CUDA setup issue

[MarkDeng1]

System Info

cuda version = 12.2
Python version = 3.9
I'm sure cuda is available, but I cannot run bitsandbytes

Reproduction

The error report is that :

(qlora) D:\working\Swinburne_work\LLMs\LLM Quantization\qlora-main\qlora-main>python -m bitsandbytes
False

===================================BUG REPORT===================================
D:\Anaconda\envs\qlora\lib\site-packages\bitsandbytes\cuda_setup\main.py:167: UserWarning: Welcome to bitsandbytes. For bug reports, please run

python -m bitsandbytes

warn(msg)

The following directories listed in your path were found to be non-existent: {WindowsPath('D')}
D:\Anaconda\envs\qlora\lib\site-packages\bitsandbytes\cuda_setup\main.py:167: UserWarning: D:\Anaconda\envs\qlora did not contain ['libcudart.so', 'libcudart.so.11.0', 'libcudart.so.12.0'] as expected! Searching further paths...
warn(msg)
CUDA_SETUP: WARNING! libcudart.so not found in any environmental path. Searching in backup paths...
The following directories listed in your path were found to be non-existent: {WindowsPath('/usr/local/cuda/lib64')}
DEBUG: Possible options found for libcudart.so: set()
CUDA SETUP: PyTorch settings found: CUDA_VERSION=121, Highest Compute Capability: 8.9.
CUDA SETUP: To manually override the PyTorch CUDA version please see:https://github.com/TimDettmers/bitsandbytes/blob/main/how_to_use_nonpytorch_cuda.md
CUDA SETUP: Loading binary D:\Anaconda\envs\qlora\lib\site-packages\bitsandbytes\libbitsandbytes_cuda121.so...
argument of type 'WindowsPath' is not iterable
CUDA SETUP: Problem: The main issue seems to be that the main CUDA runtime library was not detected.
CUDA SETUP: Solution 1: To solve the issue the libcudart.so location needs to be added to the LD_LIBRARY_PATH variable
CUDA SETUP: Solution 1a): Find the cuda runtime library via: find / -name libcudart.so 2>/dev/null
CUDA SETUP: Solution 1b): Once the library is found add it to the LD_LIBRARY_PATH: export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:FOUND_PATH_FROM_1a
CUDA SETUP: Solution 1c): For a permanent solution add the export from 1b into your .bashrc file, located at ~/.bashrc
CUDA SETUP: Solution 2: If no library was found in step 1a) you need to install CUDA.
CUDA SETUP: Solution 2a): Download CUDA install script: wget https://raw.githubusercontent.com/TimDettmers/bitsandbytes/main/cuda_install.sh
CUDA SETUP: Solution 2b): Install desired CUDA version to desired location. The syntax is bash cuda_install.sh CUDA_VERSION PATH_TO_INSTALL_INTO.
CUDA SETUP: Solution 2b): For example, "bash cuda_install.sh 113 ~/local/" will download CUDA 11.3 and install into the folder ~/local
Traceback (most recent call last):
File "D:\Anaconda\envs\qlora\lib\runpy.py", line 188, in _run_module_as_main
mod_name, mod_spec, code = _get_module_details(mod_name, Error)
File "D:\Anaconda\envs\qlora\lib\runpy.py", line 147, in get_module_details
return get_module_details(pkg_main_name, error)
File "D:\Anaconda\envs\qlora\lib\runpy.py", line 111, in get_module_details
import(pkg_name)
File "D:\Anaconda\envs\qlora\lib\site-packages\bitsandbytes_init.py", line 6, in
from . import cuda_setup, utils, research
File "D:\Anaconda\envs\qlora\lib\site-packages\bitsandbytes\research_init.py", line 1, in
from . import nn
File "D:\Anaconda\envs\qlora\lib\site-packages\bitsandbytes\research\nn_init.py", line 1, in
from .modules import LinearFP8Mixed, LinearFP8Global
File "D:\Anaconda\envs\qlora\lib\site-packages\bitsandbytes\research\nn\modules.py", line 8, in
from bitsandbytes.optim import GlobalOptimManager
File "D:\Anaconda\envs\qlora\lib\site-packages\bitsandbytes\optim_init.py", line 6, in
from bitsandbytes.cextension import COMPILED_WITH_CUDA
File "D:\Anaconda\envs\qlora\lib\site-packages\bitsandbytes\cextension.py", line 20, in
raise RuntimeError('''
RuntimeError:
CUDA Setup failed despite GPU being available. Please run the following command to get more information:

