update

Signed-off-by: yiliu30 <[email protected]>

Author: yiliu30

examples/pytorch/nlp/huggingface_models/language-modeling/quantization/auto_round/README.md

@@ -4,6 +4,9 @@
 ```bash
 export MODEL=Qwen/Qwen3-235B-A22B
 ```
+> [!TIP]
+> For quicker experimentation (shorter quantization and evaluation time, lower memory),
+> you can start with the smaller `Qwen/Qwen3-30B-A3B` model before moving to larger variants.
 
 - MXFP8
 ```bash