ExLlamaV2 kernels do not support act-order desc_act with group_size

The ExLlama kernel cannot serve this GPTQ checkpoint: act-order (desc_act=True) combined with a grouped quantization is unsupported by that kernel path. The model is fine; the kernel choice is not. This entry explains how to confirm the cause, apply the fix, and separate it from adjacent quantization failures.

Symptom

ExLlamaV2 kernels do not support act-order desc_act with group_size

Root cause

The ExLlama kernel cannot serve this GPTQ checkpoint: act-order (desc_act=True) combined with a grouped quantization is unsupported by that kernel path. The model is fine; the kernel choice is not. The decisive evidence is the first log line that precedes "ExLlamaV2 kernels do not support act-order desc_act with group_size" and differs from a healthy run.

Recommended fix

disable the ExLlama kernel and fall back to the default GPTQ CUDA kernel (disable_exllama=True in the quantization config). Slower, but correct.

How Denpex helps

Denpex investigates ExLlamaV2 kernels do not support act-order desc_act with group_size using the evidence you provide or your connected workload collects. Earlier rank, host or application evidence is needed to distinguish an initiating failure from a downstream report.

The literal signature is "ExLlamaV2 kernels do not support act-order desc_act with group_size". It is a environment failure associated with AWQ, GPTQ, bitsandbytes, GGUF, and torchao. The line identifies the failing operation or subsystem, while the surrounding evidence decides whether it is the initiating fault or a downstream symptom.

The ExLlama kernel cannot serve this GPTQ checkpoint: act-order (desc_act=True) combined with a grouped quantization is unsupported by that kernel path. The model is fine; the kernel choice is not. 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 "ExLlamaV2 kernels do not support act-order desc_act with group_size".
• 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.

• AWQ, GPTQ, bitsandbytes, GGUF, and torchao
• production-shaped multi-accelerator workloads
• containerized and bare-metal deployments of the same stack

Use this checklist to test the hypothesis against a small reproduction. No single line proves the root cause, so preserve the preceding events and compare one variable at a time.

• ✓Find the first occurrence of "ExLlamaV2 kernels do not support act-order desc_act with group_size" and preserve at least 100 lines before it.
• ✓Identify which rank, node, device, or process emitted the first related warning.
• ✓read the checkpoint's quantize_config.json, desc_act and group_size together determine which kernels can serve it. desc_act=True with group_size=-1 IS supported; the incompatibility is specifically the grouped case.
• ✓Repeat the same input after the targeted change and require the signature to disappear.
• ✓Resume from no checkpoint action; serving-side configuration only after the control passes.

• The ExLlama kernel cannot serve this GPTQ checkpoint: act-order (desc_act=True) combined with a grouped quantization is unsupported by that kernel path. The model is fine; the kernel choice is not.
• The decisive evidence is the first log line that precedes "ExLlamaV2 kernels do not support act-order desc_act with group_size" and differs from a healthy run.
• A nearby failure remains a competing hypothesis until a control separates configuration, capacity, transport, and hardware causes.

ExLlamaV2 kernels do not support act-order desc_act with group_size

Use this text as a lookup key in logs and upstream issue trackers. It is not presented as a captured customer log. Confirm the cause from your own preceding events, versions, configuration and the cited references.

disable the ExLlama kernel and fall back to the default GPTQ CUDA kernel (disable_exllama=True in the quantization config). Slower, but correct. This changes the condition in the causal diagnosis instead of hiding the outer exception. The repeated control proves ownership before recovery from no checkpoint action; serving-side configuration.

# 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 -- "ExLlamaV2 kernels do not support act-order desc_act with group_size" <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.