biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Misc] V1 LoRA support CPU offload (#15843)

Browse Source 
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
This commit is contained in:
Jee Jee Li
2025-04-02 23:04:43 +08:00
committed by GitHub
parent e86c414d6a
commit 58e234a754
1 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
vllm/config.py
View File

@@ -2434,9 +2434,9 @@ class LoRAConfig:
f"max_loras ({self.max_loras})") f"max_loras ({self.max_loras})")
def verify_with_cache_config(self, cache_config: CacheConfig): def verify_with_cache_config(self, cache_config: CacheConfig):
# TODO LoRA supports CPU offload. if cache_config.cpu_offload_gb > 0 and not envs.VLLM_USE_V1:
if cache_config.cpu_offload_gb > 0: raise ValueError(
raise ValueError("CPU offload is not supported with LoRA yet.") "V0 LoRA does not support CPU offload, please use V1.")
def verify_with_model_config(self, model_config: ModelConfig): def verify_with_model_config(self, model_config: ModelConfig):
if self.lora_dtype in (None, "auto"): if self.lora_dtype in (None, "auto"):