Support Flux quantization

[mengniwang95]

Test cmd:

for model_name in "/mengni/flux/"
do
  CUDA_VISIBLE_DEVICES="7"  python3 -m auto_round \
    --model_name  $model_name \
    --device 0 \
    --bits 4 \
    --data_type "nv_fp" \
    --group_size 16 \
    --iters 100 \
    --nsamples 1 \
    --act_bits 4 \
    --act_data_type "nv_fp4_with_static_gs" \
    --format "llmcompressor" \
    --batch_size 4 \
    --output_dir "/mengni/tmp_flux" \
    --prompt_file "captions_source.tsv" \ # download from https://github.com/mlcommons/inference/raw/refs/heads/master/text_to_image/coco2014/captions/captions_source.tsv
    --metrics "clip,clip-iqa,imagereward" \
    --mllm
done