Closed enn-nafnlaus closed 9 months ago
I got the same errors. This may be related to this other issue. I managed to fix this by:
Good luck! It reduced the GPU memory required by 3-4X, very useful!
I'm encountering this error when executing make within bitsandbytes. Any ideas?
ptxas info : Used 2 registers, 320 bytes cmem[0] /usr/include/c++/11/bits/std_function.h:435:145: error: parameter packs not expanded with ‘...’: 435 | function(_Functor&& f) | ^ /usr/include/c++/11/bits/std_function.h:435:145: note: ‘_ArgTypes’ /usr/include/c++/11/bits/std_function.h:530:146: error: parameter packs not expanded with ‘...’: 530 | operator=(_Functor&& f) | ^ /usr/include/c++/11/bits/std_function.h:530:146: note: ‘_ArgTypes’
CUDA_SETUP: WARNING! libcudart.so not found in any environmental path. Searching /usr/local/cuda/lib64... D:\Stable\LORa\kohya_ss\venv\lib\site-packages\bitsandbytes\cuda_setup\paths.py:27: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {WindowsPath('/usr/local/cuda/lib64')}
what can I do I have already installed it
Running into the same issue, still debugging. FWIW, I dearly appriciate how thorough the error message is @TimDettmers - Good lookin out for the user!
I am on windows. I had a similar problem trying to run a different model with pytorch but this might help. Ensure you have a supported cuda version with pytorch e.g. 11.8 . And have installed pytorch 11.8 with the anaconda command from the docs. Then run these commands.
1、pip uninstall bitsandbytes 2、pip install bitsandbytes-cuda116 3、https://pypi.org/project/bitsandbytes-cuda116/
I am on windows. I had a similar problem trying to run a different model with pytorch but this might help. Ensure you have a supported cuda version with pytorch e.g. 11.8 . And have installed pytorch 11.8 with the anaconda command from the docs. Then run these commands.
- pip uninstall bitsandbytes
- pip install git+https://github.com/Keith-Hon/bitsandbytes-windows.git
thank u !!!I have been trying to find this version of bitsandbytes, but I haven't found it before.
$ sudo apt install gcc-10 g++-10 $ export CC=/usr/bin/gcc-10 $ export CXX=/usr/bin/g++-10 $ export CUDA_ROOT=/usr/local/cuda $ ln -s /usr/bin/gcc-10 $CUDA_ROOT/bin/gcc $ ln -s /usr/bin/g++-10 $CUDA_ROOT/bin/g++ $ git clone https://github.com/timdettmers/bitsandbytes.git $ cd bitsandbytes $ CUDA_HOME=/usr/local/cuda-11.6 CUDA_VERSION=116 make cuda11x $ python setup.py install I succeeded!
Bitsandbytes was not supported windows before, but my method can support windows.(yuhuang) 1 open folder J:\StableDiffusion\sdwebui,Click the address bar of the folder and enter CMD or WIN+R, CMD 。enter,cd /d J:\StableDiffusion\sdwebui 2 J:\StableDiffusion\sdwebui\py310\python.exe -m pip uninstall bitsandbytes
3 J:\StableDiffusion\sdwebui\py310\python.exe -m pip uninstall bitsandbytes-windows
4 J:\StableDiffusion\sdwebui\py310\python.exe -m pip install https://github.com/jllllll/bitsandbytes-windows-webui/releases/download/wheels/bitsandbytes-0.41.1-py3-none-win_amd64.whl
Replace your SD venv directory file(python.exe Folder) here(J:\StableDiffusion\sdwebui\py310)
This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.
Got an error saying that either CUDA driver is not installed (I run CUDA apps on the GPU regularly, including StableDiffusion), CUDA is not installed (same), or I have multiple conflicting CUDA libraries (I don't). bitsandbytes == 0.35.4.
