Views
No views yet
1export exclude_layers="lm_head *.self_attn.* *.mlp.gate.*"
2python3 quantize_quark.py --model_dir Qwen/Qwen3-30B-A3B \
3 --quant_scheme mxfp4 \
4 --output_dir Qwen3-30B-A3B-MXFP4 \
5 --exclude_layers $exclude_layers \
6 --model_export hf_format