disable prefix-caching for vl model (#2825)

InternLM · Nov 27, 2024 · 3913ead · 3913ead
1 parent b5b3179
commit 3913ead
Showing 1 changed file with 5 additions and 1 deletion.
diff --git a/lmdeploy/api.py b/lmdeploy/api.py
@@ -69,7 +69,11 @@ def pipeline(model_path: str,
             if backend_config is not None else None
         model_path = get_model(model_path, download_dir, revision)
 
-    _, pipeline_class = get_task(model_path)
+    task, pipeline_class = get_task(model_path)
+    if task == 'vlm':
+        if backend_config.enable_prefix_caching:
+            backend_config.enable_prefix_caching = False
+            logger.warning('VLM does not support prefix caching.')
 
     if type(backend_config) is not PytorchEngineConfig:
         # set auto backend mode