OrgLion-ML
/
darknet
mirror of https://github.com/AlexeyAB/darknet.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
1,133 Commits 1 Branch 7 Tags 361 MiB
Commit Graph

1133 Commits

Author SHA1 Message Date
AlexeyAB e203ab9046 Minor fixes 2018-09-25 02:39:30 +03:00
Alexey 129417ae7f
Update Readme.md 2018-09-24 02:06:35 +03:00
AlexeyAB 7dd97537fb XNOR-net tiny-yolo_xnor.cfg ~2x faster than cuDNN on CUDA (nVidia GPU Maxwell) 2018-09-22 02:01:14 +03:00
AlexeyAB 0224ba3d0d Compile fix 2018-09-17 23:59:29 +03:00
AlexeyAB 03e95320a1 XNOR coalesced memory access, and avoid bank conflicts 2018-09-17 23:39:25 +03:00
Alexey 8e0b265067
Update Readme.md 2018-09-14 15:54:01 +03:00
Alexey b829fee30b
Update Readme.md 2018-09-14 15:29:52 +03:00
AlexeyAB ca43bbdaae Fixed openmp bugs for XNOR 2018-09-12 16:22:54 +03:00
AlexeyAB c0e01fd63c Test for XNOR-conv on CUDA 2018-09-08 02:46:05 +03:00
AlexeyAB 57e878b4f9 another compile fix 2018-09-07 15:25:37 +03:00
AlexeyAB b141f85cab Compile fix 2018-09-07 15:07:46 +03:00
AlexeyAB 007878393f Temporary Slow implementation of XNOR on CUDA (shared_memory) 2018-09-06 23:21:26 +03:00
AlexeyAB c4a9e3422e Temporary implementation of XNOR on CUDA 2018-08-31 02:47:58 +03:00
AlexeyAB 9753b72aeb temp fix, don't use it 2018-08-30 17:24:41 +03:00
AlexeyAB 18d5e4f39c Fixed yolov3-tiny_xnor.cfg 2018-08-24 18:29:40 +03:00
AlexeyAB 08285dbc84 Added tiny-yolo-v2-XNOR-cfg-file: tiny-yolo_xnor.cfg 2018-08-23 23:09:47 +03:00
AlexeyAB 5b6be00d4b Added yolov3-tiny_xnor.cfg 2018-08-23 15:28:34 +03:00
AlexeyAB bb48b9c992 Merge branch 'master' of github.com:AlexeyAB/darknet 2018-08-23 03:00:48 +03:00
AlexeyAB 31b6b0bad3 XNOR-net 4x acceleration on CPU for yolov2-tiny - 22 FPS (CPU Core i7 6700K) 2018-08-23 02:44:21 +03:00
AlexeyAB f606b5456e XNOR-net 21 FPS on CPU yolov2-tiny.cfg 2018-08-22 17:52:48 +03:00
Alexey 88dbaff4e6
Merge pull request #1466 from r0l1/master 
updated README for learning rate adjustment
 2018-08-22 15:53:31 +03:00
Roland Singer dbdd31ee21 updated README to include information about learning rate adjustment for multiple GPUs 2018-08-22 13:56:41 +02:00
AlexeyAB 8436251a05 maxpool_layer on CPU uses AVX2 and OpenMP 2018-08-21 20:44:18 +03:00
AlexeyAB e1801f5aec Added trim() function for label paths 2018-08-21 00:52:16 +03:00
Alexey 8e018980a8
Update Readme.md 2018-08-20 21:58:51 +03:00
Alexey 08d0e07ed6
Update Readme.md 2018-08-19 14:08:02 +03:00
AlexeyAB 552207d6f0 Minor fix 2018-08-17 19:58:05 +03:00
AlexeyAB c2a8cc2ad3 Fix GCC compile 2018-08-16 03:02:54 +03:00
AlexeyAB bd8a5c3eb1 Different checks for AVX and FMA 2018-08-16 02:26:40 +03:00
Alexey fa7bfe3412
Update Readme.md 2018-08-16 00:07:36 +03:00
AlexeyAB 1e866f0063 Fixed yolov3-tiny.cfg 2018-08-16 00:01:22 +03:00
Alexey a723e1c62a
Merge pull request #766 from HotChick91/AlexeyAB-mask 
Fix yolov3-tiny mask
 2018-08-15 23:52:09 +03:00
AlexeyAB f007f7b90e Fixes for MSVS2017 2018-08-15 23:47:56 +03:00
Alexey 140bc1882d
Merge pull request #1416 from mhaghighat/master 
Update to prevent the error at Line 74: data being one-dimensional
 2018-08-15 20:31:31 +03:00
mhaghighat 756a3144cc Update to prevent the error at Line 74: data being one-dimensional 2018-08-15 10:04:14 -07:00
AlexeyAB 7f9145729c Another one fix for MSVS2017 2018-08-14 18:38:39 +03:00
AlexeyAB e548489182 Fix for MSVS2017 for _mm256_extract_epi64 and _mm256_extract_epi32 2018-08-14 14:17:09 +03:00
AlexeyAB f92b20580a Some fixes for AVX support on CPU 2018-08-14 01:51:31 +03:00
AlexeyAB b1dddf02cc Fixed AVX compiled bug 2018-08-13 02:43:45 +03:00
AlexeyAB 1f2155b886 Experiments 2018-08-11 02:49:55 +03:00
AlexeyAB a9fef1bd66 Bug fixes. Tested im2col_cpu_custom_transpose - bad way. 2018-08-11 00:26:53 +03:00
AlexeyAB 3e856ec04e Optimized: transpose 2018-08-10 01:27:20 +03:00
Alexey 78284f791e
Update Readme.md 2018-08-09 22:53:09 +03:00
AlexeyAB d59a172a1d Fixed wrong labels 2018-08-09 14:58:26 +03:00
AlexeyAB 5c876bb918 Another one fix 2018-08-09 03:12:45 +03:00
AlexeyAB 3baf534a2d Compile error fix 2018-08-09 03:07:06 +03:00
AlexeyAB d6162af210 Optimized on CPU: gemm_bin, im2col, activation, transpose 2018-08-09 02:31:36 +03:00
AlexeyAB a284a7da8d Try to use avx_hs() - slow and requires alignment 4096 bits < (l.size*l.size*l.c) 
May be faster only from 8192 bits and more.
 2018-08-08 19:08:58 +03:00
AlexeyAB 0a326e7afe XNOR-net on CPU AVX2 2018-08-08 02:45:47 +03:00
AlexeyAB e6c97a53a7 Maxpool fixes 2018-08-07 23:19:50 +03:00