kylin9511 / crnet Goto Github PK

View Code? Open in Web Editor NEW

73.0 73.0 27.0 23 KB

Channel Reconstruction Network implemented in PyTorch

License: MIT License

Python 100.00%

crnet's People

Contributors

Stargazers

Watchers

crnet's Issues

请问下代码的test输出的Rho是什么意思？论文里没看到这个参数

One doubt from your code.

Why have you take the size of nc_expand in the evaluator function of your statics.py file to be 257? Because usually for taking FFT, the input size is taken as multiples of 2. Is there any reason you have chosen 257 instead of 256?
Please clarify.

在新的pytorch下运行出现的一些bug

我的gpu用的是1080ti
为了匹配cuda版本我装了torch1.8.0
然后就出现了一些函数无法找到的问题

尝试解决如下

如下是我的pytorch里面的fft函数的写法，不知道是不是因为函数的写法不一样造成的bug

改了之后可以在gpu上训练了但是出现了rho结果为nan+nanj的情况
由于不理解这个rho的含义不知道这个结果是不是错了呢？究竟是哪里出错的？

rho计算请教

你好，想请教一下代码中NMSE和rho计算的问题：
（1）请问在训练中使用的信道应该都是经过二维FFT之后并进行归一化的信道吧？
如果是这样的话，原始的信道应该是需要降信道进行二维的IFFT2，但是您的代码中为什么是进行1维的FFT呢？这里我不太明白。
（2）如果说FFT和IFFT2是等效的话，那么为什么频域变换成257维，然后又只取其中的125维呢？这里也不太清楚。

虽然您的README已经写的很详细了，但是我使用argparse很少，所以运行有一点错误，但暂时还不知道错误在哪里！！

wxzhu@sunlaoshilab-W560-G20-Invalid-entry-length-16-Fixed-up-to-11:~/下载/home$ tree
.
├── COST2100
│   ├── A128.mat
│   ├── A32.mat
│   ├── A512.mat
│   ├── A64.mat
│   ├── DATA_HtestFin_all.mat
│   ├── DATA_HtestFout_all.mat
│   ├── DATA_Htestin.mat
│   ├── DATA_Htestout.mat
│   ├── DATA_Htrainin.mat
│   ├── DATA_Htrainout.mat
│   ├── DATA_Hvalin.mat
│   └── DATA_Hvalout.mat
├── CRNet
│   ├── dataset
│   │   ├── cost2100.py
│   │   └── __init__.py
│   ├── LICENSE
│   ├── main.py
│   ├── models
│   │   ├── crnet.py
│   │   ├── __init__.py
│   │   └── __pycache__
│   │       ├── crnet.cpython-36.pyc
│   │       └── __init__.cpython-36.pyc
│   ├── README.md
│   └── utils
│       ├── __init__.py
│       ├── init.py
│       ├── logger.py
│       ├── parser.py
│       ├── __pycache__
│       │   ├── __init__.cpython-36.pyc
│       │   ├── init.cpython-36.pyc
│       │   ├── logger.cpython-36.pyc
│       │   ├── scheduler.cpython-36.pyc
│       │   ├── solver.cpython-36.pyc
│       │   └── statics.cpython-36.pyc
│       ├── scheduler.py
│       ├── solver.py
│       └── statics.py
└── Experiments
    ├── checkpoints
    │   ├── in_04.pth
    │   ├── in_08.pth
    │   ├── in_16.pth
    │   ├── in_32.pth
    │   ├── in_64.pth
    │   ├── out_04.pth
    │   ├── out_08.pth
    │   ├── out_16.pth
    │   ├── out_32.pth
    │   └── out_64.pth
    ├── log.out
    └── run.sh

这个是文件目录，在推测阶段，有如下错误：

wxzhu@sunlaoshilab-W560-G20-Invalid-entry-length-16-Fixed-up-to-11:~/下载/home/Experiments$ ./run.sh
Traceback (most recent call last):
  File "/home/wxzhu/下载/home/CRNet/main.py", line 4, in <module>
    from utils.parser import args
  File "/home/wxzhu/下载/home/CRNet/utils/__init__.py", line 5, in <module>
    from .solver import *
  File "/home/wxzhu/下载/home/CRNet/utils/solver.py", line 13, in <module>
    Result = namedtuple('Result', field, defaults=(None,) * len(field))
TypeError: namedtuple() got an unexpected keyword argument 'defaults'
./run.sh: 行 5: --evaluate: 未找到命令
./run.sh: 行 6: --batch-size: 未找到命令

run.sh文件如下：

python /home/wxzhu/下载/home/CRNet/main.py \
  --data-dir '/home/wxzhu/下载/home/COST2100' \
  --scenario 'in' \
  --pretrained '/home/wxzhu/下载/home/Experiments/checkpoints/in_04.pth' \  # Pretrained model loading
  --evaluate \  # Inference mode
  --batch-size 200 \
  --workers 0 \
  --cr 4 \
  --cpu \
  2>&1 | tee log.out

GPU下运行

你好请问一下按照你的代码用GPU跑的话，最后跑的时候怎么还是用CPU再运行呀

Dataset is not accessible

Some questions about your excellent work

I have read the introduction of your work and it is an excellent job. However, I have some questions:

I cannot find your paper and can you share your paper.
Do you use the same dataset used in CsiNet. If using the same, have you compared your simulation results with that of the CsiNet+[1] in Table IV?
If you have any problem, please feel free to contact me: [email protected].

[1] Guo J, Wen C K, Jin S, et al. Convolutional Neural Network based Multiple-Rate Compressive Sensing for Massive MIMO CSI Feedback: Design, Simulation, and Analysis[J]. arXiv preprint arXiv:1906.06007, 2019.

Recommend Projects

React

A declarative, efficient, and flexible JavaScript library for building user interfaces.
Vue.js

🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript

TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
TensorFlow

An Open Source Machine Learning Framework for Everyone
Django

The Web framework for perfectionists with deadlines.
Laravel

A PHP framework for web artisans
D3

Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

javascript

JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
web

Some thing interesting about web. New door for the world.
server

A server is a program made to process requests and deliver data to clients.
Machine learning

Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Visualization

Some thing interesting about visualization, use data art
Game

Some thing interesting about game, make everyone happy.

Recommend Org

Facebook

We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft

Open source projects and samples from Microsoft.
Google

Google ❤️ Open Source for everyone.
Alibaba

Alibaba Open Source for everyone
D3

Data-Driven Documents codes.
Tencent

China tencent open source team.