Merge pull request #157 from runpod-workers/m-c

Enabling model caching.
2025-02-06 11:52:19 -08:00
parent 9e8d9196b0 3f0a20d28e
commit 6fc770415d
1 changed files with 3 additions and 0 deletions
@@ -24,7 +24,10 @@ class vLLMEngine:
    def __init__(self, engine = None):
        load_dotenv() # For local development
        self.engine_args = get_engine_args()
+        if os.getenv("MODEL_CACHE_ENABLE"):
+            self.engine_args.model = f"/runpod/cache/model/{os.getenv('MODEL_NAME')}/main"
        logging.info(f"Engine args: {self.engine_args}")
+        
        self.tokenizer = TokenizerWrapper(self.engine_args.tokenizer or self.engine_args.model, 
                                          self.engine_args.tokenizer_revision, 
                                          self.engine_args.trust_remote_code)