Update for Qwen3.6 family

[Wei-Lin-Intel]

Content of this PR for Qwen3.6 family:

1. Add input scale file for Qwen3.6-27B
2. Adjust max conccurency for Qwen3.5-27B server cmd
3. Correct title for Qwen3.5-397B conversion command

[czhu15]

the commit message says it is for qwen3.6, but the data file (scripts/data/qwen3.5-input-scale.tar.gz) is for qwen3.5.
can you double check if the name is correct?
Assume this file is for both 3.5 and 3.6 (Qwen3.5-27B 转FP8权重（Qwen3.6-27B也可使用类似命令）：)

[Wei-Lin-Intel]

the commit message says it is for qwen3.6, but the data file (scripts/data/qwen3.5-input-scale.tar.gz) is for qwen3.5. can you double check if the name is correct? Assume this file is for both 3.5 and 3.6 (Qwen3.5-27B 转FP8权重（Qwen3.6-27B也可使用类似命令）：)

Yes this file is for both 3.5 and 3.6, containing all the versions.

[czhu15]

LGTM