vwenyu-L-CSDN博客

原创 tensorflow-basic tutorial

CONTACT WEIBOREFERENCEIMAGE PROCESS TF.APP TENSORBOARD SERVINGSOMETHING

2017-12-09 09:59:56 310

原创 Deep Learning

CONTACT WEIBOcapsule networkgenerative adversarial network (GAN)recurrent neural network (RNN)source of deep learning and mlpaper sources

2017-12-09 09:56:05 274

原创 pytorch-basic tutorial

CONTACT WEIBOREFERENCESTENSORCLASS MODULEDESIGN MODULECONTAINERMULTI GPUDATASETSAVE RESTOREOPTIMIZERPARAM INITERRORSTENSORBOARDXPROFILER''' by L'''import torch import...

2017-12-06 17:39:19 528

原创 pytorch-distributed traning

1. 单机多卡数据并行```model = xxx losses = torch.nn.parallel.data_parallel(model, inputs=(), device_ids=[], dim=x) # functional style```2. pytorch-1.0 distributed2.1 单机多卡 2.2 多级多卡 ----...

2019-01-28 15:25:48 990 2

原创 jupyter notebook-远程登录服务器

jupyter notebook --generate-configjupyter notebook password~/.jupyter/jupyter_notebook_config.py ~/.jupyter/jupyter_notebook_config.json c.NotebookApp.ip='*'c.NotebookApp.password = u'sha:ce..'...

2018-03-13 14:39:16 520

原创 pytorch-tensorboardX

from tensorboardX import SummaryWriterimport torchvision.utils as vutilswriter = SummaryWriter()for i, (x, y) in enumerate(dataloader): xxx writer.add_scalar('/train/loss', loss) wr...

2018-03-12 20:27:16 679

原创 tensorflow-tf.app

import tensorflow as tf import tqdmimport timetf.app.flags.DEFINE_string('var_name', 'hello world', 'var name')FLAG = tf.app.flags.FLAGSdef main(_): print(FLAG.var_name) for i in t...

2018-03-12 09:53:41 340

原创 tensorflow-serving

serving deployreference 1. http://blog.csdn.net/wangjian1204/article/details/68928656

2018-03-09 18:20:13 513

原创 tensorflow-tensorboard

tensorboardtf.summarytf.summary.scalar('/train/loss', loss)summary = tf.summary.merge_all()# writer = tf.summary.FileWriter('/outputs', tf.default_graph() )writer = tf.summary.FileWrit...

2018-03-09 16:30:16 424

原创 tensorflow-input data process

tf.gfiletf.imagereference 1. http://blog.csdn.net/White_Idiot/article/details/78825341

2018-03-09 16:26:37 325

原创 capsule network

1. DynamicRoutingBetweenCapsules2. Matrix capsules with EM Routing-----reference---------------------https://jhui.github.io/2017/11/14/Matrix-Capsules-with-EM-routing-Capsule

2017-11-25 16:24:40 1722

原创 python-conda

1. 不同版本的python管理conda update condaconda create -n python35 python=3.5

2017-11-23 21:59:38 1123

原创 vs-code tips

setting-shortcuts 功能操作 ctl+d select words [one or more times] alt+cursor select multis position or lines alt+shift+mouse select multi columnsdebugchoose and open launch.jsonstart deb

2017-11-20 20:33:24 303

原创 python-os

path.path = os.path.abspath(__file__)dir = os.path.dirname(path)os.path.split(path)os.path.exists(path)

2017-08-01 21:10:30 333

原创 linux-watch

watch -n 1 -d command

2017-07-31 13:57:10 646

原创 python-heapq

top_ka = []k = 6heapq.heappush(a, ele)heapq.heappop(a)heapq.heareplace(a, ele)

2017-07-29 15:29:13 309

原创 python-super

cls.mro()method resolution oderdef super(cls, inst): mro = inst.__class__.mro() return mro[mro.index[cls] + 1]

2017-07-29 13:55:21 280

原创 flownet2-errors

1. engine error'Layer conv1 has unknown engine'modify: CUDNN -> CAFFEhttps://github.com/BVLC/caffe/issues/26822. ‘RandomGeneratorParameter’ does not name a type, ’https://github.com/l

2017-07-28 15:24:35 1314

原创 stack

def calx(op, a, b): if op == 'mul': return str(int(a) * int(b)) elif op == 'add': return str(int(a) + int(b)) elif op == 'div': if b == '0': return 'er