_BUG REPORT Welcome to bitsandbytes. For bug reports, please submit your error trace to: https://github.com/TimDettmers/bitsandbytes/issues For effortless bug reporting copy-paste your error into this form: https://docs.google.com/forms/d/e/1FAIpQLScPB8emS3Thkp66nvqwmjTEgxp8Y9ufuWTzFyr9kJ5AoI47dQ/viewform?usp=sf_link
CUDA_SETUP: WARNING! libcudart.so not found in any environmental path. Searching /usr/local/cuda/lib64... CUDA SETUP: CUDA runtime path found: /usr/local/cuda/lib64/libcudart.so CUDA SETUP: Highest compute capability among GPUs detected: 8.6 CUDA SETUP: Detected CUDA version 120 CUDA SETUP: TODO: compile library for specific version: libbitsandbytes_cuda120.so CUDA SETUP: Defaulting to libbitsandbytes.so... CUDA SETUP: CUDA detection failed. Either CUDA driver not installed, CUDA not installed, or you have multiple conflicting CUDA libraries! CUDA SETUP: If you compiled from source, try again with
from .autograd._functions import (
File "/scratch/StableDiffusion/AUTOMATIC1111/stable-diffusion-webui/venv/lib64/python3.10/site-packages/bitsandbytes/autograd/_functions.py", line 5, in
import bitsandbytes.functional as F
File "/scratch/StableDiffusion/AUTOMATIC1111/stable-diffusion-webui/venv/lib64/python3.10/site-packages/bitsandbytes/functional.py", line 13, in
from .cextension import COMPILED_WITH_CUDA, lib
File "/scratch/StableDiffusion/AUTOMATIC1111/stable-diffusion-webui/venv/lib64/python3.10/site-packages/bitsandbytes/cextension.py", line 41, in
lib = CUDALibrary_Singleton.get_instance().lib
File "/scratch/StableDiffusion/AUTOMATIC1111/stable-diffusion-webui/venv/lib64/python3.10/site-packages/bitsandbytes/cextension.py", line 37, in get_instance
cls.instance.initialize()
File "/scratch/StableDiffusion/AUTOMATIC1111/stable-diffusion-webui/venv/lib64/python3.10/site-packages/bitsandbytes/cextension.py", line 27, in initialize
raise Exception('CUDA SETUP: Setup Failed!')
Exception: CUDA SETUP: Setup Failed!
make CUDA_VERSION=DETECTED_CUDA_VERSION
for example,make CUDA_VERSION=113
. Exception importing 8bit adam: CUDA SETUP: Setup Failed! WARNING:extensions.sd_dreambooth_extension.dreambooth.train_dreambooth:Exception importing 8bit adam: CUDA SETUP: Setup Failed! Traceback (most recent call last): File "/scratch/StableDiffusion/AUTOMATIC1111/stable-diffusion-webui/extensions/sd_dreambooth_extension/dreambooth/train_dreambooth.py", line 597, in main import bitsandbytes as bnb File "/scratch/StableDiffusion/AUTOMATIC1111/stable-diffusion-webui/venv/lib64/python3.10/site-packages/bitsandbytes/init.py", line 6, inInstalled:
cuda-toolkit-11-config-common-11.8.89-1.noarch libcudnn8-8.0.4.30-1.cuda11.1.x86_64 nvidia-driver-cuda-libs-525.60.13-1.fc36.x86_64 nvidia-driver-cuda-525.60.13-1.fc36.x86_64 cuda-toolkit-12-config-common-12.0.107-1.noarch cuda-toolkit-12-0-config-common-12.0.107-1.noarch cuda-cudart-12-0-12.0.107-1.x86_64 cuda-opencl-12-0-12.0.76-1.x86_64 cuda-nvrtc-12-0-12.0.76-1.x86_64 cuda-libraries-12-0-12.0.0-1.x86_64 cuda-nvprof-12-0-12.0.90-1.x86_64 cuda-nvml-devel-12-0-12.0.76-1.x86_64 cuda-nvdisasm-12-0-12.0.76-1.x86_64 cuda-cccl-12-0-12.0.90-1.x86_64 cuda-cudart-devel-12-0-12.0.107-1.x86_64 cuda-nvvp-12-0-12.0.90-1.x86_64 cuda-nvrtc-devel-12-0-12.0.76-1.x86_64 cuda-opencl-devel-12-0-12.0.76-1.x86_64 cuda-nsight-systems-12-0-12.0.0-1.x86_64 cuda-nsight-compute-12-0-12.0.0-1.x86_64 cuda-profiler-api-12-0-12.0.76-1.x86_64 cuda-nvtx-12-0-12.0.76-1.x86_64 cuda-nvprune-12-0-12.0.76-1.x86_64 cuda-nvcc-12-0-12.0.76-1.x86_64 cuda-nsight-12-0-12.0.78-1.x86_64 cuda-gdb-12-0-12.0.90-1.x86_64 cuda-driver-devel-12-0-12.0.107-1.x86_64 cuda-libraries-devel-12-0-12.0.0-1.x86_64 cuda-visual-tools-12-0-12.0.0-1.x86_64 cuda-documentation-12-0-12.0.76-1.x86_64 cuda-demo-suite-12-0-12.0.76-1.x86_64 cuda-cuxxfilt-12-0-12.0.76-1.x86_64 cuda-cupti-12-0-12.0.90-1.x86_64 cuda-cuobjdump-12-0-12.0.76-1.x86_64 cuda-compiler-12-0-12.0.0-1.x86_64 cuda-sanitizer-12-0-12.0.90-1.x86_64 cuda-command-line-tools-12-0-12.0.0-1.x86_64 cuda-tools-12-0-12.0.0-1.x86_64 cuda-toolkit-12-0-12.0.0-1.x86_64 cuda-toolkit-config-common-12.0.107-1.noarch
Nvidia driver installed from the runtime rather than rpm (necessary - the dependency chains have been a nightmare on FC36).