITE]Add bitnum config in post quant
...
From: @guohonhzilonghw
Reviewed-by: @ddwsky,@HilbertDavid
Signed-off-by: @HilbertDavid
4 years ago
62f846c484
!8388 link fp16 & optimize to lite
...
From: @yeyunpeng2020
Reviewed-by:
Signed-off-by:
4 years ago
9e387e3b56
!8514 dataset_attr_name
...
From: @bai-yangfan
Reviewed-by:
Signed-off-by:
4 years ago
0c2c76d037
update get rank in parallel ops
4 years ago
8612885167
!8545 Save task debug info to ir
...
From: @laiyongqiang
Reviewed-by: @kisnwang,@jjfeing
Signed-off-by: @jjfeing
4 years ago
3c6b91f6fa
!8548 handle getNextOutput and contiguous conflict
...
From: @laiyongqiang
Reviewed-by: @kisnwang,@jjfeing
Signed-off-by: @jjfeing
4 years ago
c970f30db8
!8427 fix error when execute single op in Graph mode
...
From: @chujinjin
Reviewed-by: @kisnwang,@jjfeing
Signed-off-by: @jjfeing
4 years ago
2291b7f2e6
dynamic shape check
4 years ago
848fb9b554
!8430 [MD] transform dataset to mindrecord in preprocess of gpt
...
From: @liyong126
Reviewed-by:
Signed-off-by:
4 years ago
edcb0cd86b
!8370 [MD] Use neon instruction set to accelerate the Substract and Divide ops
...
From: @jiangzhiwen8
Reviewed-by: @xulei2020
Signed-off-by:
4 years ago
301b7039a8
fix communication input address bug
4 years ago
9c8828761d
!8568 add int32 kernel reg for sparse to dense
...
From: @zhaozhenlong
Reviewed-by: @zhang_xue_tong,@zhanghaibo5
Signed-off-by: @zhang_xue_tong
4 years ago
af34a8257b
recitify warpctc script
4 years ago
f494025db8
revert pynative getnext
4 years ago
wang_shaocong
Xun Deng
lingyunli63
yelihua
mindspore-ci-bot
xulei2020
lichen_101010
Ioannis Lamprou
laiyongqiang
caifubi
fangzehua
zjun
Jonathan Yan
TFbunny
John Tzanakakis
changzherui
liangchenghui
zhengjun10
VectorSL
liuyu
yangzhenzhang
wilfChen
gengdongjie
chujinjin