Thank you and cuda issue on rtx 3070

[jvgriethuijsen]

First of all thanks so much for the wonderful work!

I think I've found a bug.
Trying the following gives an error.

model = LlamaForCausalLM.from_pretrained("TheBloke/wizardLM-7B-HF",  load_in_8bit=True,

- CUDA SETUP: WARNING! libcuda.so not found! Do you have a CUDA driver installed? If you are on a cluster, make sure you are on a CUDA machine!

I am on a windows machine (i9 / 32GB / GTX 3070 laptop 8GB vram). I tried installing an older version of cuda toolkit but it didn't work. It also seems to assume I am on Linux somehow?

==============================================================================

pip show bitsandbytes (support cuda 11.8)

Name: bitsandbytes
Version: 0.39.0
Summary: k-bit optimizers and matrix multiplication routines.
Home-page: https:/TimDettmers/bitsandbytes

nvcc --version (is cuda 11.8)

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2022 NVIDIA Corporation
Built on Wed_Sep_21_10:41:10_Pacific_Daylight_Time_2022
Cuda compilation tools, release 11.8, V11.8.89
Build cuda_11.8.r11.8/compiler.31833905_0

Pytorch 2.0.1 (with the same cuda 11.8)

Windows Path:

- added C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\bin
- added C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\libnvvp

==============================================================================

python -m bitsandbytes

===================================BUG REPORT===================================
Welcome to bitsandbytes. For bug reports, please run

python -m bitsandbytes

and submit this information together with your error trace to: https:/TimDettmers/bitsandbytes/issues

bin \env\Lib\site-packages\bitsandbytes\libbitsandbytes_cpu.so
False
CUDA_SETUP: WARNING! libcudart.so not found in any environmental path. Searching in backup paths...
\env\Lib\site-packages\bitsandbytes\cuda_setup\main.py:149: UserWarning: WARNING: The following directories listed in your path were found to be non-existent: {WindowsPath('/usr/local/cuda/lib64')}
warn(msg)
CUDA SETUP: WARNING! libcuda.so not found! Do you have a CUDA driver installed? If you are on a cluster, make sure you are on a CUDA machine!
\env\Lib\site-packages\bitsandbytes\cuda_setup\main.py:149: UserWarning: WARNING: No libcudart.so found! Install CUDA or the cudatoolkit package (anaconda)!
warn(msg)
\env\Lib\site-packages\bitsandbytes\cuda_setup\main.py:149: UserWarning: WARNING: No GPU detected! Check your CUDA paths. Proceeding to load CPU-only library...
warn(msg)
CUDA SETUP: Loading binary \env\Lib\site-packages\bitsandbytes\libbitsandbytes_cpu.so...
argument of type 'WindowsPath' is not iterable
CUDA SETUP: Problem: The main issue seems to be that the main CUDA library was not detected.
CUDA SETUP: Solution 1): Your paths are probably not up-to-date. You can update them via: sudo ldconfig.
CUDA SETUP: Solution 2): If you do not have sudo rights, you can do the following:
CUDA SETUP: Solution 2a): Find the cuda library via: find / -name libcuda.so 2>/dev/null
CUDA SETUP: Solution 2b): Once the library is found add it to the LD_LIBRARY_PATH: export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:FOUND_PATH_FROM_2a
CUDA SETUP: Solution 2c): For a permanent solution add the export from 2b into your .bashrc file, located at ~/.bashrc
Traceback (most recent call last):
File "", line 189, in run_module_as_main
File "", line 148, in get_module_details
File "", line 112, in get_module_details
File "\env\Lib\site-packages\bitsandbytes_init.py", line 6, in
from . import cuda_setup, utils, research
File "\env\Lib\site-packages\bitsandbytes\research_init.py", line 1, in
from . import nn
File "\env\Lib\site-packages\bitsandbytes\research\nn_init.py", line 1, in
from .modules import LinearFP8Mixed, LinearFP8Global
File "\env\Lib\site-packages\bitsandbytes\research\nn\modules.py", line 8, in
from bitsandbytes.optim import GlobalOptimManager
File "\env\Lib\site-packages\bitsandbytes\optim_init_.py", line 6, in
from bitsandbytes.cextension import COMPILED_WITH_CUDA
File "\env\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:/TimDettmers/bitsandbytes/issues

[ToukoH]

From requirements on README:

"The bitsandbytes library is currently only supported on Linux distributions. Windows is not supported at the moment."

[TimDettmers]

This issue might be resolve by upgrading via pip install -U bitsandbytes. However, note that Windows is currently not supported. You can find more information here: #30

[swumagic]

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:/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）****