not running on GPU
geocine opened this issue · 2 comments
When I am running script from https://github.com/pykeio/diffusers/blob/main/examples/stable-diffusion.rs . Its using up my CPU and system Memory. I don't see the GPU being utilized
I have the following setup:
May Cargo.toml
looks like this
[dependencies]
pyke-diffusers = "0.1"
Output of nvidia-smi
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 527.56 Driver Version: 527.56 CUDA Version: 12.0 |
|-------------------------------+----------------------+----------------------+
| GPU Name TCC/WDDM | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 NVIDIA GeForce ... WDDM | 00000000:0A:00.0 On | N/A |
| 0% 50C P5 38W / 220W | 988MiB / 8192MiB | 11% Default |
| | | N/A |
+-------------------------------+----------------------+----------------------+
Output of nvcc -v
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2021 NVIDIA Corporation
Built on Fri_Dec_17_18:28:54_Pacific_Standard_Time_2021
Cuda compilation tools, release 11.6, V11.6.55
Build cuda_11.6.r11.6/compiler.30794723_0
From C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6\include\cudnn_version.h
version is 8.6.0
#define CUDNN_MAJOR 8
#define CUDNN_MINOR 6
#define CUDNN_PATCHLEVEL 0
I have successfully run CUDA workloads on this machine outside of this library
Make sure you've enabled the ort-cuda
Cargo feature. You may have to cargo clean
and rebuild. Sorry, I should've been more clear in the docs.
Thanks, you did mention in the docs to use ort-cuda
. I forgot to uncomment when I was trying to make fp16
work on the other issue I opened. All good now