Arch Linux User Repository

This package provides the cli version of waifu2x-caffe. A readme for all of the flags is located in /usr/share/docs/waifu2x-caffe/README. The models are located at /usr/share/waifu2x-caffe/models.

@sommio Feel free to alter the PKGBUILD to have the older versions of cuda and cudnn on your local system. Once you have cloned this repository, make the necessary changes by updating the fields in the depends section to the older versions. Then you can commit that change to your local git repo. What this will do is allow for any changes in the future to be merged with the changes you added and you may not need to make them every time I release an update.

If the older versions work let me know.

Not only is this compatible with older nvidia gpu's, but caffe has higher performance in cudnn7 than cudnn8!

Can you make a waifu2x-caffe pkgbuild that uses cuda10 and cudnn7? Thanks!

@ginnokami The first time I tried to use this program, it seems that my graphics card does not work with the latest CUDA...

@sommio Has this application ever worked for you or is this the first time you are trying to use it? Taking a look at CUDNN Capability Matrix shows that your CUDA capability may no longer be supported on the newer CUDNN releases.

Are you able to run the tool without CUDNN and just CUDA? That would be just removing the -cudnn option.

@ginnokami

deviceQuery

./deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "NVIDIA GeForce GTX 650"
  CUDA Driver Version / Runtime Version          11.4 / 11.6
  CUDA Capability Major/Minor version number:    3.0
  Total amount of global memory:                 973 MBytes (1020526592 bytes)
  (002) Multiprocessors, (192) CUDA Cores/MP:    384 CUDA Cores
  GPU Max Clock rate:                            1058 MHz (1.06 GHz)
  Memory Clock rate:                             2500 Mhz
  Memory Bus Width:                              128-bit
  L2 Cache Size:                                 262144 bytes
  Maximum Texture Dimension Size (x,y,z)         1D=(65536), 2D=(65536, 65536), 3D=(4096, 4096, 4096)
  Maximum Layered 1D Texture Size, (num) layers  1D=(16384), 2048 layers
  Maximum Layered 2D Texture Size, (num) layers  2D=(16384, 16384), 2048 layers
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       49152 bytes
  Total shared memory per multiprocessor:        49152 bytes
  Total number of registers available per block: 65536
  Warp size:                                     32
  Maximum number of threads per multiprocessor:  2048
  Maximum number of threads per block:           1024
  Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
  Max dimension size of a grid size    (x,y,z): (2147483647, 65535, 65535)
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             512 bytes
  Concurrent copy and kernel execution:          Yes with 1 copy engine(s)
  Run time limit on kernels:                     Yes
  Integrated GPU sharing Host Memory:            No
  Support host page-locked memory mapping:       Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support:                        Disabled
  Device supports Unified Addressing (UVA):      Yes
  Device supports Managed Memory:                Yes
  Device supports Compute Preemption:            No
  Supports Cooperative Kernel Launch:            No
  Supports MultiDevice Co-op Kernel Launch:      No
  Device PCI Domain ID / Bus ID / location ID:   0 / 1 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 11.4, CUDA Runtime Version = 11.6, NumDevs = 1
Result = PASS

@ginnokami I tried using the makechrootpkg build, but it still doesn't work

mkarchroot $CHROOT/root base-devel
makechrootpkg -c -r $CHROOT

my log and pkg: Google Drive

@ginnokami Yes, I am using the latest version of cudnn and cuda and pkgbuild, but my nvida driver version is 470xx.

community/cuda 11.6.0-1 
community/cudnn 8.3.1.22-1
extra/opencv 4.5.5-3

local/lib32-nvidia-470xx-utils 470.103.01-1
local/lib32-opencl-nvidia-470xx 470.103.01-1
local/libxnvctrl-470xx 470.103.01-1
local/nvidia-470xx-dkms 470.103.01-1
local/nvidia-470xx-settings 470.103.01-1
local/nvidia-470xx-utils 470.103.01-1
local/opencl-nvidia-470xx 470.103.01-1

waifu2x-caffe -p cudnn --gpu 0 -m noise_scale -n 2 -i in.jpg -o out.jpg

E20220226 12:29:23.864383 51447 common.cpp:167] Cannot create Cublas handle. Cublas won't be available.
Could not create log file: File exists
COULD NOT CREATE LOGFILE '20220226-122923.51447'!
F20220226 12:29:23.877106 51447 cudnn_conv_layer.cpp:54] Check failed: status == CUDNN_STATUS_SUCCESS (1 vs. 0)  CUDNN_STATUS_NOT_INITIALIZED
*** Check failure stack trace: ***
    @     0x7f1a608facbd  google::LogMessage::Fail()
    @     0x7f1a608fd298  google::LogMessage::SendToLog()
    @     0x7f1a608fa7e6  google::LogMessage::Flush()
    @     0x7f1a608fd90a  google::LogMessageFatal::~LogMessageFatal()
    @     0x7f1a60b6a3e5  caffe::CuDNNConvolutionLayer<>::LayerSetUp()
    @     0x7f1a60c60689  caffe::Net<>::Init()
    @     0x7f1a60c61fce  caffe::Net<>::Net()
    @     0x55a1495dcb5d  cNet::ConstractNet()
    @     0x55a1496154b3  Waifu2x::Init()
    @     0x55a1495bfa87  main
    @     0x7f1a4bee5310  __libc_start_call_main
    @     0x7f1a4bee53c1  __libc_start_main_alias_2
    @     0x55a1495bb0a5  _start

@sommio are you on the latest version (my last compile was on 2/7 when I updated to support building in chroot)? I am able to successfully run using cudnn using the following command:

waifu2x-caffe -p cudnn --gpu 0 -m noise_scale -n 2 -c 256 -h 1440 -i in.jpg -o out.jpg | nkf
変換に成功しました

Unable to use CUDA and CUDNN

CUDA:

変換に失敗したファイルがあります

CUDNN:

Log file created at: 2022/02/24 11:48:49
Running on machine: arch
Running duration (h:mm:ss): 0:00:00
Log line format: [IWEF]yyyymmdd hh:mm:ss.uuuuuu threadid file:line] msg
E20220224 11:48:49.032555 136521 common.cpp:167] Cannot create Cublas handle. Cublas won't be available.
F20220224 11:48:49.043658 136521 cudnn_conv_layer.cpp:54] Check failed: status == CUDNN_STATUS_SUCCESS (1 vs. 0)  CUDNN_STATUS_NOT_INITIALIZED

nvcc -v

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2021 NVIDIA Corporation
Built on Fri_Dec_17_18:16:03_PST_2021
Cuda compilation tools, release 11.6, V11.6.55
Build cuda_11.6.r11.6/compiler.30794723_0