Nemo 2.0 Finetuned model export or merge to base model

[kanic1111]

Hello right now i'm testing Finetuning Llama4 with Nemo 2.0 and Lora which do succeed

But right now i can't fine any method to merge it or export to some target like hf
by using llm.export_ckpt which only works for normal model not finetuned model
Is there any tutorial of how to merge the model to be base model so i can use the llm.export_ckpt or any other support method?

[kanic1111]

i has found the Tutorial https://docs.nvidia.com/nemo-framework/user-guide/latest/sft_peft/peft_nemo2.html
but unfortunate Llama4 right now doesn't not support straight convert to hf-peft
So i think only merge model are the avaliable right now