[Core] Refactor model loading code (#4097)

2024-04-16 11:34:39 -07:00
parent 05434764cd
commit 69e1d2fb69
67 changed files with 1054 additions and 963 deletions
--- a/vllm/executor/cpu_executor.py
+++ b/vllm/executor/cpu_executor.py
@@ -40,6 +40,7 @@ class CPUExecutor(ExecutorBase):
            scheduler_config=self.scheduler_config,
            device_config=self.device_config,
            cache_config=self.cache_config,
+            load_config=self.load_config,
            local_rank=0,
            rank=0,
            distributed_init_method=distributed_init_method,