[SAVE] Missing `generation_config.json` file for post-quantized `qwen2`

[zzh-www]

I notice it will not save the default generation_config.json of qwen2 model after quantization.

[Qubitium]

@zzh-www Thanks for reporting this issue. Please send us the HF or ModelScope link for the Qwen2 model used. We need the exact model repo link to reproduce.

[zzh-www]

I use Qwen/Qwen2.5-3B-Instruct from HF

[Qubitium]

@zzh-www Fixed. Please try to compile and run main again and confirm.