Commit Graph

219 Commits

Author SHA1 Message Date
sunsuodong 032246837f add 1.6 release notes 2022-01-10 03:33:46 -08:00
i-robot 388e96c2ff
!28408 [MS][LITE] dpico custom kernel memory optimize
Merge pull request !28408 from jianghui58/dpico_dev_mem
2022-01-05 08:06:55 +00:00
jianghui58 3af80892be fix register bug 2022-01-05 09:41:02 +08:00
chenping 68e93e3b18 model_zoo模型在Ascend710上的在线推理调试 2021-12-26 15:28:03 +08:00
i-robot 5503948658 !27610 Support Ascend710 device to run converter and benchmark.
Merge pull request !27610 from chenping/master
2021-12-20 12:20:38 +00:00
xuanyue 1dd50a0a6f open ConverterContext and open special nodes 2021-12-18 23:07:29 +08:00
i-robot ae6ac1a52e !27772 [MSLITE][GPU] opengl use old benchmark bugfix and remove opengl Dependent library
Merge pull request !27772 from Greatpan/master_gl
2021-12-17 02:32:30 +00:00
chenping eacab8afe8 Support Convert and Benchmark for Ascend710. 2021-12-16 20:02:49 +08:00
greatpanc c8e043597b opengl old benchmark bugfix and remove opengl Dependent library 2021-12-16 19:47:36 +08:00
zhengjun10 8394ee9878 fix albert loss not found issue 2021-12-15 18:57:58 +08:00
greatpanc 29b852f46d opengl bugfix 2021-12-07 09:15:02 +08:00
zhengyuanhua c0332b4fc7 310 support image size 2021-11-24 14:15:43 +08:00
Fazzie 8db6239b4a fix opengl texture 2021-11-22 14:51:41 +08:00
Fazzie 25bd921971 enable opengl Texture 2021-11-17 14:56:11 +08:00
Liu_Xuu db5f132127 [MSLITE] add get rank id in tensorrt delegate 2021-11-16 14:20:07 +08:00
i-robot 525df1ca30 !26101 【MS】【LITE】infershape support get attr from kernel
Merge pull request !26101 from chenjianping/code_check
2021-11-11 04:32:21 +00:00
chenjianping 9579b20faf infershape support getAttr from kernel 2021-11-11 09:14:47 +08:00
i-robot 2c3ce1d73c !24811 [MSLITE] change ElementsNum data type to int64_t
Merge pull request !24811 from zhanyuan/super_large
2021-11-08 11:00:13 +00:00
zhanyuan 7860a94547 [MSLITE] change ms tensor's ElementsNum date type to int64_t 2021-11-08 16:29:25 +08:00
i-robot 0f143d65b2 !25834 reduce memcpy while big model load
Merge pull request !25834 from hangq/fuzz
2021-11-08 08:04:53 +00:00
i-robot f029f8de1a !25912 【MS】【LITE】support get attr and config from kernel
Merge pull request !25912 from chenjianping/code_check
2021-11-05 08:23:43 +00:00
i-robot 279a099936 !25138 [MS][LITE][ToD] mindir_fp16
Merge pull request !25138 from Nizzan/export_nizzan
2021-11-05 06:28:01 +00:00
chenjianping 8acc951a47 support get config and attr from kernel 2021-11-05 11:14:24 +08:00
hangangqiang 30ab8a4540 reduce memcpy while big model load 2021-11-05 10:12:13 +08:00
nizzan 665e2a738e Support mix precision mindir file 2021-11-04 12:28:22 +02:00
Emir Haleva 8433b1be54 API to change TensorsWeights for Site AI 2021-11-03 16:02:09 +02:00
i-robot 1639a33c07 !25351 310 support dynamic batchsize
Merge pull request !25351 from zhengyuanhua/model_br1
2021-10-26 08:15:36 +00:00
zhengyuanhua 200a18103e support dynamic batch size 2021-10-26 09:11:34 +08:00
yangruoqi713 d071a7e91f [MSLITE][DEVELOP] clean code check warnings 2021-10-25 19:06:47 +08:00
xuanyue 35e7c5a9f1 rename parser_context.h 2021-10-20 13:22:59 +08:00
xuanyue 1615e4907f make external nodeparser lightweight and add nodeparser ut 2021-10-19 11:05:32 +08:00
yoni 097e09d541 mini_yolo bit exact 2021-10-14 09:19:11 +03:00
chenjianping b3353a15c7 clean code 2021-10-13 22:17:17 +08:00
xuanyue 7f75d09d8b open the function of setting basic info to exported model 2021-09-27 15:48:37 +08:00
zhengyuanhua 5440f9fbc1 external interface modify 2021-09-22 21:05:38 +08:00
xuanyue 6f738d2e4a draw out model_parser basic class 2021-09-16 14:29:40 +08:00
xuanyue ecafae75d5 abstract out passs 2021-09-13 20:35:31 +08:00
i-robot 901124c9bd !23198 context support to set Ascend310 config
Merge pull request !23198 from zhengyuanhua/code_review
2021-09-13 01:41:54 +00:00
zhengyuanhua e25ca34ab0 context support set 310 and add parse config file 2021-09-12 21:35:48 +08:00
zengxianglong ab6ed9a494 tflite converter support multiple subgraphs 2021-09-11 12:40:09 +08:00
chaijun c515fcd361 gpu demo 2021-09-09 10:34:42 +08:00
zhengjun10 dcff9ade54 lite train support mix precision model convert 2021-09-06 20:57:55 +08:00
zhengjun10 fc4e6dc533 lite train support mix precision model 2021-09-02 20:50:28 +08:00
Emir Haleva a14eac9fb2 Added Federated Learning API 2021-09-01 22:45:39 +03:00
sunsuodong 062dcaaaa1 vs build converter 2021-08-29 20:36:50 +08:00
i-robot 207ac21853 !22414 [MSLITE] add support gpu device context set device info for tensorrt
Merge pull request !22414 from Liu_Xuu/trt_0825_cudadevice
2021-08-28 09:45:20 +00:00
Liu_Xuu 678b8cf784 [MSLITE] add support gpu device context set device info for tensorrt 2021-08-27 18:06:19 +08:00
i-robot 2a2c2c661b !22317 [MS][LITE][Develop] benchmark new api support string tensor and fix vs build
Merge pull request !22317 from sunsuodong/code_review
2021-08-26 08:00:45 +00:00
sunsuodong 6c7728ae0b review 2021-08-26 14:30:29 +08:00
i-robot 0d1e0b0f5c !22038 [lite]open node_parser register
Merge pull request !22038 from 徐安越/master
2021-08-26 03:56:58 +00:00