[model] add mimo7b

[Kuangdd01]

What does this PR do?

Fixes #7939 (issue)

Just test in chat mode and lora finetune.

Besides, please comment this line https://huggingface.co/XiaomiMiMo/MiMo-7B-SFT/blob/main/modeling_mimo.py#L64 for ddp training because it is not used during training.

Before submitting

• Did you read the contributor guideline?
• Did you write any new necessary tests?

[hiyouga]

See the above comments

[Kuangdd01]

Reasoning data && General data pretraining -> MiMo-7B-Base
MiMo-7B-Base + SFT -> MiMo-7B-SFT
MiMo-7B-SFT + GRPO -> MiMo-7B-RL
MiMo-7B-Base + GRPO -> MiMo-7B-RL-Zero

so did I fix that mismatch? And there is no hard switch like Qwen3 think mode.