fix(nemotron-omni-cp): use HF model ID in ep8cp2 yaml

Replace local lustre path with the canonical
``nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16`` model ID,
matching the existing ``nemotron_omni_cord_v2.yaml`` and ``_peft.yaml``.

Also resolves a secrets-detector false positive (``Base64 High Entropy
String`` triggered on the long lustre path at line 27).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Signed-off-by: HuiyingLi <willwin.lee@gmail.com>

Author: HuiyingLi

examples/vlm_finetune/nemotron_omni/nemotron_omni_v3_cord_v2_ep8cp2.yaml

@@ -24,7 +24,7 @@ rng:
 
 model:
   _target_: nemo_automodel.NeMoAutoModelForImageTextToText.from_pretrained
-  pretrained_model_name_or_path: /lustre/fs1/portfolios/coreai/users/chcui/nomni-workspace/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16
+  pretrained_model_name_or_path: nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16
   trust_remote_code: true
   torch_dtype: torch.bfloat16
   backend:
@@ -39,7 +39,7 @@ model:
 
 processor:
   _target_: transformers.AutoProcessor.from_pretrained
-  pretrained_model_name_or_path: /lustre/fs1/portfolios/coreai/users/chcui/nomni-workspace/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16
+  pretrained_model_name_or_path: nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16
   trust_remote_code: true
 
 checkpoint: