[Feature]: AutoDeploy: nano 3 custom modeling code follow-ups

[lucaslie]

🚀 The feature, motivation and pitch

see https://nvidia.slack.com/archives/D07R57B36KU/p1764006341019689?thread_ts=1763681902.419839&cid=D07R57B36KU

1. fix leading 0 from d_mlp
2. reset _custom_class_mapping in children of AutoModelForCausalLMFactory
3. switch keying from model_type to type(config).name

Alternatives

No response

Additional context

No response

Before submitting a new issue...

• Make sure you already searched for relevant issues, and checked the documentation and examples for answers to frequently asked questions.