support one more mxfp4 variant

[wenhuach21]

Description

for model_name in  "/models/Qwen3-8B"   "/models/Llama-3.1-8B-Instruct"
do
  CUDA_VISIBLE_DEVICES=$device python3 -m auto_round \
  --model_name  $model_name \
  --scheme mxfp4_rceil \
  --data_type mx_fp4_rceil_v2  \
  --act_data_type mx_fp4_rceil_v2 \
  --group_size 32 \
  --format fake \
  --iters 200 \
  --enable_torch_compile \
  --eval_bs $eval_bs \
  --tasks 'lambada_openai,hellaswag,piqa,winogrande,truthfulqa_mc1,openbookqa,boolq,arc_easy,arc_challenge,mmlu,gsm8k' \
  --output_dir "/data2/wenhuach/mxfp4_rceil_tuning" \
  2>&1 | tee -a test_rceil_tuning.txt
done

Type of Change

Bug fix

Related Issues

Fixes or relates to #

Checklist Before Submitting

• My code has been tested locally.
• Documentation has been updated as needed.
• New or updated tests are included where applicable.
• The CUDA CI has passed. You can trigger it by commenting /azp run Unit-Test-CUDA-AutoRound.

[AutoRoundBot]

/azp run Unit-Test-CUDA-AutoRound

[azure-pipelines]

Azure Pipelines successfully started running 1 pipeline(s).

[xin3he]

LGTM