    python -m bitsandbytes

    Inspect the output of the command and see if you can locate CUDA libraries. You might need to add them
    to your LD_LIBRARY_PATH. If you suspect a bug, please take the information from python -m bitsandbytes
    and open an issue at: https://github.com/TimDettmers/bitsandbytes/issues

Expected behavior

using bitsandbytes with cuda

[mwzkhalil]

having same issue...
`RuntimeError: Failed to import transformers.integrations.bitsandbytes because of the following error (look up to see its traceback):

    CUDA Setup failed despite GPU being available. Inspect the CUDA SETUP outputs above to fix your environment!
    If you cannot find any issues and suspect a bug, please open an issue with detals about your environment:
    https://github.com/TimDettmers/bitsandbytes/issues`

[virtualpeer]

having same issue...

[moyuii]

having same issue...

[mwzkhalil]

If the issue occurred while you were preparing the model, please utilize the prepare_model_for_kbit_training function.

[moyuii]

having same issue...

System Info

cuda version = 12.2
Python version = 3.9

compile from source：

mkdir -p build
mkdir -p dependencies
ENVIRONMENT

CUDA_VERSION: 122

NVCC path: /usr/bin/nvcc
GPP path: /usr/bin/g++ VERSION: g++ (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
CUDA_HOME: /usr
CONDA_PREFIX:
PATH: /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin:/snap/bin
LD_LIBRARY_PATH:

/usr/bin/nvcc -gencode arch=compute_75,code=sm_75 -gencode arch=compute_80,code=sm_80 -gencode arch=compute_86,code=sm_86 -gencode arch=compute_89,code=sm_89 -gencode arch=compute_90,code=sm_90 -Xcompiler '-fPIC' --use_fast_math -Xptxas=-v -dc /home/user/tests/bitsandbytes-main/csrc/ops.cu /home/user/tests/bitsandbytes-main/csrc/kernels.cu -I /usr/include -I /home/user/tests/bitsandbytes-main/csrc -I /include -I /home/user/tests/bitsandbytes-main/include -L /usr/lib64 -lcudart -lcublas -lcublasLt -lcusparse -L /lib --output-directory /home/user/tests/bitsandbytes-main/build
nvcc fatal : Unsupported gpu architecture 'compute_89'
make: *** [Makefile:111: cuda12x] Error 1

[Sturmkater]

Have the same issue, but on Linux:-(

[motoko33]

Having the same issue on windows......

[MatthiasKohl]

Please take a look at whether you can work with the instructions described here: #1042

specifically:

cmake -B . -DCOMPUTE_BACKEND=cuda -DCOMPUTE_CAPABILITY=[your compute capability] -S .
cmake --build .

since you're likely running into an issue with some of the compute capabilities not supported for your version of nvcc

[Nethramudhiraj]

If the issue occurred while you were preparing the model, please utilize the prepare_model_for_kbit_training function.

Hi @MatthiasKohl can u elaborate more about 'prepare_model_for_kbit_training` '

[TonyLv]

Have the same issue on Ubutu

[Titus-von-Koeller]

Dear all,

Since the current release (last week, 8th of March) we now have official support for Windows 🎉 (which we did not have before) via

pip install bitsandbytes>=0.43.0

We're closing all old Windows issues and are asking everyone to try installing with this new version as outlined above and validate the install with python -m bitsandbytes which should spit out a bunch of stuff and then SUCCESS. Please let us know if everything worked correctly in this new umbrella / catch-all issue. Thanks 🤗