holmesshuan / oisr-pytorch Goto Github PK
View Code? Open in Web Editor NEWPyTorch implementation of "ODE-inspired Network Design for Single Image Super-Resolution"(CVPR2019)
License: BSD 2-Clause "Simplified" License
PyTorch implementation of "ODE-inspired Network Design for Single Image Super-Resolution"(CVPR2019)
License: BSD 2-Clause "Simplified" License
主要就是基本模块改了,框架是EDSR吗?
Thanks for your sharing.、
python main.py --data_test Set5+Set14+B100+Urban100+DIV2K --data_range 801-900 --scale 2 --test_only --ext bin --dir_data /home/data/ --n_resblocks 8 --n_feats 122 --pre_train ./RK2-m/x2.pt --gpu-id 4,5,6,7 --chop --n_GPUs 4 --save_results
Is the dataset Set5 path format supposed to be /home/data/benchmark/Set5/HR
and /home/data/benchmark/Set5/LR_bicubic/X2
?
(NTIRE2019)In train.sh the n_feats=64, when I use pre_trained params the n_feats=512, why?
i tried to run RK3/main.py with RK3 pretrained models but below error occurred:
D:\Artificial Intelligence\SuperResolution\New folder\RK3>python main.py --test_only --save_results
Making model...
Load the model from weight/x4.pt
Traceback (most recent call last):
File "D:\Artificial Intelligence\SuperResolution\New folder\RK3\model\edsr.py", line 60, in load_state_dict
own_state[name].copy_(param)
RuntimeError: The expanded size of the tensor (64) must match the existing size (256) at non-singleton dimension 0. Target sizes: [64, 3, 3, 3]. Tensor sizes: [256, 3, 3, 3]
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "main.py", line 21, in <module>
model = model.Model(args, checkpoint)
File "D:\Artificial Intelligence\SuperResolution\New folder\RK3\model\__init__.py", line 35, in __init__
cpu=args.cpu
File "D:\Artificial Intelligence\SuperResolution\New folder\RK3\model\__init__.py", line 109, in load
if load_from: self.get_model().load_state_dict(load_from, strict=False)
File "D:\Artificial Intelligence\SuperResolution\New folder\RK3\model\edsr.py", line 66, in load_state_dict
.format(name, own_state[name].size(), param.size()))
RuntimeError: While copying the parameter named head.0.weight, whose dimensions in the model are torch.Size([64, 3, 3, 3]) and whose dimensions in the checkpoint are torch.Size([256, 3, 3, 3]).
In NTIRE2019/OISR/experiment/log.txt,a record 1600/15960 , could you tell me how to get the 15960 please?
论文中图片中你提出的block都是固定值,为什么代码不是固定值,而是可学习的呢
self.scale1 = nn.Parameter(torch.FloatTensor([0.5]), requires_grad=True)
参数requires_grad=Ture 是自动学习权重吗?
can you share your paper?
When I download the training dataset, the system prompts the extraction code(ij6f) error.
I am facing an implementation error in import _DataLoaderIter from torch.utils.data.dataloader in dataloader.py
Im using pytorch 1.4.0 and python 3.7
I tried to change _DataLoaderIter to DataLoaderIter. Even that doesn't work.
I tried many ways to debug the error. However, I get the error:
ImportError Traceback (most recent call last)
in ()
4 from torch._C import _set_worker_signal_handlers, _set_worker_pids, _remove_worker_pids, _error_if_any_worker_fails
5 from torch.utils.data.dataloader import DataLoader
----> 6 from torch.utils.data.dataloader import DataLoaderIter
ImportError: cannot import name 'DataLoaderIter'
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.