Add required lora_load_format and lr params to multi-model example

These parameters are now required when lora_rank > 0.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Author: binary-husky

tutorial/example_train_multi_model/trans_roll_lora.py

@@ -51,6 +51,8 @@ def main():
         logging="swanlab",
         lora_rank=32,
         lora_alpha=32,
+        lora_load_format="safetensors",
+        lr=3e-4,
     )
 
     job_7b = AgentJetJob(
@@ -64,6 +66,8 @@ def main():
         logging="swanlab",
         lora_rank=32,
         lora_alpha=32,
+        lora_load_format="safetensors",
+        lr=3e-4,
     )
 
     # Original (sequential) version: