Nigel
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
7,488 Commits 24 Branches 0 Tags 3.9 GiB
Commit Graph

7488 Commits

Author SHA1 Message Date
lizhenyu d667d6ee92 bugfix:SigmoidCrossEntropyWithLogitsGrad need multiply dout 2020-08-17 17:00:43 +08:00
sunsuodong 676e44a130 fix_topkv2_parser 2020-08-17 17:00:32 +08:00
songhonglei413 b9e69b2714 add op_constantofshape and testcase 2020-08-17 16:54:36 +08:00
jonyguo 8bb4449fa8 add testcase for padded dataset with decode op 2020-08-17 16:43:23 +08:00
mindspore-ci-bot f41ca6b5c6 !4558 soft_dvpp_decode_resize_jpeg_op issue resolved, input size can be an int, break the constraints of input size being sequence (height, width) 
Merge pull request !4558 from magemomou/my/work
 2020-08-17 16:40:43 +08:00
hangq 4f886e8dcd fix java head 2020-08-17 16:40:25 +08:00
mindspore-ci-bot 490a515b1f !4583 [MS][LIte][DEV]fix transpose bug 
Merge pull request !4583 from 张学同/to_merge
 2020-08-17 16:29:28 +08:00
mindspore-ci-bot 6fd5323356 !4458 change wide&deep training batchsize 
Merge pull request !4458 from zyli2020/change_wide_and_deep_model_zoo_script
 2020-08-17 16:22:54 +08:00
mindspore-ci-bot 50e78118fb !4573 Fix do fission when UnsegmentSum's input0 shape size is 1 
Merge pull request !4573 from huanghui/unsorted-segment-sum-fission-pass
 2020-08-17 16:08:15 +08:00
zhangxuetong 67000aec29 fix bug 2020-08-17 16:00:05 +08:00
tao_yunhao fd9bd7a205 modify arm cpu op: Arithmetic_fp16 2020-08-17 15:45:24 +08:00
Zhang Qinghua 42218a1dc3 Fix the codedex issues and other issues. 2020-08-17 15:45:07 +08:00
mindspore-ci-bot c7f461acc7 !4549 fp16 concat 
Merge pull request !4549 from zhaozhenlong/lite/op/fp16/concat
 2020-08-17 15:38:07 +08:00
mindspore-ci-bot 094f259abc !4550 fp16 reduce mean 
Merge pull request !4550 from zhaozhenlong/lite/op/fp16/transpose_mean_concat
 2020-08-17 15:37:12 +08:00
mindspore-ci-bot 51a38c04e9 !4539 add op fp16 transpose 
Merge pull request !4539 from zhaozhenlong/lite/ops/fp16/transpose
 2020-08-17 15:36:28 +08:00
mindspore-ci-bot 2da8b5d9aa !4575 optimize fp16 conv3x3 post func 
Merge pull request !4575 from fuzhiye/tmp
 2020-08-17 15:32:29 +08:00
mindspore-ci-bot e2daa61072 !4514 refactor CreateNewCNode 
Merge pull request !4514 from Margaret_wangrui/refactor_CreateNewCNode
 2020-08-17 15:31:58 +08:00
mindspore-ci-bot d94a8e02ca !4454 fix int8/uint8 print bug 
Merge pull request !4454 from zjun/fix_int8_print
 2020-08-17 15:30:26 +08:00
mindspore-ci-bot 31ff088789 !4547 [MS][LITE]deconvolution fp16 post function 
Merge pull request !4547 from ling/deconv
 2020-08-17 15:29:50 +08:00
mindspore-ci-bot c9062e1428 !4574 fix post training quant 
Merge pull request !4574 from xutianchun/quant_st
 2020-08-17 15:22:06 +08:00
mindspore-ci-bot 58700369d3 !4353 refactor tflite parsers 
Merge pull request !4353 from lyvette/tflite_parser
 2020-08-17 15:13:01 +08:00
fuzhiye a2f05a8daa optimize fp16 conv3x3 post func 2020-08-17 15:05:35 +08:00
hanhuifeng2020 9e2b38c961 The teacher use fp16 calculations to optimize the performance of tinybert on the gpu 2020-08-17 15:01:01 +08:00
mayang c410c6a6e2 soft_dvpp_decode_resize_jpeg_op issue resolved, input size can be an int, break the constraints of input size being sequence (height, width) 2020-08-17 14:57:04 +08:00
xutianchun fcb7bf9b77 add post quant st 2020-08-17 14:56:48 +08:00
ling fb6b50e8c2 [MS][LITE]deconv post func 2020-08-17 14:52:41 +08:00
zhaozhenlong 5787dd642d reduce mean fp16 2020-08-17 14:41:22 +08:00
mindspore-ci-bot 5f27ff4afe !4556 [MS][LITE]add fp16 split 
Merge pull request !4556 from 张学同/to_merge
 2020-08-17 14:40:42 +08:00
mindspore-ci-bot 305e36a5e3 !4568 fix fp16 converter bug 
Merge pull request !4568 from cjh9368/fp16
 2020-08-17 14:38:29 +08:00
huanghui 71d839dec7 fix do fission when UnsegmentSum's input0 shape size is 1 2020-08-17 14:36:18 +08:00
mindspore-ci-bot a3b8b4c2d6 !4377 Supports weight slice while training with multi servers 
Merge pull request !4377 from ZPaC/master-multi-server-weight-slice
 2020-08-17 14:30:19 +08:00
huanghui 0d962a372f add ScatterNdUpdate cpu kernel 2020-08-17 14:14:26 +08:00
mindspore-ci-bot d341461baa !4510 GPU fix codex for replace_bn_grad_cast 
Merge pull request !4510 from VectorSL/fix-codex
 2020-08-17 14:10:27 +08:00
mindspore-ci-bot 1b28f77be4 !4198 New optimization pass to remove redundant Max ops 
Merge pull request !4198 from thlinh/dev_Aug08_eliminate_more_redundant_Max
 2020-08-17 12:48:36 +08:00
lyvette 123c2024a5 refactor tflite parsers. 
append ut

