bitsandbytes was compiled without GPU support libbitsandbytes_cuda.so could not be loaded

bitsandbytes loaded its CPU-only build, or the CUDA build it selected does not match the installed toolkit or the GPU architecture. Quantized (8-bit / 4-bit) paths silently fall back or fail outright. This entry explains how to confirm the cause, apply the fix, and separate it from adjacent huggingface failures.

The literal signature is "bitsandbytes was compiled without GPU support libbitsandbytes_cuda.so could not be loaded". It is a environment failure associated with Hugging Face Transformers, Accelerate, and PEFT. The line identifies the failing operation or subsystem, while the surrounding evidence decides whether it is the initiating fault or a downstream symptom.

bitsandbytes loaded its CPU-only build, or the CUDA build it selected does not match the installed toolkit or the GPU architecture. Quantized (8-bit / 4-bit) paths silently fall back or fail outright. The failure becomes visible at this call site because the operation first requires the missing resource, valid state, healthy peer, or correct result. Earlier log lines and a known-good control carry more causal value than the final wrapper exception.

• The workload stops or loses forward progress after emitting "bitsandbytes was compiled without GPU support libbitsandbytes_cuda.so could not be loaded".
• A retry on the same configuration reproduces the failure because the causal state has not changed.
• The outer framework exception can hide the rank, node, allocation, or dependency that failed first.
• Increasing timeouts or reducing workload size can suppress the symptom without correcting the cause.

• Hugging Face Transformers, Accelerate, and PEFT
• production-shaped multi-accelerator workloads
• containerized and bare-metal deployments of the same stack

Apply the following checklist to a small reproduction: each box below is a positive signal that you are looking at this exact failure rather than a sibling in the same taxonomy.

• ✓Find the first occurrence of "bitsandbytes was compiled without GPU support libbitsandbytes_cuda.so could not be loaded" and preserve at least 100 lines before it.
• ✓Identify which rank, node, device, or process emitted the first related warning.
• ✓the usual causes are (a) a wheel built for a different CUDA minor version, (b) CUDA_HOME/LD_LIBRARY_PATH pointing at a different toolkit than the one torch was built against, (c) a GPU newer than the wheel supports, "no kernel image is available" on sm_90/sm_100 means the binary has no cubin for that architecture.
• ✓Repeat the same input after the targeted change and require the signature to disappear.
• ✓Resume from job start only after the control passes.

bitsandbytes was compiled without GPU support libbitsandbytes_cuda.so could not be loaded

Timestamps and exact values vary across runs, but the pattern. An info-level start, an early WARN, an ERROR carrying the symptom. Is the actual fingerprint you should alert on. The Denpex platform flags this combination automatically.

python -m bitsandbytes, the self-test prints which binary it selected and which CUDA version it expects. Compare against nvcc --version and nvidia-smi. This changes the condition in the causal diagnosis instead of hiding the outer exception. The repeated control proves ownership before recovery from job start.

# Preserve evidence before restarting
rg -n -i 'error|exception|timeout|failed' <log-file>
nvidia-smi
python -m torch.utils.collect_env

# Find the exact signature in the complete log
rg -n -F -- "bitsandbytes was compiled without GPU support libbitsandbytes_cuda.so could not be loaded" <log-file>

Adapt the snippet to your framework. The same pattern holds for PyTorch Lightning, Hugging Face Trainer, DeepSpeed, Megatron-LM, and vLLM training wrappers. Where the wrapper exposes a config flag (for examplelr_scheduler_type in Trainer), prefer the flag over the imperative API to keep the schedule declarative and reproducible.

Select the traceback or open the failed terminal, then run Denpex locally to see the initiating rank, collateral failures, exact fix, and verification command without uploading the log.

CUDA reports errors asynchronously, so the traceback usually points at whatever line synchronised next rather than the one at fault. The hub covers every common CUDA error and how to make it report honestly.

• bitsandbytes loaded its CPU-only build, or the CUDA build it selected does not match the installed toolkit or the GPU architecture. Quantized (8-bit / 4-bit) paths silently fall back or fail outright.
• The decisive evidence is the first log line that precedes "bitsandbytes was compiled without GPU support libbitsandbytes_cuda.so could not be loaded" and differs from a healthy run.
• A nearby failure remains a competing hypothesis until a control separates configuration, capacity, transport, and hardware causes.