Required library version not found: libbitsandbytes_cuda124_nocublaslt.so

[MostHumble]

System Info

False

===================================BUG REPORT===================================
/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/site-packages/bitsandbytes/cuda_setup/main.py:167: UserWarning: Welcome to bitsandbytes. For bug reports, please run

python -m bitsandbytes

warn(msg)

/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/site-packages/bitsandbytes/cuda_setup/main.py:167: UserWarning: /home/sklioui/miniconda3/envs/luma_unbias did not contain ['libcudart.so', 'libcudart.so.11.0', 'libcudart.so.12.0'] as expected! Searching further paths...
warn(msg)
/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/site-packages/bitsandbytes/cuda_setup/main.py:167: UserWarning: /trinity/shared/apps/tr17.10/x86_64/cuda-12.4/lib64 did not contain ['libcudart.so', 'libcudart.so.11.0', 'libcudart.so.12.0'] as expected! Searching further paths...
warn(msg)
The following directories listed in your path were found to be non-existent: {PosixPath('/trinity/shared/apps/tr17.10/x86_64/cuda-12.4/doc/man')}
The following directories listed in your path were found to be non-existent: {PosixPath('/usr/lib64/qt-3.3/include')}
The following directories listed in your path were found to be non-existent: {PosixPath('//matplotlib_inline.backend_inline'), PosixPath('module')}
The following directories listed in your path were found to be non-existent: {PosixPath('cuda/12.4'), PosixPath('userspace/all')}
The following directories listed in your path were found to be non-existent: {PosixPath('/run/user/2569/vscode-git-2cb3a4d16d.sock')}
The following directories listed in your path were found to be non-existent: {PosixPath('/run/user/2569/vscode-ipc-c2b58e1d-ebe4-4b31-a2eb-0f592a570bed.sock')}
The following directories listed in your path were found to be non-existent: {PosixPath('() { eval /usr/bin/modulecmd bash $*\n}')}
DEBUG: Possible options found for libcudart.so: {PosixPath('/trinity/shared/apps/tr17.10/x86_64/cuda-12.4/lib64/libcudart.so')}
CUDA SETUP: PyTorch settings found: CUDA_VERSION=124, Highest Compute Capability: 7.0.
CUDA SETUP: To manually override the PyTorch CUDA version please see:https://github.com/TimDettmers/bitsandbytes/blob/main/how_to_use_nonpytorch_cuda.md
/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/site-packages/bitsandbytes/cuda_setup/main.py:167: UserWarning: WARNING: Compute capability < 7.5 detected! Only slow 8-bit matmul is supported for your GPU! If you run into issues with 8-bit matmul, you can try 4-bit quantization: https://huggingface.co/blog/4bit-transformers-bitsandbytes
warn(msg)
CUDA SETUP: Required library version not found: libbitsandbytes_cuda124_nocublaslt.so. Maybe you need to compile it from source?
CUDA SETUP: Defaulting to libbitsandbytes_cpu.so...

================================================ERROR=====================================
CUDA SETUP: CUDA detection failed! Possible reasons:

1. You need to manually override the PyTorch CUDA version. Please see: "https://github.com/TimDettmers/bitsandbytes/blob/main/how_to_use_nonpytorch_cuda.md
2. CUDA driver not installed
3. CUDA not installed
4. You have multiple conflicting CUDA libraries
5. Required library not pre-compiled for this bitsandbytes release!
   CUDA SETUP: If you compiled from source, try again with make CUDA_VERSION=DETECTED_CUDA_VERSION for example, make CUDA_VERSION=113.
   CUDA SETUP: The CUDA version for the compile might depend on your conda install. Inspect CUDA version via conda list | grep cuda.
   ================================================================================

CUDA SETUP: Something unexpected happened. Please compile from source:
git clone https://github.com/TimDettmers/bitsandbytes.git
cd bitsandbytes
CUDA_VERSION=124_nomatmul
python setup.py install
CUDA SETUP: Setup Failed!
Traceback (most recent call last):
File "/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/runpy.py", line 187, in _run_module_as_main
mod_name, mod_spec, code = _get_module_details(mod_name, _Error)
File "/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/runpy.py", line 146, in _get_module_details
return _get_module_details(pkg_main_name, error)
File "/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/runpy.py", line 110, in _get_module_details
import(pkg_name)
File "/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/site-packages/bitsandbytes/init.py", line 6, in
from . import cuda_setup, utils, research
File "/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/site-packages/bitsandbytes/research/init.py", line 1, in
from . import nn
File "/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/site-packages/bitsandbytes/research/nn/init.py", line 1, in
from .modules import LinearFP8Mixed, LinearFP8Global
File "/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/site-packages/bitsandbytes/research/nn/modules.py", line 8, in
from bitsandbytes.optim import GlobalOptimManager
File "/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/site-packages/bitsandbytes/optim/init.py", line 6, in
from bitsandbytes.cextension import COMPILED_WITH_CUDA
File "/home/sklioui/miniconda3/envs/luma_unbias/lib/python3.10/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

Reproduction

python -m bitsandbytes

Expected behavior

True ?

[djbyrne]

I am also seeing this, any ideas on how to fix?

[MostHumble]

I haven't had time to investigate this. For me, it might be caused by one of the following:

• Either the fact that I'm using SLURM with MODULE, and the paths are not being handled correctly
• Or because I'm using a V100, which might not be supported for some reason

[djbyrne]

I am seeing the same on H200 just using kubernetes batch jobs

[matthewdouglas]

V100 is supported, as well as H200. Please upgrade to the latest bitsandbytes release.