fix typo

README.md

@@ -100,10 +100,10 @@ prompt = "Once upon a time,"
 
 pip3 install lm-eval==0.4.7
 
-we found lm-eval is very unstable for this model. Please set `add_bos_token=True `to align with the origin model. Please use autogptq format
+we found lm-eval is very unstable for this model. Please set `add_bos_token=True `to align with the origin model. **Please use autogptq format**
 
 ```bash
-lm-eval --model hf --model_args pretrained=OPEA/Llama-3.3-70B-Instruct-int3-sym-inc,add_bos_token=True   --tasks leaderboard_mmlu_pro,leaderboard_ifeval,lambada_openai,hellaswag,piqa,winogrande,truthfulqa_mc1,openbookqa,boolq,arc_easy,arc_challenge,mmlu,gsm8k --batch_size 16
+lm-eval --model hf --model_args pretrained=OPEA/Llama-3.3-70B-Instruct-int2-sym-inc,add_bos_token=True   --tasks leaderboard_mmlu_pro,leaderboard_ifeval,lambada_openai,hellaswag,piqa,winogrande,truthfulqa_mc1,openbookqa,boolq,arc_easy,arc_challenge,mmlu,gsm8k --batch_size 16
 ```
 |           Metric           |   BF16(lm-eval==0.4.5)   | W2G32 With BOS            | BF16(lm-eval==0.4.7 with BOS) |      WO BOS       |
 | :------------------------: | :----------------------: | ------------------------- | ----------------------------- | :---------------: |