colin
/
witnn
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Commit Graph

  • c796977b6b add amp debug master Colin 2022-01-14 09:16:26 +0000
  • c8e26afd67 Add AMP Colin 2022-01-10 08:05:23 +0000
  • 486e9726c5 code refine colin 2020-07-30 14:48:58 +0800
  • 56b961459c remove debug optput colin 2020-07-29 15:50:20 +0800
  • 206ba35c93 refine export ,add layer2-4 colin 2020-07-28 13:50:58 +0800
  • ce1aca9531 mv shape to single file colin 2020-07-28 13:35:50 +0800
  • e5410cb1b2 add shape output and CheckResnet50 colin 2020-07-27 20:27:00 +0800
  • e2e2b01b54 add layer1 inner print colin 2020-07-27 16:30:54 +0800
  • 6309c2f066 fix auto layer1 lose two layer colin 2020-07-24 14:03:07 +0800
  • 7e3d3bab6f remove unuse print colin 2020-07-23 17:07:18 +0800
  • 7a3dfde615 fix resnet to eval static , fix batch normal error colin 2020-07-23 15:26:40 +0800
  • ee0a12f7dd add verify resnet colin 2020-07-20 17:35:05 +0800
  • 039310704e fix byte count with float * 4 colin 2020-07-20 16:00:32 +0800
  • 3d062150e9 add input and output in resnet50 demo colin 2020-07-20 15:45:02 +0800
  • a14983b756 add bn weight and bias colin 2020-07-20 13:05:12 +0800
  • 7ae864fe48 add ; in cc colin 2020-07-20 11:21:02 +0800
  • ef0dd5bc00 add ResNet weight dump colin 2020-07-18 11:23:58 +0800
  • 7ab55557cc add MnistDemo colin 2020-07-13 20:14:01 +0800
  • 67c5cc13a0 Fix readme colin 2020-03-08 17:11:54 +0800
  • d3046d1131 fix as 4 kernels in first layer. c 2020-01-02 15:51:24 +0800
  • 346159068c Speedup to search. c 2019-12-30 17:24:27 +0800
  • 18126824ee Add more train to test entropy resualt. c 2019-12-27 18:04:51 +0800
  • e0de909d62 Revert "Refine kernal Random to w*h*16 kinds possiable." c 2019-12-19 16:07:29 +0800
  • 675cd8567e Refine kernal Random to w*h*16 kinds possiable. c 2019-12-19 15:20:47 +0800
  • 8404d93880 Speed up by multi cpu to handle histogram c 2019-12-18 18:22:59 +0800
  • 418e79df60 Add Directed search 8 weights instead of work with two steps. c 2019-12-17 13:53:27 +0800
  • fa7a066b87 Test 7x7 image of Mnist. c 2019-12-09 22:23:29 +0800
  • a368c4eae5 Try deep search ,and test in Minist DataSet c 2019-11-28 21:06:06 +0800
  • f105932c26 Add readme. c 2019-11-25 14:17:03 +0800
  • 8bb2c5562d Add bn layer to 3335,Get if it is benify that. Colin 2019-11-06 15:17:14 +0800
  • c2e149fb68 Check if train kernel is benifit to pretrain. colin 2019-10-31 14:35:47 +0800
  • 8a9acaa94e New test colin 2019-10-15 15:26:03 +0800
  • febad1f9c3 Fix filter must sum=0 colin 2019-10-03 19:50:45 +0800
  • 9628b2c17e add minactive return in search colin 2019-10-03 16:30:52 +0800
  • 130a6942b9 refine Search Weight calculater time. colin 2019-09-30 18:33:47 +0800
  • 76c9b7f05e Refine line display visdom colin 2019-09-23 10:02:01 +0800
  • 7f39f884fc Refine Distance of kernels colin 2019-09-20 12:48:45 +0800
  • b0246d2a95 Refine find best 8 colin 2019-09-17 23:20:24 +0800
  • f858ae1821 Add UnsuperLearnFindBestWeight to find best from 4000 colin 2019-09-13 21:03:38 +0800
  • 6201163e14 Add UnsuperLearnTrainWeight to train new weight kernel colin 2019-09-02 19:28:04 +0800
  • d4a22775d4 Add two method of real wit kernel genarator. colin 2019-09-01 12:21:22 +0800
  • 5220bf6885 start to creat 8 from 4000 colin 2019-08-27 23:06:24 +0800
  • d9df29f04a train with pretrain colin 2019-08-23 23:27:45 +0800
  • 81d8931842 init git colin 2019-08-19 15:53:10 +0800