refactor tflite parsers

modify tflite parser, ut and model

supplement caffe flatten parser

fix the weight tensor format of deconv bug

fix bug when idx=-1

fix the weight tensor format of depthConv bug.
 2020-08-17 11:52:45 +08:00
cjh9368 8a29d90d3c solve fp16tofp32 bug 2020-08-17 11:50:43 +08:00
mindspore-ci-bot dda4c26e7c !4423 add reshape type of tensor 
Merge pull request !4423 from lianliguang/add-reshape-type-to-tensor
 2020-08-17 11:34:11 +08:00
mindspore-ci-bot de857161ed !4554 fix anf exporter problem 
Merge pull request !4554 from wangchangkai/master
 2020-08-17 11:27:54 +08:00
yelihua 31e61f71eb add data saver module 2020-08-17 11:18:40 +08:00
mindspore-ci-bot f20e68a879 !4565 fix post training quant 
Merge pull request !4565 from xutianchun/quant_0817
 2020-08-17 11:16:10 +08:00
mindspore-ci-bot 86ab206a00 !4466 akg add Chinese version readme 
Merge pull request !4466 from looop5/akg_readme
 2020-08-17 11:04:17 +08:00
ZPaC 348c67ac00 PS support multi server. 2020-08-17 11:01:40 +08:00
kai00 892668389f anf exporter fixed 2020-08-17 10:59:53 +08:00
xutianchun 3f68f575c4 fix post training quant 2020-08-17 10:57:44 +08:00
He Wei e80e19d063 Refactory Tensor bool type and ToString() 
1. Use C++ native bool type for tensor:
Since we changed tensor data from vector<T> to unique_ptr<T[]>,
boolean tensor can be implemented using C++ native bool type.

2. Improve ToString():
Try to make tensor string representation looks better.
 2020-08-17 10:48:52 +08:00
mindspore-ci-bot cde696477c !4560 [MS][LITE][Develop]unsqueeze support int32 
Merge pull request !4560 from chenjianping/lite_dev2
 2020-08-17 10:36:20 +08:00
zhangxuetong 96bc761754 add split 2020-08-17 10:26:35 +08:00
mindspore-ci-bot f384444487 !4482 add_field_in_strategy_ckpt 
Merge pull request !4482 from yao_yf/add_field_in_strategy_ckpt
 2020-08-17 10:11:40 +08:00
王明贵 32e4e2040f Add arm op Slice for int8 and testcases 2020-08-17 10:00:24 +08:00
chenjianping baf64e498e unsqueeze support int32 2020-08-17 09:59:48 +08:00