2017-07-26 22:20:14 334

原创 python-argparse

parser = argparse.ArgumentParser()parser.add_argument('echo')parser.add_argument('-num', typr=int, default=0)args = parser.parse_args()print(args.echo, args.num)

2017-07-24 14:03:05 279

原创 python-pypdf2

from PyPDF2 import PdfFileWriter, PdfFileReaderpdf_input = PdfFileReader(open('xx.pdf', 'rb'))pdf_output = PdfFileWriter()print(pdf_input.numPages)page = pdf_input.getPage(0) #pdf_output.addPa

2017-07-21 17:07:51 1378

原创 linux-screen

screen screen -lsscreen -r xxctrl + a+ d

2017-07-21 13:14:16 349

原创 python-bisect/np-searchsorted

bisect()np.searchsorted()

2017-07-20 13:28:11 404

原创 visual studio-third part lib

Configuration Properties -> VC++Directories -> Include Directories(.h) Library Directories(.lib) Linker/Input add .lib nameabout .dll file add directories into Configuration Properties/Debu

2017-07-14 22:40:51 370

原创 N/D to float number

#include "stdafx.h"#include#include#include#include#includeusing namespace std;int _tmain(int argc, _TCHAR* argv[]){ int n, d; cin >> n >> d; int tmpn = n; int tmpd = d; int x = 0; w

2017-07-12 22:15:02 271

原创 mxnet-install using source code

1. openblasmake make PREFIX=/PATH/ installexport LD_LIBRARY_PATH=/PATH/bin:$LD_LIBRARY_PATHor.modify ~/.bashrcsource ~/.bashrcmxnet/make.config.mdUSE_BALS=openblasA

2017-07-04 19:47:36 491

原创 python-skimage

1. skimage.utilcroppadrandom_noiseregular_gridview_as_windowsview_as_blcks2. skimage.transformimg = np.array(img, dtype=float)resize(img, [], order=1) 0-nearest 1-bilinear 3-cub

2017-07-02 21:08:40 1690

原创 python-PIL

----------------------------https://www.daniweb.com/programming/software-development/code/216637/resize-an-image-python

2017-06-30 22:30:39 335

1. nn.DataParallelmodel = nn.DataParallel(model.cuda(1), device_ids=[1,2,3,4,5])criteria = nn.Loss() # i. .cuda(1) 20G-21G ii. cuda() 18.5G-12.7G iii. nothing 16.5G-12.7G. these all use almost same...

2017-06-30 10:46:19 10756 1

原创 python-classdictsetattrgetattr

In [153]: class test(object): ...: def __init__(self, a, b): ...: self._cont = {} ...: self.a = a ...: self.b = b ...: #self._cont = {}

2017-06-22 15:54:50 465

原创 pytorch-containers

In [90]: class MM(nn.Module): ...: def __init__(self): ...: super(MM, self).__init__() ...: #1. ...: #self.dicx = OrderedDict() .

2017-06-22 11:15:54 553

原创 paper-image/video segmentation

fcndeeplab-v2 deeplab-v3deepmask sharpmask----------------1. http://www.dongzhuoyao.com/deeplab-semantic-image-segmentation-with-deep-convolutional-nets-atr

2017-06-21 15:23:19 528

原创 pytorch-tensor data type

1. torch.FloatTensor(2,3).normal_(0,1)torch.LongTensor(2,3).random(0,4)torch.from_numpy(arr).float() data.float()model.float()------------------------------------

2017-06-21 12:10:58 2659

原创 pytorch-design NEW Function and Module

-------------------------1. https://www.qcloud.com/community/article/831497

2017-06-12 22:31:25 504

原创 pytorch-errors

0.RuntimeError: save_for_backward can only save input or output tensors, but argument 0 doesn't satisfy this conditionWhen custom Funciton & Module, and the module need backward, the input should ...

2017-06-12 10:50:05 10273 5

原创 pytorch-loss function

-----------------reference---------------1.http://blog.csdn.net/zhangxb35/article/details/72464152?utm_source=itdadao&utm_medium=referral

2017-06-11 15:28:34 717

原创 83. Remove Duplicates from Sorted List

/** * Definition for singly-linked list. * struct ListNode { * int val; * ListNode *next; * ListNode(int x) : val(x), next(NULL) {} * }; */class Solution {public: ListNode* de

2017-06-09 09:09:03 260