Refactor dynamic LoRA update to fix incorrect handling of variant weight shapes (#7844)

This commit is contained in:
Lifu Huang
2025-07-13 18:36:01 -07:00
committed by GitHub
parent b5dd5e8741
commit e2ed9d049a
10 changed files with 840 additions and 227 deletions

View File

@@ -891,6 +891,8 @@ class ModelRunner:
lora_backend=self.server_args.lora_backend,
tp_size=self.tp_size,
tp_rank=self.tp_rank,
max_lora_rank=self.server_args.max_lora_rank,
target_modules=self.server_args.lora_target_modules,
)
result = self.lora_manager.load_lora_adapters(self.server_args.lora_paths)
if result.success: